Nonlinear Control Model of Synchronous Motor with Excitation and Damper Windings
重复控制在永磁电机低速控制系统中的设计
重复控制在永磁电机低速控制系统中的设计摘要:随着电力电子技术、微电子技术、电机控制理论以及稀土永磁材料的快速发展,永磁同步电机(P M S M )得以迅速使用。
除了具有一般同步电机的工作特性以外,永磁同步电机还具有效率高、结构简单、转动惯量小、维修性好等特点。
因此,广泛应用于柔韧性制造系统、工业机器人、办公自动化、数控机床以及航空航天等领域。
由于永磁同步电机是一个非线性、多变量、强藕荷的系统,采用传统的 P I D 控制方法很容易受电机参数变化和负载扰动等不确定因素的影响。
为了提高控制系统的动态和静态特性,可以采用新型控制理论和智能控制理论代替 PID 控制,比如滑模控制、神经网络控制、模糊控制等。
本文将分析重复控制用于永磁电机低速控制系统中的设计。
关键词:重复控制;永磁电机;低速控制系统;设计原理Abstract:with the rapid development of power electronic technology, microelectronics technology, motor control theory and rare earth permanent magnet materials, the permanent magnet synchronous motor (P M S M) can be used quickly. In addition to the general characteristics of synchronous motor, permanent magnet synchronous motor has the advantages of high efficiency, simple structure, small inertia, good maintainability and so on. Therefore, it is widely used in flexible manufacturing systems,industrial robots, office automation, CNC machine tools and aerospace, etc.. Because of the permanent magnet synchronous motor is a nonlinear system, multivariable, strong coupling, using the traditional P I D control method is easily affected by motor parameter variations and load disturbances and other uncertain factors. In order to improve the dynamic and static characteristics of the control system, the new control theory and intelligent control theory can be used instead of PID control, such as sliding mode control, neural network control, fuzzy control and so on. In this paper, the repetitive control is used to design the low speed control system of permanent magnet motor.Key words: repetitive control; permanent magnet motor; low speed control system; design principle目录摘要 (I)关键词 (I)Abstract: (I)一、绪论 (1)(一)课题研究的背景及意义 (1)(二)永磁同步电机控制系统的国内外研究现状 (1)二、重复控制系统 (2)(二)重复控制发展现状 (2)(二)重复控制研究 (3)三、永磁同步电机低速控制系统 (4)四、重复控制在永磁电机低速控制系统中的设计 (5)(一)重复控制方法是通过内部模型作为基础而运行 (5)(二)仿真波形分析 (6)结论 (7)参考文献 (8)致谢 (9)一、绪论(一)课题研究的背景及意义一个国家的综合实力如何,可以通过观察其国家的航空航天事业的水平来确定。
基于非线性建模与拟合的永磁同步电机转子初始位置精确估计方法
第28卷㊀第2期2024年2月㊀电㊀机㊀与㊀控㊀制㊀学㊀报Electri c ㊀Machines ㊀and ㊀Control㊀Vol.28No.2Feb.2024㊀㊀㊀㊀㊀㊀基于非线性建模与拟合的永磁同步电机转子初始位置精确估计方法姚培煜1,㊀冯国栋1,㊀吴轩2,㊀彭卫文1,㊀丁北辰3(1.中山大学智能工程学院,广东深圳518107;2.湖南大学电气与信息工程学院,湖南长沙410082;3.中山大学先进制造学院,广东深圳518107)摘㊀要:针对永磁同步电机转子初始位置估计的精度与收敛速度受限问题,提出一种基于高频信号注入的非线性建模与拟合实现的初始位置估计方法㊂首先,建立初始位置与高频信号响应的关联模型,表明高频响应可用于直接计算初始位置,但直接计算结果在大部分转子位置易受测量噪声的影响㊂为此,提出基于多项式模型建立位置估计非线性模型,选取合适的模型参数,利用少量测试点拟合该模型,即可实现初始位置的快速精确估计,有效提高了估计精度与系统抗干扰能力㊂实验与仿真结果表明,相比现有方法,提出的方法易于实现,无需复杂滤波器与观测器设计,仅需要选取少量测试点即可快速估计精确转子初始位置,在保证估计精度的同时改进了传统估计方法收敛速度慢问题㊂关键词:永磁同步电机;高频信号注入;转子初始位置估计;多项式模型;非线性模型DOI :10.15938/j.emc.2024.02.014中图分类号:TM351文献标志码:A文章编号:1007-449X(2024)02-0142-10㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀收稿日期:2022-09-24基金项目:国家自然科学基金(52105079,62103455)作者简介:姚培煜(1999 ),男,硕士研究生,研究方向为永磁同步电机无位置传感控制;冯国栋(1988 ),男,博士,副教授,硕士生导师,研究方向为新能源汽车电机系统控制关键技术;吴㊀轩(1983 ),男,博士,副教授,研究方向为电力电子与电力传动㊁大型风力发电技术㊁特种车辆电驱动技术;彭卫文(1987 ),男,博士,副教授,研究方向为系统可靠性㊁智能系统的状态监测㊁故障预测与健康管理;丁北辰(1990 ),男,博士,副教授,研究方向为机器人控制与新能源汽车动力系统控制㊂通信作者:丁北辰High precision initial rotor position estimation method for permanent magnet synchronous motor based on nonlinear modeling and fittingYAO Peiyu 1,㊀FENG Guodong 1,㊀WU Xuan 2,㊀PENG Weiwen 1,㊀DING Beichen 3(1.School of Intelligent Systems Engineering,Sun Yat-sen University,Shenzhen 518107,China;2.College of Electrical and Information Engineering,Hunan University,Changsha 410082,China;3.School of Advanced Manufacturing,Sun Yat-sen University,Shenzhen 518107,China)Abstract :Aiming at the problem that the accuracy and convergence speed of rotor initial position estima-tion of permanent magnet synchronous motor are limited,a nonlinear modeling and fitting method basedon high-frequency signal injection was proposed.Firstly,the correlation model between the initial posi-tion and the high-frequency signal response was established,which shows that the high-frequency re-sponse can be used to calculate the initial position directly,but the direct calculation results are vulnera-ble to the measurement noise in most rotor positions.To solve this issue,a polynomial model was used toestablish the nonlinear model of location estimation,suitable model parameters were selected and a few oftest points were used to fit the polynomial model to achieve rapid and accurate calculation of the initialposition,which effectively improves the estimation accuracy and anti-interference ability of the system. The experimental and simulation results show that compared with the existing methods,in the proposed method it is easy to implement,complex filter and observer design is not needed,and only a few test points need to be selected to quickly estimate the initial position of the precise rotor,which ensures the estimation accuracy and improves the problem of slow convergence of the traditional estimation methods. Keywords:permanent magnet synchronous motor;high frequency signal injection;initial rotor position es-timation;polynomial model;nonlinear model0㊀引㊀言永磁同步电机(permanent magnet synchronous motor,PMSM)因其结构简单,高效率,高能量密度等优点而被广泛应用于新能源汽车等多个领域[1-3]㊂对于永磁同步电机伺服系统,转子初始位置是保证电机启动性能的重要参数㊂具体而言,精确的初始位置能够提高电机控制性能,若初始位置误差过大,会降低启动性能,甚至会导致电机反转与启动失败[4-6]㊂转子位置可通过光电编码器,旋转变压器等获取,但增加了系统成本和体积,在低成本应用如家用电器以及超高速电机应用中,无位置传感控制技术被广泛应用㊂初始位置估计是无位置传感控制的重要环节,可有效地提高系统启动与控制的可靠性㊂因此,转子初始位置估计对永磁同步电机伺服系统十分关键㊂转子初始位置估计在文献中已有广泛研究㊂其中,利用电感饱和效应是近年来解决转子初始位置估计的重要手段,可分为脉冲电压法[7-10],高频信号注入法[11-23]㊂脉冲电压法通过注入一系列脉冲电压矢量,利用电流响应估计转子位置㊂然而,脉冲电压注入可导致转子转动,且过程耗时长㊂高频信号注入法实现简单,无需电机参数和额外硬件,可分高频旋转电压注入[11-16]和高频脉振电压注入[17-23]㊂高频旋转电压注入法依赖于转子凸极效应,且需要通过坐标变换和滤波器提取转子位置㊂文献[11]对高频电流响应进行低通滤波,根据电流幅值随转子位置变化实现转子位置估计㊂文献[14]对三相高频电流正㊁负序分量分离,利用任意一相正负序相角差估计转子位置㊂文献[15]分析了旋转高频注入方法受采样㊁滤波器的影响,并提出一种补偿算法提高位置观测精度㊂高频脉振电压注入法对凸极性要求不高,适用于表贴式电机㊂文献[17]针对相移问题,改用交直轴响应电流解调去除高频分量㊂文献[18]通过对虚拟直轴施加高频电压产生一系列振动信号实现初始位置估计㊂但该方法需要振动传感器,且在转动惯量较大的应用中,需要较大电流诱导转子振动㊂文献[20]在脉振注入基础上引入载波频率成分法判断磁极极性,避免二次信号注入,简化了实现步骤㊂现有高频信号注入估计方法大多通过滤波环节分离高频信号,再通过观测器估计转子初始位置㊂但滤波器对高频信号的幅值和相位产生影响,限制了系统带宽,无法同时保证转子位置的辨识精度和辨识速度㊂同时,观测器的设计也依赖高频信号响应和电机参数㊂针对以上问题,本文提出一种基于高频信号注入的非线性建模与拟合方法,实现转子初始位置估计㊂在虚拟直轴注入高频信号,解调高频电流响应即可获得初始位置,但易受转子所在位置的影响㊂在此基础上,提出基于非线性建模的初始位置估计方法,利用少数测试对非线性模型辨识,实现对转子位置的精确估计㊂此方法无需复杂滤波器和观测器设计,避免相位偏移和收敛速度慢等问题㊂此外,采用测试点快速拟合估计模型有效提高初始位置估计精度和收敛速度㊂仿真与实验结果验证提出方法的有效性㊂1㊀高频信号注入建模永磁同步电机d-q轴电压方程可表示为:u d=Ri d+L dd i dd t-ωL q i q;u q=Ri q+L qd i qd t+ωL d i d+ωλ0㊂üþýïïïï(1)式中:u d/q㊁i d/q和L d/q分别表示d-q轴电压㊁电流和电感;λ0是永磁磁链;R是绕组电阻;ω是电角速度㊂对应的高频信号注入模型可表示为:u dh=R h i dh+L dhd i dhd t;u qh=R h i qh+L qhd i qhd t㊂üþýïïïï(2)341第2期姚培煜等:基于非线性建模与拟合的永磁同步电机转子初始位置精确估计方法式中下标h 表示高频分量㊂例如L dh /qh 表示高频电感,R h 表示高频电阻,初始转速为0㊂不失一般性,假设电机转子的初始位置为θ0㊂定义一个虚拟d -q 轴,其虚拟d 轴的位置为θv ,而θ0和θv 间的误差定义为Δθ=θv -θ0,虚拟d -q 轴与真实d -q 轴的关系如图1所示㊂图1㊀虚拟d -q 轴与真实d -q 轴的关系Fig.1㊀Relationship between virtual and actualdq-axis为估计初始位置θ0,将高频电压信号注入虚拟d 轴,可表达为u dh,v =V dh cos(ωh t )㊂(3)式中:u dh,v 表示高频电压;V dh 为幅值;ωh 为频率㊂基于旋转变换可得注入实际d 轴的高频电压信号为:u dh =u dh,v cosΔθ;u qh =u dh,vsinΔθ㊂}(4)式中u dh 和u qh 为注入到真实d -q 轴的高频电压㊂将式(3)和式(4)代入式(2)可得d -q 与α-β轴下的高频电流响应为:㊀i dh =I dd sin(ωh t -φd )cosΔθ;i qh=I dqsin(ωht -φq)sinΔθ㊂}(5)㊀i αh =I dd sin(ωh t +φd )cosΔθcos θ0-I dq sin(ωh t +φq )sinΔθsin θ0;i βh =I dd sin(ωh t +φd )cosΔθsin θ0+I dqsin(ωht +φq)sinΔθcos θ0㊂üþýïïïï(6)㊀I dd =V dh Z dh ;I dq =V dh Z qh;Z 2dh =R 2h +ω2h L 2dh ;Z 2qh =R 2h +ω2h L 2qh ;tan φd =R h ωh L dh ;tan φq =R h ωh L qh㊂üþýïïïïïï(7)式中i αh 和i βh 可由abc 相电流计算获取㊂对α-β轴高频电流进行如下运算,即:M αs ≜avg(i αh sin ωh t )=I 1cosΔθcos θ0-I 2sinΔθsin θ0;M βs≜avg(i βhsin ωht )=I 1cosΔθsin θ0+I 2sinΔθcos θ0㊂}(8)式中: avg(x ) 表示x 在一个或多个周期内的平均值(例如信号x 的5个周期),I 1和I 2表示如下:I 1=0.5I dd cos φd ;I 2=0.5I dq cos φq ㊂}(9)2㊀转子初始位置直接计算2.1㊀高频注入直接计算法原理式(8)存在3个未知数,至少需要两组数据确定θ0㊂为此,将高频信号分别注入2个虚拟d 轴,对应位置分别为θv0和θv1,其中:1)将V dh0cos(ωh0)注入虚拟d 轴θv0,得到i αh0和i βh0;2)将V dh1cos(ωh1)注入虚拟d 轴θv1,得到i αh1和i βh1㊂基于式(8)以及i αh i 和i βh i ,i =0㊁1,可得:M αs0=I 1cos(θv0-θ0)cos θ0-I 2sin(θv0-θ0)sin θ0;M βs0=I 1cos(θv0-θ0)sin θ0+I 2sin(θv0-θ0)cos θ0;M αs1=I 1cos(θv1-θ0)cos θ0-I 2sin(θv1-θ0)sin θ0;M βs1=I 1cos(θv1-θ0)sin θ0+I 2sin(θv1-θ0)cos θ0㊂üþýïïïï(10)不难看出,基于式(10)可直接计算转子初始位置,定义计算出的位置为θr ㊂特别地,当选择虚拟位置满足θv0=0和θv1=π/2时,θr 可表示为:2θr =arccos(cos2θ0),sin2θ0ȡ0;2π-arccos(cos2θ0),sin2θ<0㊂{(11)其中:sin2θ0=B2C -A 2;cos2θ0=DA 2C -A 2㊂üþýïïïï(12)A =M αs0+M βs1=I 1+I 2;B =2M αs1=(I 1-I 2)sin2θ0;C =M 2αs0+M 2βs1+2M 2αs1=I 21+I 22;D =M 2αs0-M 2βs1=(I 21-I 22)cos2θ0㊂üþýïïïïï(13)图2给出了直接计算法的实施流程,高频信号依次注入得到α-β轴高频电流响应,通过式(10)~式(13)计算出转子初始位置的估计值θr ,最后使用短脉冲注入方法辨识转子磁极极性[24]㊂2.2㊀直接计算法估计误差分析不难看出直接计算法的估计误差与高频信号注入的虚拟位置θv0与θv1相关㊂定义直接计算法的估计误差为Δθe =θr -θ0㊂本节研究θv0与θv1的选择与441电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀估计误差Δθe 的关系,指导θv0与θv1的选择㊂图2㊀直接计算法框图Fig.2㊀Block diagram of direct calculation method2.2.1㊀虚拟位置θv0和θv1选择与误差Δθe 的关系直接计算法是将式(3)中的高频信号分别注入虚拟位置θv0和θv1,获得α-β轴高频响应,对其进一步处理得方程组(10),包含3个未知量,利用数值计算可获得估计结果㊂图3为分别在2个转子初始位置θ0下选择任意不同θv0和θv1时,直接计算法估计误差的分布图,图中每个误差点都是在噪声强度为30dB 仿真环境下2000次随机试验的平均值㊂下文若无特别说明,仿真环境中的噪声强度统一为30dB㊂不难看出,当θv0和θv1越接近,Δθe 越大;当θv0=θv1时,式(10)中的方程式个数变为2个,方程组无解;当θv0和θv1的差值越大,估计误差受噪声影响越小㊂θv0和θv1分别取0和π/2时估计误差相对最小㊂图3㊀不同θv0和θv1的估计误差分布Fig.3㊀Estimation error distributions of different θv0and θv12.2.2㊀不同转子位置的误差Δθe 分析本节探讨转子在不同初始位置直接计算法的估计误差㊂图4给出了不同转子位置的估计误差㊂其中,虚拟位置设置为θv0=0和θv1=π/2;每个误差点都是对同一位置2000次随机试验的平均值㊂可以看出θ0在[0,π]上的估计误差Δθe 呈现三角函数规律变化,在θ0=0㊁π/2㊁π/4附近时θr 的误差Δθe 较小,最小误差约为0.01rad,而在θ0=π/4㊁3π/4附近时θ0的误差Δθe 非常大,最大误差为0.063rad,最大误差是最小误差的6倍以上㊂导致误差呈三角函数规律变化的原因如下:在式(10)中噪声来源于M αs 和M βs ,而在使用式(10)求解θr 时,对cos2θ0进行反三角变化求解θr ㊂对式(10)等式右边变换拆解,提取含有cos2θ0的部分为:S αs =0.5(cos θv (I 1-I 2)cos2θ0)M αs ;S βs =0.5(sin θv (I 1-I 2)cos2θ0)M βs㊂üþýïïïï(14)式中:S αs 和S βs 可以近似表示信号与噪声的比例,即信噪比(signal to noise ratio,SNR)㊂当θ0接近π/4㊁3π/4时,cos2θ0趋于0,S αs 和S βs 趋于0㊂θ0趋于0㊁π/2㊁π时,cos2θ0趋于1,S αs 和S βs 远大于0㊂即Δθe 随着cos2θ0变化而波动㊂不难发现,由于测量噪声的存在,基于式(10)的直接计算法的估计误差在不同转子位置的波动非常大,特别是转子位置在π/4㊁3π/4附近的估计误差比最小误差增加了6倍㊂因此,本文提出基于非线性建模与拟合的方法估计初始位置,提高估计精度和降低估计误差的波动㊂图4㊀直接计算法在不同转子位置的误差变化Fig.4㊀Error variation of direct calculation method atdifferent rotor positions3㊀基于非线性建模与拟合的初始转子位置估计3.1㊀基于多项式建模与曲线拟合的估计方法基于式(8),定义M s ≜M 2αs +M 2βs =I 22+(I 21-I 22)cos 2(θv -θ0)㊂(15)541第2期姚培煜等:基于非线性建模与拟合的永磁同步电机转子初始位置精确估计方法式中M s 以虚拟d 轴位置θv 为自变量的函数,且M s在θv 满足下式时取最大值:Δθ=θv -θ0=0or π㊂(16)如图5所示,考虑在一个周期内,函数M s (θv )在θv <θ0时递增,在此处后递减,这表明转子初始位置θ0可在函数曲线M s (θv )的最大值处得到㊂图5㊀θ0=π/2时M s (θv )曲线Fig.5㊀Curve of M s (θv )at θ0=π/2考虑到直接计算法受测量噪声影响较大,本文提出利用多项式函数对M s (θv )建模,进而在M s (θv )的最大值处确定初始位置θ0㊂不失一般性,本文使用k 阶多项式对M s (θv )建模,即M s (θv )=a k θk v +a k -1θk -1v+ +a 1θv +a 0㊂(17)式中a 0, ,a k -1,a k 为k 阶多项式的系数,可通过曲线拟合估计㊂当a 0, ,a k -1,a k 确定,初始位置θ0可以通过求解下式获得:d M s (θv )d θv =ka k θk -1v +(k -1)a k -1θk -2v+ +2a 2θv +a 1=0㊂(18)当k =2或3时,θ0的估计为:θ0=-a 12a 2,k =2;-a 2ʃa 22-3a 3a 13a 3ɪ[0,π2],k =3㊂ìîíïïïï(19)综上,基于提出的初始位置估计分为两步:第一步:设置N 个虚拟d 轴位置,注入高频测试信号并采集数据用于拟合M s (θv );第二步:基于最小二乘估计a 0, ,a k -1,a k ,并用式(19)计算初始位置θr ㊂图6给出了第一步的图解,假设N 个虚拟d 轴位置为{θv1,θv2, ,θv N },通过电流计算获得{M s1,M s2, ,M s N }㊂基于上述数据与最小二乘法拟合的多项式系数可表示为a =(ϕT ϕ)-1ϕT M ㊂(20)式中:a =[a 0,a 1, ,a k ]T ;ϕ=θk v1θk -1v1θv11θk v2θk -1v2 θv21︙︙︙︙θk v N θk -1v N θv N 1éëêêêêêêùûúúúúúú;M =[M s1,M s2, ,M s N ]T ㊂üþýïïïïïïïïï(21)图6㊀第一步的步骤图Fig.6㊀Diagram of the first step图7给出了此方法的实施框图㊂定义测试点固定间距为θL ,高频电压信号依次注入d 轴虚拟位置θv i =θv i -1+θL ,i =1, ,N ㊂采集α-β轴电流响应,利用式(15)计算M s (θv )用于建模与拟合,利用式(19)计算初始位置θ0㊂图7㊀拟合估计法框图Fig.7㊀Block diagram of fitting estimation method3.2㊀多项式模型参数选择首先,讨论如何选择合适的参数k ㊂一般选择k =2~4可满足估计精度要求㊂考虑到实际环境中的测量噪声,图8为使用不同阶次的多项式拟合M s (θv )㊂从表1不难发现,曲线拟合误差随着k 的增加而越小,但在θ0附近使用二阶多项式拟合即可实现较好的拟合精度㊂641电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀表1㊀不同阶次多项式的拟合精度比较Table 1㊀Comparison of fitting precision between differentorder polynomials参数转子位置/rad 拟合误差/rad真实位置θ00.7854 二阶多项式0.83080.0454三阶多项式0.82730.0419四阶多项式0.82560.0402图8㊀不同阶次多项式拟合M s (θv )Fig.8㊀Fitting M s (θv )with different order polynomials拟合k 次多项式最少需要k +1个拟合点,即N ȡk +1㊂其次,研究如何选取合适的虚拟位置{θv1,θv2, ,θv N },保证初始位置估计精度㊂图9给出了选择k =2㊁N =3㊁4㊁5时的估计误差㊂从图9中不难发现拟合点数量N =5较N =4拟合精度提升并不明显,但需要增加测试点;而N =4较于N =3估计精度有显著提高,且N =4对应的估计精度已满足应用需求㊂综合实现复杂度与估计精度要求,本文选择N =4个拟合点实现多项式模型的拟合㊂图9㊀不同拟合点数量的估计误差Fig.9㊀Estimation error between different number offitting points直接计算法估计的θr 可用于确定一个θ0的粗略分布区域㊂假定θ0=π/4㊁k =2㊁N =4㊂分别在区间R 1=[0,π/2]㊁R 2=[π/8,3π/8]和R 3=[3π/16,5π/16]内随机选取拟合点进行曲线拟合估计,表2是进行2000次随机实验的平均误差,表明通过θr 确定一个合适的区间可以有效地提高估计精度㊂表2㊀不同拟合点选取区间的拟合精度比较Table 2㊀Comparison of fitting precision between differentselection interval of fitting points参数转子位置/rad 拟合误差/rad 真实位置θ00.7854R 10.95280.1674R 20.95680.1714R 30.89680.1114M s (θv )曲线在峰值附近以峰值为中心左右对称,因此在两侧对称选取拟合点能有效提高拟合效果㊂考虑到估计的θr 接近峰值位置,因此本文选择在θr 左右对称地选取拟合点㊂具体而言,首先确定左侧第一个拟合点,其次在当前位置叠加θL 确定下一拟合点位置,该过程可表示为θ2=θ1+θL , ,θN =θN -1+θL ㊂(22)式中θL 对拟合结果有显著影响㊂假定θ0=π/4㊁k =2㊁N =4,图10给出了选择不同θL 时估计误差的变化曲线㊂不难看出,选择θL =0.558rad 估计误差最小㊂综上,本文选择二阶多项式四点拟合,其中拟合点以直接计算值θr 左右对称等间距θL =0.558rad 选取㊂图10㊀不同拟合点间距的估计误差Fig.10㊀Estimation error under different θL3.3㊀多项式曲线拟合法仿真实验本节通过仿真结果验证提出方法的有效性㊂上文分析得出k 阶多项式参数k =2㊁N =4以及拟合点741第2期姚培煜等:基于非线性建模与拟合的永磁同步电机转子初始位置精确估计方法间距选择θL =0.558rad,具有较高的估计精度,下文仿真实验都将使用此模型参数㊂图11是假定初始位置θ0=π/4时,分别使用直接计算法和拟合估计法进行2000次随机实验的估计误差分布㊂不难发现,相比于直接计算法,曲线拟合估计法在同一转子位置上的估计误差和误差波动都更小㊂图11㊀2000次随机实验的估计误差分布Fig.11㊀Estimated error distributions for 2000randomized tests图12为使用高频注入直接计算法和曲线拟合估计法在不同转子位置上的估计误差比较,图12(a)㊁(b)分别为30dB 和40dB 测量噪声下的结果㊂图中每点都是进行了2000次实验的平均估计误差㊂可以发现在θ0=π/4㊁3π/4附近的大部分区域,拟合误差远小于直接计算误差,差值最大的位置拟合误差较直接计算误差减小了0.0352rad,减小了56%㊂另外,对比不同噪声强度环境可以发现,曲线拟合估计法在不同噪声强度下都能够保持较大幅度的估计精度提升㊂曲线拟合法在超过80%的转子位置上估计误差小于直接计算法,在一些位置误差能减小50%以上㊂但在θ0=0㊁π/2㊁π附近其余20%的位置上,因信噪比较大,直接计算法估计误差小于曲线拟合法㊂因此在一个电角度周期内,可以采用两种方法混合估计,当θ0在0㊁π/2㊁π附近小部分区域时令θr 为最终估计结果,否则进一步实施拟合方法估计初始位置,如表3所示㊂图12㊀不同转子位置上估计误差对比Fig.12㊀Comparison of estimated errors between differ-ent rotor positions表3㊀不同转子位置上3种方法的区别Table 3㊀Difference of three methods between differentrotor positions方法θ0在0㊁π/2㊁π附近其他位置直接计算法直接计算直接计算拟合估计法拟合估计拟合估计混合估计法直接计算拟合估计在所有位置上,θr 的平均误差为0.0432rad,拟合θ0的平均误差为0.0268rad,混合估计法可使平均误差进一步减小到0.0248rad㊂整体估计精度提高40%,且拟合估计值的误差波动更小㊁更平稳㊂4㊀实验验证在图13所示的PMSM 样机实验平台上验证本文所提出的方法㊂实验电机的设计参数如表4所示㊂测试电机配备高分辨率光学编码器,单转脉冲数(PPR)为2500㊂从该编码器测量的转子位置将被用来评估提出估计方法的性能,不参与实际控制㊂在实验平台验证方法过程中,电机的转速与转矩都为0㊂注入高频信号的参数为:注入信号频率ωh =150Hz,注入信号幅值V dh =20V㊂选择的非线性模型参数为:k =2㊁N =4㊁θL =0.558rad㊂图14出了使用此参数对M s (θv )进行建模估计θ0的例子㊂841电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀图13㊀实验装置Fig.13㊀Experimental device 表4㊀实验电机的设计参数Table 4㊀Design parameters of experimental motor图14㊀实验验证的拟合估计法例子Fig.14㊀Examples of fitting estimation method verifiedby experiment首先,实验一在不同转子位置进行实验以评估提出估计方法的效果㊂图15(a)给出了电机一个电角度周期内8个位置的估计结果,不难发现估计结果与真实位置十分接近,具体误差分布见图15(b)㊂从图15可以看出,一个电角度周期内,最大拟合误差0.0412rad,最小拟合误差0.0035rad,平均拟合误差约为0.018rad㊂结果表明,曲线拟合估计法能精确估计转子初始位置㊂其次,实验二对比直接计算法与拟合估计法的实验结果㊂直接计算法从α-β轴高频响应电流计算转子初始位置,曲线拟合估计法采用二阶多项式四点非线性建模与拟合估计转子位置㊂估计结果对比如图16(a)所示,2种方法的估计误差对比如图16(b)所示㊂可以看出,直接计算法的平均估计误差为0.034rad,最大估计误差0.114rad,拟合估计的平均拟合误差为0.016rad,最大拟合误差0.042rad㊂实验证明提出的方法相比于传统高频注入法大幅提升了估计精度,降低了误差波动㊂图15㊀实验一的转子初始位置估计结果Fig.15㊀Rotor initial position estimation results inexperiment 1图16㊀实验二的转子初始位置估计结果比较Fig.16㊀Comparison of rotor initial position estimationresults in experiment 25㊀结㊀论本文提出一种基于高频注入的非线性建模与拟合的转子初始位置估计方法,并通过仿真和实验验941第2期姚培煜等:基于非线性建模与拟合的永磁同步电机转子初始位置精确估计方法证提出方法的有效性㊂提出的方法利用少数测试点对位置估计非线性模型快速拟合,实现简单,不依赖电机参数,无需复杂滤波器和观测器的设计㊂实验结果表明,最大误差小于0.05rad,平均误差小于0.02rad㊂与现有方法相比,提出的方法具有估计精度高,收敛速度快,易于实现等优势,工程实用价值高㊂此外,该方法同样在无位置传感器控制技术上有潜在的应用前景㊂参考文献:[1]㊀SHOU W,KANG J,DEGANO M,et al.An accurate wide-speedrange control method of IPMSM considering resistive voltage drop and magnetic saturation[J].IEEE Transactions on Industrial E-lectronics,2020,67(4):2630.[2]㊀朱元,肖明康,陆科,等.电动汽车永磁同步电机转子温度估计[J].电机与控制学报,2021,25(6):72.ZHU Yuan,XIAO Mingkang,LU Ke,et al.Rotor temperature estimation for permanent magnet synchronous motors in electric ve-hicles[J].Electric Machines and Control,2021,25(6):72. [3]㊀王晓远,刘铭鑫,陈学永,等.电动汽车用ANGN带滤波补偿三阶滑模自抗扰控制[J].电机与控制学报,2021,25(11):25.WANG Xiaoyuan,LIU Mingxin,CHENG Xueyong,et al.Third-order sliding mode active disturbance rejection control of PMSM with filter compensation for electric vehicle[J].Electric Machines and Control,2021,25(11):25.[4]㊀BRIZ F,DEGNER M.Rotor position estimation[J].IEEE Indus-trial Electronics Magazine,2011,5(2):24.[5]㊀YEH H,YANG S.Phase inductance and rotor position estimationfor sensorless permanent magnet synchronous machine drives at standstill[J].IEEE Access,2021(9):32897.[6]㊀贾洪平,贺益康.基于高频注入法的永磁同步电动机转子初始位置检测研究[J].中国电机工程学报,2007,27(15):15.JIA Hongping,HE Yikang.Study on inspection of the initial rotor position of a PMSM based on high-frequency signal injection[J].Proceedings of the CSEE,2007,27(15):15.[7]㊀张树林,康劲松,母思远.基于等宽电压脉冲注入的永磁同步电机转子初始位置检测方法[J].中国电机工程学报,2020,40(19):6085.ZHANG Shulin,KANG Jinsong,MU Siyuan.Initial rotor position detection for permanent magnet synchronous motor based on identi-cal width voltage pulse injection[J].Proceedings of the CSEE, 2020,40(19):6085.[8]㊀王宾,彭皆彩,于水娟.一种电流合成的PMSM转子初始位置检测方法[J].电机与控制学报,2020,24(8):67.WANGBin,PENG Jiecai,YU Shuijuan.Method to detect the ini-tial rotor position of PMSM based on current synthesis[J].Elec-tric Machines and Control,2020,24(8):67.[9]㊀孟高军,余海涛,黄磊,等.一种基于线电感变化特征的永磁同步电机转子初始位置检测新方法[J].电工技术学报, 2015,30(20):1.MENG Gaojun,YU Haitao,HUANG Lei,et al.A novel initial rotor position estimation method for PMSM based on variation be-havior of line inductances[J].Transactions of China Electrotech-nical Society,2015,30(20):1.[10]㊀WU X,LU Z,LING Z,et al.An improved pulse voltage injec-tion based initial rotor position estimation method for PMSM[J].IEEE Access,2021(9):121906.[11]㊀鲁家栋,刘景林,卫丽超.永磁同步电机转子初始位置检测方法[J].电工技术学报,2015,30(7):105.LU Jiadong,LIU Jinglin,WEI Lichao.Estimation of the initialrotor position for permanent magnet synchronous motors[J].Transactions of China Electrotechnical Society,2015,30(7):105.[12]㊀JIN X,NI R,CHEN W,et al.High-frequency voltage-injectionmethods and observer design for initial position detection of per-manent magnet synchronous machines[J].IEEE Transactions onPower Electronics,2018,33(9):7971.[13]㊀王华斌,施金良,陈国荣,等.内嵌式永磁同步电机转子初始位置检测[J].电机与控制学报,2011,15(3):40.WANG Huabin,SHI Jinliang,CHEN Guorong,et al.Initial ro-tor position detection of IPMSM[J].Electric Machines and Con-trol,2011,15(3):40.[14]㊀刘景林,鲁家栋.基于相电流正负序分量相角差的高精度内置式永磁同步电机转子初始位置检测方法[J].电工技术学报,2016,31(23):63.LIU Jinglin,LU Jiadong.High-precision estimation method of in-itial rotor position for IPMSM based on phase difference of posi-tive and negative sequence current component[J].Transactionsof China Electrotechnical Society,2016,31(23):63. [15]㊀杨健,杨淑英,李浩源,等.基于旋转高频电压注入的永磁同步电机转子初始位置辨识方法[J].电工技术学报,2018,33(15):3547.YANG Jian,YANG Shuying,LI Haoyuan,et al.Initial rotor po-sition estimation for IPMSM based on high frequency rotating volt-age injection[J].Transactions of China Electrotechnical Society,2018,33(15):3547.[16]㊀SHUANG B,ZHU Z.A novel sensorless initial position estima-tion and startup method[J].IEEE Transactions on Industrial E-lectronics,2021,68(4):2964.[17]㊀于安博,刘利,阚志忠,等.高频脉振信号注入永磁同步电机无滤波器初始位置辨识方法[J].电工技术学报,2021,36(4):801.YU Anbo,LIU Li,KAN Zhizhong,et al.Initial position identi-051电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀fication of PMSM with filterless high frequency pulse signal injec-tion method[J].Transactions of China Electrotechnical Society,2021,36(4):801.[18]㊀FU X,XU Y,HE H,et al.Initial rotor position estimation bydetecting vibration of permanent magnet synchronous machine[J].IEEE Transactions on Industrial Electronics,2021,68(8):6595.[19]㊀ZHANG X,LI H,YANG S,et al.Improved initial rotor positionestimation for PMSM drives based on HF pulsating voltage signalinjection[J].IEEE Transactions on Industrial Electronics,2018,65(6):4702.[20]㊀李洁,周波,刘兵,等.表贴式永磁同步电机无位置传感器起动新方法[J].中国电机工程学报,2016,36(9):2513.LI Jie,ZHOU Bo,LIU Bing,et al.A novel starting strategy ofsensorless control for surface mounted permanent magnet synchro-nous machines[J].Proceedings of the CSEE,2016,36(9):2513.[21]㊀TANG Q,SHEN A,LUO X,et al.PMSM sensorless control byinjecting hf pulsating carrier signal into ABC frame[J].IEEETransactions on Power Electronics,2017,32(5):3767. [22]㊀吕德刚,姜国威,纪堂龙.永磁同步电机低速域改进高频脉振注入控制[J].哈尔滨理工大学学报,2022,27(6):32.LÜDegang,JIANG Guowei,JI Tanglong.Improved high fre-quency pulse injection control inlow speed domain of permanentmagnet synchronous motor[J].Journal of Harbin University ofScience and Technology,2022,27(6):32.[23]㊀WU T,LUO D,HUANG S,et al.A fast estimation of initial ro-tor position for low-speed free-running IPMSM[J].IEEE Trans-actions on Power Electronics,2020,35(7):7664. [24]㊀XUAN W,YAO F,XIAO L,et al.Initial rotor position detec-tion for sensorless interior PMSM with square-wave voltage injec-tion[J].IEEE Transactions on Magnetics,2017,53(11):1.(编辑:刘琳琳)151第2期姚培煜等:基于非线性建模与拟合的永磁同步电机转子初始位置精确估计方法。
永磁同步电机伺服系统的滑膜变结构控制研究
➢法国学者Laghrouche S,
Plestan F, Glumineau A, et al.[7]等设计了一个二阶 SMC调节器控制 PMSM调 速系统。
➢I.俄Ut罗ki斯n.学[9]者提V出ad了im滑模变结
构控制器在交流传动系统中 的设计方法和原则,详细分 析降阶、参数扰动的不敏感 性以及功率变换器实现方法。
建模并进行仿真
3
分析,撰写硕士
论文初稿
4
2014.1-2014.3 总结并完成硕士
论文的撰写
进度安排及参考文献
参考文献
[1]唐任远等.现代永磁电机[M].机械工业出版社.1997. [2]郭庆鼎等.交流伺服系统[M].机械工业出版社.1994. [3]王丰尧.滑模变结构控制[M].机械工业出版社.1998. [4]陈志梅.滑膜变结构控制理论及应用[M].电子工业出版社.2012. [5]高为炳.变结构控制理论基础[M].中国科学技术出版社.1990. [6]In Cheol Baik, Kyeong Hwa Kim. Robust nonlinear speed control of PM
课题的国内、外研究现状
拟选课题在该领域目前存在的问题
(1)永磁同步电机滑模设计主要来解决运动点到达滑模 状态后的收敛速度,即动态快速性问题。 (2)抖振问题是阻碍滑模变结构控制在永磁同步电机交
访谈结果流与伺析服系统中应用的突出障碍,解决抖振问题实际上是
解决系统稳态的稳定性和精度问题。
课题拟研究内容
[11]童克文,张兴,等.基于新型趋近律的永磁同步电动机滑模变结构控制[J]. 中国电机工程学报,2008,28(21):102-106.
[12]葛宝明,郑琼林,蒋静坪,等.基于离散时间趋近率控制与内模控制的永 磁同步电动机传动系统[J].中国电机工程学报,2004,24(11):106-111.
带有负载转矩观测器的PMSM自适应反步控制
ZHANG Hu , ZHU Xiaohong , Sun Mingshan
(1.Suzhou Higher Vocational School,Suzhou 215009,China;2.College of Electr ical Engineer ing and Control Science,Nanjing TECH University,Nanjing 21 1816,China;3.Suzhou Institute of
迫札 再控 制 应闭2016,43(4)
控制与应用技术 i E№ A
带有 负载转 矩观测器 的 PMSM 自适应 反步控 制
张 虎 , 朱 晓虹 , 孙 明 山。 (1.苏州高等职业技术学校 ,江苏 苏州 215009; 2.南京工 业 大学 电 气工程 与控 制科 学学 院 ,江 苏 南京 21 1816; 3.中科 院 苏州 生物 医学工程技 术研 究所 ,江 苏 苏州 215163)
Biomedical Engineer ing and Technology Chinese Academy of Sciences,Suzhou 215 163,China)
Abstract:The backstepping control m ethod was proposed due to the nonlinear characteristics of perm anent m agnet synchronous m otor.The design of the method was sim ple,and it can decouple the control system completely SO the method can achieve better control than traditional PI control m ethod.Because of the temperature changes, resistance value will chan ge also.W e present a combination method of Adaptive Control and t he backstepping control met h od to overcome th e instability of the system influenced by param eters.In addition,we design a load torque sliding mode observer to reduce the influences of load change. The observed load torque was used for feed—f orward
基于特征模型的永磁同步直线电机自适应控制
第28卷㊀第3期2024年3月㊀电㊀机㊀与㊀控㊀制㊀学㊀报Electri c ㊀Machines ㊀and ㊀Control㊀Vol.28No.3Mar.2024㊀㊀㊀㊀㊀㊀基于特征模型的永磁同步直线电机自适应控制曹阳,㊀郭健(南京理工大学自动化学院,江苏南京210094)摘㊀要:为了解决永磁同步直线电机系统的参数不确定性㊁建模不确定性及饱和非线性等问题,提出一种基于特征模型的自适应控制器㊂依据特征模型理论描述永磁同步直线电机系统,采用自适应和鲁棒控制方法设计控制器㊂建立永磁同步直线电机的特征模型,并给出具体建立步骤,使得控制器设计变得简单,易于工程实现㊂通过设计参数自适应律对系统未知特征参数进行估计,可实现对系统模型的精确补偿,同时在控制器中添加带有误差积分的鲁棒控制项,提高系统对不确定参数及未知干扰的鲁棒性㊂此外,由于饱和特性的存在,导致控制器产生windup 问题,给系统的控制性能和稳定性造成不利影响㊂因此,该控制器中还带有抗饱和控制项,能够提升系统的抗饱和能力㊂最后,通过对比实验验证了所提控制器的有效性㊂关键词:永磁同步直线电机;参数不确定性;建模不确定性;饱和非线性;特征模型;自适应控制;抗饱和DOI :10.15938/j.emc.2024.03.013中图分类号:TM351文献标志码:A文章编号:1007-449X(2024)03-0131-10㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀收稿日期:2022-07-04基金项目:国家自然科学基金(61673219)作者简介:曹㊀阳(1993 ),男,博士研究生,研究方向为电机系统分析与控制;郭㊀健(1974 ),男,博士,教授,博士生导师,研究方向为智能系统与智能控制㊁机器人系统㊁高精度电机控制等㊂通信作者:郭㊀健Adaptive control of permanent magnet synchronous linear motorbased on characteristic modelCAO Yang,㊀GUO Jian(School of Automation,Nanjing University of Science and Technology,Nanjing 210094,China)Abstract :To address the problems of parameter uncertainty,modeling uncertainty and saturation nonlin-earity in the permanent magnet synchronous linear motor system,an adaptive controller based on charac-teristic model was proposed.A characteristic model was used to describe the permanent magnet synchro-nous linear motor system,and the controller was designed using adaptive and robust control methods.The characteristic model was established based on the system dynamics and parameters,and the specific steps were presented.This simplifies the controller design and facilitates the engineering implementation.An online parameter adaptation law was employed to estimate the unknown characteristic parameters of the system and achieve accurate compensation for the system model.Furthermore,an integral-type robust control term was incorporated into the controller,which improves the robustness of the system against un-certain parameters and unknown disturbances.In addition,the saturation nonlinearity leads to the windup problem in the controller,which has adverse effects on the control performance and stability of the sys-tem.Therefore,an anti-windup control scheme was devised for the controller,which can enhance the an-ti-saturation ability of the system.Finally,comparative experiments with other control methods were con-ducted to verify effectiveness of the proposed controller.Keywords:permanent magnet synchronous linear motor;friction nonlinearity;saturation nonlinearity;ar-mature mass variation;characteristic model;adaptive control;anti-windup0㊀引㊀言相比于旋转同步电机,永磁同步直线电机(per-manent magnet synchronous linear motor,PMSLM)具有更高的推力密度和更快的动态响应,特别适用于对速度和精度要求较高的场合,已被广泛应用在高精密加工㊁轨道交通传输等现代工业领域[1-2]㊂但是由于采用直接驱动方式,PMSLM控制系统对参数摄动及扰动等因素变得更加敏感[3],这会严重影响系统的控制性能㊂因此,保证PMSLM系统的高精度跟踪性能与抗扰动能力十分重要,对提高机床加工精度㊁提升交通传输效率具有重要的意义㊂针对PMSLM系统的高精度跟踪问题,国内外已有众多学者对其进行了研究㊂文献[4]设计了一种带模型参考自适应观测器的预测电流控制策略,经过实验验证该控制策略可以实现对速度进行在线准确辨识,进而提高电流的跟踪性能㊂文献[5]利用扩张状态观测器和非线性状态误差反馈对PMSLM的自抗扰控制器进行优化,提高了系统的动态响应性能和抗干扰能力㊂文献[6]提出一种基于周期性扰动学习的自适应滑模控制方法,采用滑模控制确保PMSLM系统对不确定性因素具有较强的鲁棒性㊂文献[7]在系统模型反馈线性化的基础上,将Hɕ鲁棒控制方法与D-K迭代法相结合,提高了系统对不确定性因素影响的抑制能力㊂姚斌等[8]提出一种自适应鲁棒控制方法,所开发的控制器成功应用在多种控制系统中[9-11]㊂为了解决非光滑饱和非线性的影响,文献[12]构造了一种新的近似饱和模型,该模型能够以任意规定的精度平滑地逼近实际饱和㊂此外,通过添加积分器技术,使得控制器可以消除与表面误差和边界层误差有关的耦合项㊂但是该方法在控制器的设计中需要对虚拟控制量重复微分,如果系统模型阶数高,会增加设计的复杂性㊂文献[13]提出一种考虑LuGre 摩擦的自适应鲁棒控制方法,针对陀螺框架伺服系统未知惯量和阻尼系数㊁LuGre摩擦参数不确定性及未知外部干扰上界,设计参数更新律对其进行估计,该控制律提高了系统的跟踪精度并通过仿真结果验证了所提方法的有效性㊂但该方法需要被控对象的精确数学模型,另外估计的未知参数过多,多个自适应参数需要反复调试,增加了实际应用时的难度㊂自适应鲁棒控制可以估计系统未知参数,但如果系统模型复杂㊁未知参数多㊁某些状态不可测时,控制器的设计将面临巨大挑战㊂针对这些问题,吴宏鑫院士等[14-15]提出特征建模的思想,特征模型一般用一阶或二阶差分方程/微分方程来描述,有关信息都压缩到几个特征参数中,并不丢失原有的信息㊂特征模型建立的形式比原对象动力学方程简单,为实际复杂系统的建模问题提供了一条途径㊂文献[16]基于永磁同步电机的特征模型,设计一个以非线性黄金分割自适应控制为主的控制方案㊂通过安排过渡过程和特征模型参数的在线辨识,该控制方案实现了控制器参数的在线自适应调节㊂文献[17]将特征建模方法推广到具有惯性变化的齿轮传动伺服系统中,设计了一个自适应二阶离散终端滑模控制器,并实现了有限时间有界性㊂然而上述基于特征模型所设计的控制器没有进行抗饱和(anti-windup)研究㊂windup现象是指由于被控对象的输入限制,使得被控对象的实际输入与控制器的输出不等,引起系统闭环响应变差(如超调变大,调节时间变长,甚至使系统失去稳定)的现象㊂实际的PMSLM是个物理限制系统,转速控制器的输出必须限定在一定的范围内,使得实际电机的控制输入量不能大于一个预先设定值㊂当控制器输出受到饱和限制时,特别是含有积分项的控制信号仍然增加时,就会出现windup现象,使实际闭环系统的性能下降,因此对PMSLM系统设计抗饱和控制是有必要的[18-19]㊂基于上述分析,针对PMSLM系统存在的参数不确定性㊁建模不确定性及饱和非线性等问题,提出一种基于特征模型的抗饱和自适应鲁棒控制器(an-ti-windup adaptive robust control based on characteris-tic model,AARC)㊂利用特征模型简化PMSLM系统的描述,并对其进行验证㊂然后,设计一种基于参数投影的自适应律,实现对系统模型的在线补偿㊂同时,将系统的不确定参数和未知干扰视为集总的干231电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀扰项,引入误差积分的鲁棒控制项进行抑制㊂此外,为了解决积分环节可能引起的windup 现象,加入抗饱和控制项,提高系统的抗饱和能力㊂最后,基于Lyapunov 函数证明闭环系统的稳定性,并通过实验验证所提控制器的有效性和鲁棒性㊂1㊀PMSLM 的特征建模与验证1.1㊀PMSLM 模型PMSLM 的运动方程为m d y d t =3π2τn p i q [ψf+(L d -L q )i d ]-F fric (y )㊂(1)式中:m 为等效质量;ψf 为磁链;y 为动子速度;i d ㊁i q 分别为d㊁q 轴电流;τ为极距;n p 为极对数;L d ㊁L q 分别为d㊁q 轴电感;F fric (y )为摩擦力㊂由式(1)可得y ㊃㊃=1.5πn p mτ[ψf i ㊃q +(L d -L q )(i ㊃d i q +i ㊃q i d )]- F fric y㊃m y㊂(2)设PMSLM 的采样周期为T ,将式(2)离散化可得㊀y (k +1)-2y (k )+y (k -1)T 2=[1.5πmTτn p ψf +1.5n p (L d -L q )i d (k )mTτ]i q (k )-[1.5πmTτn p ψf +1.5n p (L d -L q )i d (k )mTτ]i q (k -1)+1.5πn p (L d -L q )i q (k )mTτ[i d (k )-i d (k -1)]-1mT F firc (y (k )-y (k -1))y ㊂(3)在式(3)两边同乘T 2,可以重新写为y (k +1)=[1.5πmτn p ψfT +1.5n p (L d -L q )i d (k )Tmτ]i q (k )+[2-1m F firc T v ]y (k )+[1m F firc T v-1]y (k -1)+[1.5n p (L d -L q )i d (k )T mτ-1.5πmτn p ψfT ]i q (k -1)+1.5πn p (L d -L q )i q (k )Tmτˑ[i d (k )-i d (k -1)]=β1(k )i q (k )+α1(k )y (k )+α2(k )y (k -1)+Δ(k )㊂(4)式中:y (k )为系统输出;i q (k )为系统输入;α1㊁α2㊁β1为系统的特征参数,定义为:α1(k )=[2-1m F firc Tv];α2(k )=[1m F firc Tv -1];β1(k )=[1.5πmτn p ψf T +1.5n p (L d -L q )i d (k )T mτ]㊂üþýïïïïïïï(5)Δ(k )表示集总未知非线性函数,包括建模误差和未知扰动,定义为Δ(k )=[1.5n p (L d -L q )i d (k )Tmτ-1.5πmτn p ψfT ]i q (k -1)+1.5πn p (L d -L q )i q (k )Tmτˑ[i d (k )-i d (k -1)]㊂(6)通过式(4)可以看出,特征模型是将模型结构的模型不确定性和参数摄动等不确定信息压缩成几个未知的特征参数,使其与实际模型等价而不是近似㊂使用特征建模不仅能简化控制器设计,而且更利于工程应用㊂1.2㊀特征模型验证特征模型验证过程如图1所示㊂首先,分别给予PMSLM 系统和特征模型相同的输入信号u ㊂然后,采样PMSLM 的输入输出信号,采用传统投影梯算法[16]在线辨识特征参数,并计算特征模型输出㊂最后,通过比较特征模型输出y ^与PMSLM 系统输出y ,得到误差e 0㊂将输入设为1sin(2.09t )A 的正弦信号,并且设PMSLM 的采样频率为80μs㊂特征模型验证结果如图2所示㊂实验结果表明,在相同的控制输入作用下,特性模型输出与实际系统输出的误差很小,说明特征模型可以很好地描述PMSLM 系统的输入输出特征,可以利用该特征模型来设计控制器㊂331第3期曹㊀阳等:基于特征模型的永磁同步直线电机自适应控制图1㊀特征模型验证Fig.1㊀Verification block diagram of characteristicmodel图2㊀特征模型验证结果Fig.2㊀Verification results of characteristic model2㊀非线性自适应控制器设计2.1㊀自适应控制设计针对PMSLM 系统中存在的参数不确定㊁饱和非线性以及外界干扰,设计基于特征模型的自适应鲁棒控制律,对系统的不确定性和干扰进行估计和补偿,实现PMSLM 的速度跟踪控制㊂设计的自适应控制结构如图3所示,控制器包括模型补偿项u a ㊁线性反馈项u s1㊁积分鲁棒控制律u s2和抗饱和控制律k cw η,i qmax =0.03㊁i qmin =-0.03为饱和限制上下界㊂图3㊀自适应抗饱和控制结构框图Fig.3㊀Structure diagram of adaptive anti-windupcontroller将特征模型写成如下二阶时变辨识模型:y (k +1)=φ(k )T θ(k )㊂(7)式中:φ(k )=[y (k )y (k -1)u (k )]T ;θ(k )=[α1(k )α2(k )β1(k )]T ㊂在下面的部分中,㊃j 表示向量㊃的第j 个分量,并且针对2个向量的运算 < 是根据向量的相应元素来执行的㊂用θ^表示θ的估计值,θ~表示估计误差(θ~=θ^-θ)㊂结合式(7),一种不连续投影可以定义为proj θ^j {㊃j }=0,if θ^j =θj max and㊃j >0;0,if θ^j =θj min and㊃j <0;㊃j ,otherwise㊂ìîíïïïï(8)式中:j =1,2,3;proj θ^j{㊃j }可以保证估计参数在有界凸闭集D s 内㊂为保证参数估计值的有界性,设计未知参数估计自适应律为:θn (k )=θ^(k -1)+Γτλ+φT(k -1)φ(k -1);θ^(k )=proj θ^(θn(k ))㊂}(9)式中:Γ>0,λ>0为待设计的可调参数;τ为待合成的自适应函数;θ^(k )为系统参数θ(k )的估计值,利用基于不连续投影的参数自适应律可以估计出未知的特征参数α1㊁α2㊁β1㊂特征模型式(4)可被重写为y (k +1)=[α^1(k )-α~1(k )]y (k )+[α^2(k )-α~2(k )]y (k -1)+[β^1(k )-β~1(k )]u (k )+β1η(k )+Δ(k )㊂(10)式中α~1(k )=α^1(k )-α1(k ),α~2(k )=α^2(k )-α2(k ),β~1(k )=β^1(k )-β1(k )为辨识误差㊂所以式(10)可以改写为431电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀y(k+1)=α^1(k)y(k)+α^2(k)y(k-1)+β^1(k)u(k)+β1η+Δ(k)-θ~(k)φ(k)㊂(11)其中θ~(k)φ(k)=α~1(k)y(k)+α~2(k)y(k)+β~1(k)u(k)表示模型估计误差㊂假设1:从工程实践中可知,对于稳定对象,参数不确定性和不确定非线性的程度已知,即θɪΩθ {θ:θminɤθɤθmax};ΔɪΩd {Δ:|Δ(k)-Δ(k-1)|ɤδd(k)}㊂}(12)式中:θmin=[θ1min, ,θ3min]T;θmax=[θ1max, ,θ3max]T;δd是已知的㊂控制目标是设计自适应控制器使得系统的输出y(k)跟踪期望输出y d(k),定义跟踪误差函数为e(k)=y(k)-y d(k)㊂(13)定义s(k)为s(k)=e(k)-k1e(k-1)㊂(14)其中0<k1<1为待设计的可调参数㊂所以有s(k+1)=e(k+1)-k1e(k)㊂(15)自适应抗饱和控制律可以设计为:u(k)=1β^1(k)[u a(k)+u s1(k)+u s2(k)];u a(k)=-α^1(k)y(k)-α^2(k)y(k-1)+ y d(k+1)+k1e(k)-k cwη;u s1(k)=k s s(k);u s2(k)=-E1(k)㊂üþýïïïïïïïï(16)式中:k cwȡ β1 max为抗饱和反馈增益;|k s|<1是待设计的可调参数;E1(k)表达式为E1(k)=E1(k-1)+k s k2s(k-1)+βsat(s(k-1))㊂(17)式中:k2>0为可调系数;sat(㊃)为饱和函数㊂设计参数自适应律τ=s(k)φ(k-1),将式(9)改写为:θn(k)=θ^(k-1)+Γs(k)φ(k-1)λ+φT(k-1)φ(k-1);θ^(k)=projθ^(θn(k))㊂üþýïïï(18) 2.2㊀稳定性分析定理1:对于特征模型式(10)所描述的PMSLM,所有信号都是有界的㊂采用自适应控制律式(16)和参数更新规律式(18),能使闭环系统的跟踪误差渐近收敛至0㊂证明:将式(16)代入式(10)中,并结合式(18)可得s(k+1)=[y(k+1)-y d(k+1)]-k1e(k)=α^1(k)y(k)+α^2(k)y(k-1)+β^1(k)u(k)-α~1(k)y(k)-α~2(k)y(k-1)-β~1(k)u(k)+Δ(k)=-θ~T(k)φ(k)+β1η(k)-k cwη(k)+k s s(k)-E1(k)+Δ(k)㊂(19)取k cwȡ β1 max,然后对式(19)进行差分可得s(k+1)-s(k)=-(θ~T(k)φ(k)-θ~T(k-1)φ(k-1))+k s(s(k)-s(k-1))-(E1(k)-E1(k-1))+Δ(k)-Δ(k-1)㊂(20)考虑到采样周期很小,通过线性外推法预测可知s(k+1)=2s(k)-s(k-1)㊂(21)构建Lyapunov函数为V(k)=s(k)λ+φT(k-1)φ(k-1)+θ~(k) 2Γ㊂(22)首先考虑式(22)的第2项,根据投影参数自适应律式(18)可得θ~(k) 2ɤ θn(k)-θ(k) 2= θ~(k-1) 2+2Γs(k)φT(k-1)θ~(k-1)λ+ φ(k-1)Tφ(k-1) +(Γs(k))2 φ(k-1) 2(λ+ φ(k-1) 2)2ɤ2Γs(k)φT(k-1)θ~(k-1)λ+ φ(k-1) 2+Γ2s2(k)λ+ φ(k-1) 2+ θ~(k-1) 2㊂(23)将式(16)㊁式(20)和式(21)代入式(23)可得 θ~(k) 2- θ~(k-1) 2ɤ2Γs(k)[-(s(k)-s(k-1))+k s(s(k-1)-s(k-2))]λ+ φ(k-1) 2+ 2Γs(k)[-θ~T(k-2)φ(k-2)+k s k2s(k-1)-βsign(s(k-1))]λ+ φ(k-1) 2+531第3期曹㊀阳等:基于特征模型的永磁同步直线电机自适应控制2Γs (k )[(Δ(k -1)-Δ(k -2)]λ+ φ(k -1) 2+Γ2s 2(k )λ+ φ(k -1) 2㊂(24)选取βȡ| θM φmax +δd |,进一步可得 θ~(k ) 2- θ~(k -1) 2ɤ2Γs (k )(k s -1)(s (k )-s (k -1))+2Γk s k 2s (k )s (k -1)λ+ φ(k -1) 2+Γ2s 2(k )λ+ φ(k -1) 2㊂(25)引理1[20]:(Young 不等式)假设a ㊁b 为非负实数,P >1,1p +1q =1,那么ab ɤa p p +b pq ,当且仅当a p=b q时,等号成立㊂根据引理1可得:2s (k )s (k -1)ɤ s (k ) 2+ s (k -1) 2; θ~(k ) 2- θ~(k -1) 2ɤ-Γ(3-3k s -k s k 2)s 2(k )λ+ φ(k -1) 2+Γ(k s +k s k 2-1)s 2(k -1)λ+ φT (k -1) 2㊂üþýïïïïïï(26)对Lyapunov 函数式(22)进行差分,并联立式(26)可得ΔV (k )=V (k )-V (k -1)ɤs 2(k )λ+ φT (k -1) 2-s 2(k -1)λ+ φT (k -2) 2+-(3-3k s -k s k 2)s 2(k )λ+ φ(k -1) 2+(k s +k s k 2-1)s 2(k -1)λ+ φT (k -1) 2+Γs 2(k )λ+ φT (k -1) 2ɤ-(2-3k s -k s k 2-Γ)s 2(k )λ+ φT (k -1) 2+(k s +k s k 2-1)s 2(k -1)λ+ φT (k -1) 2-s 2(k -1)λ+ φT (k -2) 2ɤ-As 2(k )-Bs 2(k -1)㊂(27)式中:A =2-3k s -k s k 2-Γλ+ φT (k -1) 2;B =1λ+ φT (k -2) 2-1-k s -k s k 2λ+ φT (k -1) 2㊂通过选取合适的参数k s ㊁k 2㊁Γ㊁λ使得A >0,B >0㊂根据式(27),对Δ(k )从1到k 求和可得ðki =1[As 2(k )+Bs 2(k -1)]ɤV (1)-V (k )ɤV (1)㊂(28)当k ңɕ时,As 2(k )+Bs 2(k -1)ȡ0,由于φ(k ) 有界,可知lim k ңɕ|s (k )|=0㊂(29)根据式(29)可知,∃N ,当k >N 时,有|s (k )|ɤ0㊂(30)由式(15)可得|e (k )|ɤ|k 1||e (k -1)|+|s (k )|ɤ|k 1|k -N|e (N )|+|k 1|k -N -1|s (N +1)|+ +s (k )ɤ|k 1|k -N|e (N )|+0㊂(31)因为|k s |<1,所以有lim k ңɕsup |e (k )|=0㊂(32)3㊀实验结果比较为了说明上述方法的可行性和有效性,在实验室建立一个验证平台如图4所示,PMSLM 的基本参数列于表1㊂该平台由MOSFET 三相逆变桥㊁磁栅尺㊁相电流采样电路㊁TMS320F28062(DSP)及外围电路㊁IR2181S 驱动电路㊁系统电源电路组成㊂此外,为了模拟不同的工作条件,对直线电机的动子进行了调整㊂通过直接在动子上安装标准化铁块,准确地改变其质量m ,以模拟不同的惯性效应㊂图4㊀PMSLM 实验平台Fig.4㊀PMSLM experimental platform631电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀表1㊀PMSLM 的基本参数Table 1㊀Parameters of PMSLM㊀㊀参数数值极对数n p7极距τ/mm(180ʎ)12d 轴电感L d /mH 8q 轴电感L q /mH 8永磁体磁链ψf /Wb0.61PMSLM 矢量控制系统框架如图5所示㊂它由PMSLM㊁空间矢量脉宽调制(space vector pulse widthmodulation,SVPWM)模块㊁Park 和Clark 坐标变换㊁电压源逆变器㊁电流调节器和速度控制器组成㊂本文设计一种速度控制器,电流控制器采用PI 控制㊂图5㊀矢量控制总体结构框图Fig.5㊀Overall structure diagram of vector control为了验证所提控制器的可行性和有效性,本文对以下3种控制器进行比较㊂1)AARC㊂本文设计的抗饱和自适应鲁棒控制器参数设置如下:k 1=0.15,k 2=0.0006,k s =0.1,β=0.04,k cw =0.1,Γ=0.05,λ=0.995,θ^(0)=[1.9,-0.9,0.00001]T ㊂2)抗饱和自适应控制器(anti-windup adaptivecontrol based on characteristic model,AAC)㊂未添加鲁棒项u s2的抗饱和自适应控制器,其他参数与AARC 一致㊂3)抗饱和PID 控制器(anti-windup proportional-integral-differential,APID)㊂控制器的增益设置为k p =150,k i =1,k d =0,k cw =0.1㊂此外,将使用跟踪误差的最大值㊁平均值和标准差来衡量每个控制算法的质量,定义如下:1)最大跟踪误差的绝对值为M e =max i =1, ,N{|e (i )|}㊂(33)2)平均跟踪误差定义为μ=1N ðNi =1|e (i )|㊂(34)3)跟踪误差的标准差为δ=1N ðNi =1[|e (i )|-μ]2㊂(35)其中N 是所记录的数字信号的个数㊂首先将给定速度设置为y d =0.56sin(3.14t)m/s㊂系统跟踪结果如图6所示,性能指标如表2所示㊂从这些实验结果可以看出,所提出的AARC 控制器在瞬态和最终跟踪误差方面优于其他两种控制器,因为AARC 采用了基于参数自适应的补偿和鲁棒控制项,可以同时处理参数和未建模不确定性㊂虽然AAC 中也包含参数自适应,但对于建模的不确定性和未知扰动的抑制效果不佳㊂通过表2可以看出,AARC 添加鲁棒项后各种误差指标会比AAC 小,验证了鲁棒控制项u s2的有效性㊂在3种控制器中,线性抗饱和PID 的误差指标最差,达到了AARC 的2倍以上,这说明基于非线性模型的控制器设计方法具有更大的优势㊂图6㊀无铁块情况下PMSLM 的跟踪结果Fig.6㊀Tracking results of PMSLM without iron表2㊀最后两个周期的性能指标Table 2㊀Performance indexes during the last two cycles控制方法M e /(m /s)μ/(m /s)δ/(m /s)APID 0.055420.013360.00971AAC0.026890.008100.00572AARC 0.025220.006000.00490731第3期曹㊀阳等:基于特征模型的永磁同步直线电机自适应控制为了进一步验证控制器对参数变化的自适应能力,设定了不同的动子质量来进行实验㊂给PMSLM 的动子上添加1.33kg 的铁块㊂系统跟踪结果如图7所示,表3列出了最后两个周期的性能指标㊂从图7可以看出,使用AARC 控制方法的控制系统,在面对动子质量变化时,其反应速度快,并且波动较小㊂从表3可知,APID 的最大跟踪误差没有增大,意味着APID 中存在大的积分增益对该扰动也有一定的抑制效果㊂但与上一个实验情况相比,APID 的μ和δ指标增大明显,仍然比其他2个控制器差㊂适当的参数自适应在一定程度上也可以削弱动子质量变化给系统带来的参数不确定性影响,就像AAC 那样㊂AARC 的各项误差指标是3个控制器中最好的,再次证明了该控制器的有效性㊂图7㊀铁块质量为1.33kg 时PMSLM 的跟踪结果Fig.7㊀Tracking results of PMSLM when iron massis 1.33kg表3㊀最后两个周期的性能指标Table 3㊀Performance indexes during the last two cycles控制方法M e /(m /s)μ/(m /s)δ/(m /s)APID 0.043890.015370.01061AAC0.029620.008440.00605AARC 0.025320.005980.00496最后将动子上的铁块增加到2.64kg,此时PMSLM 受到的摩擦非线性和扰动进一步增大,3个控制器的跟踪性能都有所变差㊂实验结果如图8所示,误差指标见表4㊂在这个测试用例中,APID 中的跟踪误差抖动变大,而AARC 的跟踪误差则相当平滑㊂APID 控制器表现出最差的跟踪性能,最大跟踪误差为0.094,表明APID 在该跟踪任务中已经达到了其局限性㊂另外,即使在增大动子质量情况下,所提出的AARC 控制器仍然可以对模型进行补偿并衰减未建模的扰动,从而在所有比较的控制器中达到最好的跟踪性能㊂图8㊀铁块质量增加到2.64kg 情况下PMSLM 的跟踪结果Fig.8㊀Tracking results of PMSLM when the mass ofiron is increased to 2.64kg 表4㊀最后两个周期的性能指标Table 4㊀Performance indexes during the last two cycles控制方法M e /(m /s)μ/(m /s)δ/(m /s)APID 0.093700.027090.01934AAC0.034620.008410.00643AARC 0.028870.005860.005054㊀结㊀论本文针对PMSLM 系统提出一种基于特征模型的自适应控制方法,该方法能够有效地解决PMSLM 系统的参数不确定性㊁建模误差和外部干扰等问题㊂首先利用二阶变差分方程对PMSLM 系统进行简化831电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀建模,然后设计了一种基于特征模型的自适应控制器,仅利用系统的输入和输出信号,实现了对PMSLM系统的精确速度跟踪控制㊂为了提高系统的鲁棒性和抗饱和能力,还引入了鲁棒补偿项和抗饱和控制项,并严格证明了闭环系统的稳定性㊂最后,通过实验结果验证了所提控制方法的有效性㊂本文控制器的参数是固定的,需要通过反复调试来确认㊂当实验条件和环境发生改变时,可能导致参数不一定是最优的㊂因此,在未来工作中将考虑进一步研究控制器参数的自动调整技术[21],采用自学习的方法来替代控制器中参数的人工调整部分㊂参考文献:[1]㊀龚夕霞,李焱鑫,卢琴芬.模块化永磁直线同步电机考虑制造公差的推力鲁棒性优化[J].电工技术学报,2024,39(2):465.GONG Xixia,LI Yanxin,LU Qinfen.Thrust robustness optimiza-tion of modular permanent magnet linear synchronous motor ac-counting for manufacture tolerance[J].Transactions of China Electrotechnical Society,2024,39(2):465.[2]㊀张春雷,张辉,叶佩青.高霍尔位置检测精度的圆筒型永磁同步直线电机设计[J].电工技术学报,2022,37(10):2481.ZHANG Chunlei,ZHANG Hui,YE Peiqing.Design of tubular permanent magnet synchronous linear motor by reliability-based ro-bust design optimization[J].Transactions of China Electrotechni-cal Society,2022,37(10):2481.[3]㊀缪仲翠,苏乙,张磊,等.梯形Halbach交替极无铁心永磁同步直线电机特性分析与优化设计[J].电机与控制学报, 2024,28(1):164.MIAO Zhongcui,SU Yi,ZHANG Lei,et al.Characteristic analy-sis and optimization design of air-core permanent magnet linear synchronous motor with trapezoidal Halbach array consequent-pole [J].Electric Machines and Control,2024,28(1):164. [4]㊀李争,安金峰,肖宇,等.基于自适应观测器的永磁同步直线电机模型预测控制系统设计[J].电工技术学报,2021,36(6):1190.LI Zheng,AN Jinfeng,XIAO Yu,et al.Design of model predic-tive control system for permanent magnet synchronous linear motor based on adaptive observer[J].Transactions of China Electrotech-nical Society,2021,36(6):1190.[5]㊀李争,张梓豪,王康涛,等.基于无模型的PMLSM改进自适应滑模自抗扰控制[J].电机与控制学报,2024,28(1):142.LI Zheng,ZHANG Zihao,WANG Kangtao,et al.Improved adap-tive sliding mode active disturbance rejection control for PMLSM based on model-free theory[J].Electric Machines and Control, 2024,28(1):142.[6]㊀张康,王丽梅.基于周期性扰动学习的永磁直线电机自适应滑模位置控制[J].电机与控制学报,2021,25(8):132.ZHANG Kang,WANG Limei.Adaptive sliding mode position con-trol for permanent magnet linear motor based on periodic disturb-ance learning[J].Electric Machines and Control,2021,25(8): 132.[7]㊀孙宜标,毛爽,夏加宽.直线电机悬浮平台的μ-Hɕ鲁棒控制[J].沈阳工业大学学报,2014,36(1):7.SUN Yibiao,MAO Shuang,XIA Jiakuan.μ-Hɕrobust control for linear motor levitation platform[J].Journal of Shenyang Uni-versity of Technology,2014,36(1):7.[8]㊀YAO B,BU F,REEDY J,et al.Adaptive robust control of sin-glerod hydraulic actuators:theory and experiments[J].IEEE/ ASME Transactions on Mechatronics,2000,5(1):79. [9]㊀CHEN S,CHEN Z,YAO B,et al.Adaptive robust cascade forcecontrol of1-DOF hydraulic exoskeleton for human performance augmentation[J].IEEE/ASME Transactions on Mechatronics, 2017,22(2):589.[10]㊀HELIAN B,CHEN Z,YAO B.Adaptive robust motion control ofa pump direct drive electro-hydraulic system with meter-out pres-sure regulation[J].IFAC-Papers OnLine,2020,53(2):9005.[11]㊀HAN J,YANG S,XIA L,et al.Deterministic adaptive robustcontrol with a novel optimal gain design approach for a fuzzy2doflower limb exoskeleton robot system[J].IEEE Transactions onFuzzy Systems,2020,29(8):2373.[12]㊀LIU Y.Saturated robust adaptive control for uncertain nonlinearsystems using a new approximate model[J].IET Control Theoryand Applications,2017,11(6):870.[13]㊀王璐,郭毓,钟晨星,等.控制力矩陀螺框架伺服系统期望补偿自适应鲁棒控制[J].控制理论与应用,2017,34(9):1143.WANG Lu,GUO Yu,ZHONG Chenxing,et al.Desired com-pensation adaptive robust control for control moment gyroscopesgimbal servo system[J].Control Theory and Applications,2017,34(9):1143.[14]㊀CHANG Y,JIANG T,PU Z.Adaptive control of hypersonic ve-hicles based on characteristic models with fuzzy neural networkestimators[J].Aerospace Science and Technology,2017,68:475.[15]㊀常亚菲.一类不确定非线性系统基于特征模型的复合自适应控制[J].控制理论与应用,2019,36(7):1137.CHANG Yafei.Characteristic model-based composite adaptivecontrol for a class of uncertain nonlinear systems[J].ControlTheory and Applications,2019,36(7):1137. [16]㊀王永,窦晓华,方浩,等.永磁同步电机非线性黄金分割自适应转速控制[J].电机与控制学报,2017,21(10):23.WANG Yong,DOU Xiaohua,FANG Hao,et al.Nonlineargolden-section adaptive speed control of permanent magnet syn-chronous motor[J].Electric Machines and Control,2017,21(10):23.931第3期曹㊀阳等:基于特征模型的永磁同步直线电机自适应控制[17]㊀WANG X,WU Y,ZHANG E,et al.Adaptive terminal sliding-mode controller based on characteristic model for gear transmis-sion servo systems[J].Transactions of the Institute of Measure-ment and Control,2019,41(1):219.[18]㊀QI L,BAO S,SHI H.Permanent-magnet synchronous motor ve-locity control based on second-order integral sliding mode controlalgorithm[J].Transactions of the Institute of Measurement andControl,2015,37(7):875.[19]㊀张兴华,姚丹.感应电机直接转矩控制系统的 抗饱和 控制器设计[J].电工技术学报,2014,29(5):181.ZHANG Xinghua,YAO Dan.Anti-windup speed controller de-sign for direct torque controlled induction motor drives[J].Trans-actions of China Electrotechnical Society,2014,29(5):181.[20]㊀IGHACHANE M A,AKKOUCHI M.Further refinements ofYoung's type inequality for positive linear maps[J].Revista de laReal Academia de Ciencias Exactas,Físicas y Naturales.SerieA.Matemáticas,2021,115(2):10.[21]㊀BENOSMAN M.Learning-based adaptive control for nonlinearsystems[C]//European Control Conference(ECC),June24-27,2014,Strasbourg,France.2014:920-925.(编辑:邱赫男)041电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀。
永磁同步电机逆变器非线性补偿控制
测器观测出反电势,进而经过一系列计算间接获
取转子位置和转速信息。由此看来,反电势观测
的准确度成为无位置传感器滑模控制的关键。
然而,随着滑模观测法的永磁同步电机无传感
器控制应用领域逐渐向高性能场合发展,一些关
键问题逐渐体现:如逆变器非线性、磁饱和以及
转子磁场谐波导致电机转子位置估计精度降低
波法[4]、模型参考自适应法[5]、载波频率成分法[6]
观测算法广泛应用于无位置传感器永磁同步电
机转子位置以及转速检测中[1]。首先通过滑模观
以及滑模观测法[7-8]等。其中,滑模观测法以结构
简单、抗干扰性强的优点,被广泛应用于永磁同
步电机无传感器驱动控制实际工程中。文献[9]
提出一种新型滑模观测器来实现永磁同步电机
等等[2]。
近年来,针对永磁同步电机无位置传感器控
制问题,国内外学者提出多种转子位置以及转速
基金项目:2017 年度校科研基金项目(11130300117014);
2017 年江苏省知识产权软科学研究计划项目(JSIP-2017-R-9)
作者简介:李玮(1972-),
女,
本科,
副教授,
Email:
czgclw@
中图分类号:TM301
文献标识码:
A
DOI:
10.19457/j.1001-2095.dqcd18994
Nonlinear Compensation Control of Inverter for Permanent Magnet Synchronous Motor
LI Wei
(School of Intelligent Equipment and Information Engineering,Changzhou Vocational
基于反推自适应控制的永磁同步电机摩擦力矩补偿策略
⎧ did ⎪ L dt = ud − Rs id + Lpωiq ⎪ ⎪ diq = uq − Rs iq − Lpωid − ψ f pω ⎪L ⎨ dt ⎪T = 1.5ψ pi f q ⎪ e ⎪ dω = Te − TL − Tf ⎪J ⎩ dt
(1)
由式(3)可知,摩擦力矩表现出强烈的非线性特性, 传统的 PI 控制系统带宽有限, 很难有效抑制高频扰 动。ω<0 的情况与上述情况类似。因此,低速下摩 擦力矩对系统的影响比较明显。
76
第 33 卷 第 33 期 2013 年 11 月 25 日
中
国 电 机 工 程 学 Proceedings of the CSEE 中图分类号:TM 351
报
Vol.33 No.33 Nov.25, 2013 ©2013 Chin.Soc.for Elec.Eng. 学科分类号:470⋅40
文章编号:0258-8013 (2013) 33-0076-09
Fs>Fc。
稳态时,dz/dt=0,以转速ω>0 为例,式(2)表示 的 LuGre 模型可以写为一静态模型:
Tf = σ 0 Fc + σ 0 ( Fs − Fc )e− (ω /ωs ) + σ 2ω
2 2
(3)
1 数学模型及摩擦力矩特性分析
1.1 永磁同步电机数学模型 本文以表贴式永磁同步电机为被控对象,即 Ld=Lq=L。 永磁同步电机在 d-q 两相旋转坐标系中的 数学模型为
LuGre 摩擦力矩模型,其表达式为
⎧ g (ω ) = Fc + ( Fs − Fc )e− (ω /ωs ) ⎪ ⎪ dz ω ⎪ =ω − z ⎨ dt g (ω ) ⎪ ⎪ dz Tf = σ 0 z + σ 1 + σ 2ω ⎪ dt ⎩
基于扩张状态观测器的PMSM积分时变滑模控制
电气传动2021年第51卷第7期摘要:为了实现永磁同步电机(PMSM )驱动系统的高精度跟踪控制,提出了新型积分时变滑模控制策略,该策略考虑到系统的非线性和耦合特性对动、静态性能的影响,首先采用反馈线性化原理将系统模型线性化,然后为了加快动态响应过程,采用单回路结构取代串级结构设计积分时变滑模控制器。
针对负载扰动的问题,设计了一种以负载转矩为观测对象的扩张状态观测器,并将观测值反馈到控制器中以克服扰动对性能的影响。
最后在永磁同步电机实验平台上开展了对比实验研究,通过实验结果可以看出,积分时变滑模控制器使系统具有无超调、快速性的优点,提高了系统的动态和稳态性能,扩张状态观测器能够快速跟踪负载的变化,增强了控制器对负载扰动的鲁棒性。
关键词:永磁同步电机;积分时变滑模控制;反馈线性化;扩张状态观测器中图分类号:TP351文献标识码:ADOI :10.19457/j.1001-2095.dqcd21086Integral and Time -varying Sliding Mode Control of PMSM Based on Extended State ObserverLIU Jiawen ,YU Haisheng(College of Automation ,Qingdao University ,Qingdao 266071,Shandong ,China )Abstract:A novel integral and time-varying sliding mode control strategy was investigated to realize the high accuracy tracing control for the permanent magnet synchronous motor (PMSM )drive system.The influence of the nonlinear and the coupling characteristic on the dynamic and static performance of the system was considered in this strategy.Firstly ,the linearization model of PMSM was derived from feedback linearization technology.Then ,in order to accelerate the dynamic response process ,the single-loop structure was adopted instead of the cascade control to design the integral and time-varying sliding mode controller.Aiming at the problem of load disturbance ,an extended state observer with load torque as observation object was designed ,and the estimated value was fed back into the controller to overcome the influence of disturbance on performance.Finally ,the comparative experiment study was carried out on the PMSM experimental platform.Experimental results show that the integral and time-varying sliding mode controller can make the system has the advantages of rapidity and no over-shoot ,improve the dynamic and static performances.The extended state observer can observe the change of the load torque rapidly and enhance the robustness of the controller to load disturbance.Key words:permanent magnet synchronous motor (PMSM );integral and time-varying sliding mode control ;feedback linearization ;extended state observer基金项目:国家自然科学基金(61573203)作者简介:刘佳雯(1994—),女,硕士研究生,Email :*****************通讯作者:于海生(1963—),男,博士,教授,博导,Email :***************.cn基于扩张状态观测器的PMSM 积分时变滑模控制刘佳雯,于海生(青岛大学自动化学院,山东青岛266071)永磁同步电机(permanent magnet synchro⁃nous motor ,PMSM )由于其固有的低转子惯性、高效率、结构坚固、高功率密度等优点,在电动汽车、风力发电系统、机器人等各种工业应用中受到了广泛关注[1-3]。
永磁同步电机矢量控制系统建模与仿真
永磁同步电机矢量控制系统建模与仿真王涛;李勇;王青;贾克军【摘要】基于永磁同步电机具有多变量、非线性的复杂特性,为研究需要,对其物理模型进行简化,建立了电机的数学模型及其基本方程.在矢量控制众多方法中采用最为简单的使直轴电流id=0方法进行研究,得到了基于转子磁场定向矢量控制下的电机电磁转矩方程.在Matlab/Simulink搭建整个系统仿真模型、转速和电流控制模块,并对这些模块进行仿真.仿真结果表明所得波形符合理论分析,系统响应快、超调量小,系统运行稳定,具有良好的动、静态特性.该模型的建立和分析对电机的实际控制提供了新的研究思路.%Based on the complex system of Permanent Magnetic Synchronous Motor (PMSM) with multi-variable and nonlinear, in this paper, the physical model of PMSM is simplified and the mathematical model of the motor is established in order to facilitate research. This paper uses id = 0 control manner which is the simplest manner in vector control methods, motor electromagnetic torque equation is established based on rotor field oriented vector control. The system model,speed and current control block are built and simulated with Matlab/Simulink. Simulation results show that the waveform is consistent with theoretical analysis; the model has fast response and small overshoot. The system runs stably with good dynamic and static characteristics. So,the establishment and analysis of PMSM model provide a new study for its actual control.【期刊名称】《河北大学学报(自然科学版)》【年(卷),期】2011(031)006【总页数】5页(P648-652)【关键词】永磁同步电机;矢量控制;建模;仿真【作者】王涛;李勇;王青;贾克军【作者单位】河北大学质量技术监督学院,河北保定071002;北京科技大学车辆工程研究所,北京100083;河北大学质量技术监督学院,河北保定071002;河北大学质量技术监督学院,河北保定071002【正文语种】中文【中图分类】TH39永磁同步电机与励磁同步电机相比取消了励磁电源和励磁绕组,取而代之的是能够产生稳定磁场的永磁体,这就使得永磁同步电机结构更加紧凑,重量减轻,体积减小,又由于同时也取消了励磁系统的损耗,其效率、功率因数得到了很大的提高[1-2].永磁同步电机的励磁磁场由转子上的永磁体产生,按转子磁场定向的矢量控制实现类似于直流电机对转矩和转子磁链的分别控制,从而获得类似于直流电机的宽范围调速性能.随着电力电子技术和控制技术的发展,永磁同步电机具有精度高、动态性能好、调速范围大以及定位控制准确等优点,常被应用于伺服系统和高性能的调速系统,因此引起了国内外越来越多学者的广泛关注[3].本文对永磁同步电机建立数学模型得到其基本方程,对矢量控制众多控制方法中最为简单的id=0方法进行研究,在Matlab/Simulink平台下建立该控制方法的仿真模型并进行仿真,并对仿真结果进行分析.该模型的建立和分析对电机的实际控制提供了新的研究思路.1.1 永磁同步电机基本结构永磁同步电机的定子与一般交流电机的定子绕组相同,采用三相交流绕组.定子铁心由带有齿和槽的冲片叠成,在槽中嵌入交流绕组.当三相对称电流通入三相对称绕组时,在气隙中产生同步旋转磁场,为简化问题同时又不影响数学模型的精度,常作如下假设:1)气隙磁场即永磁体产生的励磁磁场和三相绕组产生的电枢反应磁场呈正弦分布,定子三相绕组磁通产生的感应电动势也呈正弦分布;2)由于永磁同步电机的气隙比较大,所以不计定子磁路的饱和和铁损;3)转子上没有阻尼绕组,永磁体没有阻尼作用[4-5].1.2 永磁同步电机基本方程将永磁同步电机模型建立在三相静止坐标系(abc坐标系)上,可得到其各绕组电压平衡方程[6-7]式中,ea,eb,ec 为永磁体磁场在a,b,c三相电枢绕组中感应的旋转电动势,Rs 为定子绕组电阻,La,Lb,Lc 为定子绕组自感,Mab,Mbc,Mca为绕组间的互感.由于转子结构不对称,将abc坐标系(三相静止坐标系)中的a,b,c三相绕组先变换到αβ坐标系(两相静止坐标系),然后再由αβ坐标系变换到dq坐标系(两相旋转坐标系)中.采用的坐标变换关系式为[8-11]得到dq坐标系上的电压方程为dq向abc转换关系如式(5)所示.式中,Ld,Lq 为定子绕组自感,id,iq 为d,q轴电流分量,Rs 为定子绕组电阻,ud,uq 为d,q轴电压分量,ωr 为转子角速度,ψf =ψfm/2,ψfm 为与定子a,b,c三相绕组交链的永磁体磁链的幅值.电机在dq坐标系中转矩方程为永磁同步电机的矢量控制方法有很多种,其中使直轴电流id=0控制是最常用的方法.此时电流矢量随负载状态的变化在q轴上移动.根据式(4),id=0时的电磁转矩为.采用该方法消除了直轴电流带来的电枢反应,电机所有电流都用来产生电磁转矩,电流控制效率得到提高,产生最大的电磁转矩.永磁同步电机矢量控制结构图1所示.根据永磁同步电机矢量控制结构图[12-15],在Matlab/Simulink中搭建仿真模型,如图2所示.本文采用永磁同步电机电流、速度的双闭环控制,如图3所示.内环为电流环,外环为速度环.将电流环看作是速度调节系统中的一个环节,其作用是提高系统的快速性,抑制电流环内部干扰,限制最大电流以保障系统安全运行,速度环的作用是增强系统抗负载扰动的能力,抑制速度波动[16].转速调节模块如图4所示.该模块由PI调节器和限幅输出模块组成.通过反复调整kp,ki参数使系统输出达到最佳状态.电流调节其实就是转矩调节模块,将转速调节器的输出电流作为转矩调节器的输入.该模块也由PI调节器和限幅输出模块组成,电流调节模型图与转速调节模型图相同[17-18].仿真参数设置:逆变器直流电源电压380V,永磁同步电机定子绕组电阻Rs=2.67Ω,d轴电感Ld=0.007H,q轴电感Lq=0.007H,极对数p=2,电机转动惯量J=0.006kg·m2.电机空载启动,启动转速给定n=3 000r/min;待系统进入稳态后在0.05s时突加Tl=6N·m的负载,仿真时间t=0.1s.仿真结果如图5a-c 所示.从图5a中可以看出电机在启动后的0.02s内转速快速上升,并在经过0.01s的波动之后迅速达到稳定状态,电机动态响应性能良好.图5b中看出0.03s之前出现很大的振荡,这是因为电机启动初期转子转速低于定子旋转磁场转速,定子磁链和永磁体磁链产生的转矩在较短的时间内起到制动作用.当牵引转矩小于制动转矩时,电机总转矩下降,从而出现振荡现象.在0.05s突加6N·m的负载时,转速、转矩均有相应响应,但经过短暂的波动之后均达到稳定状态.由于仿真过程中使用PWM逆变器供电,定子电流中出现一定的谐波分量,影响到电磁转矩,使转矩和转速均出现一定的脉动,但不影响系统的稳定性.图5c为电机的机械特性曲线,可以看出机械特性较为理想.在分析永磁同步电机数学模型的基础之上,建立了电机的数学方程,通过数学的方法去研究永磁同步电机,并在Matlab/Simulink里搭建模型并进行仿真.由电机仿真波形可以看出,系统响应快速且平稳,转速和转矩超调量非常小,系统起动后保持恒定转矩;突加扰动时系统波动较小,充分说明系统具有较好的鲁棒性.仿真结果证明了本文所提出的永磁同步电机仿真建模方法的有效性.【相关文献】[1]曾毅.变频调速控制系统的设计和维护[M].2版.济南:山东科学技术出版社,2002.[2]张铁军.永磁同步电机数字化控制系统研究[D].长沙:湖南大学,2006.[3]王成元.电机现代控制技术[M].北京:机械工业出版社,2007.[4]杨文峰,孙韶元.参数自调整模糊控制交流调速系统的研究[J].电工技术杂志,2001(9):11-13.[5]BARRERO F,GONZÁLEZ A ,TORRALBA A,et al.Speed control of induction motors using a novel fuzzy sliding mode structure[J].IEEE Transactions on Fuzzy Systems,2002,10(3):375-380.[6]薛峰,谢运祥,吴捷.直接转矩控制系统的转速估算模型及其参数补偿方法[J].电工技术学报,1998,13(5):26-30.[7]EBERHART R,KENNEDY J.A new optimizer using particl swarm theory[Z].Proceedings of Sixth International Symposium MicroMachine and Human Science,Nagoya,Japan,1995.[8]陈伯时.电力拖动自动控制系统[M].2版.北京:机械工业出版社,2001.[9]陈荣.永磁同步电机伺服系统研究[D].南京:南京航空航天大学,2004.[10]黄永安,马路,刘慧敏.MATLAB 7.1/Simulink 6.1建模仿真开发与高级工程应用[M].北京:清华大学出版社,2005.[11]李学文,李学军.基于SIMULINK的永磁同步电机建模与仿真[J].河北大学学报:自然科学版,2007,27(S1):28-31.[12]BOUCHIKER S,CAPOLINO G A.Vector control of a permanent magnet synchronous motor using AC matrix converter[J].IEEE Transactions on Power Electronics,1998,13(6):1089-1099.[13]沈艳霞,吴定会,李三东.永磁同步电机位置跟踪控制器及Backstepping方法建模[J].系统仿真学报,2005,17(6):1318-1321.[14]薛花,姜建国.基于EKF永磁同步电机FMRC方法的仿真研究[J].系统仿真学报,2006,18(11):3324-3327.[15]林伟杰.永磁同步电机两种磁场定向控制策略的比较[J].电力电子技术,2007,41(1):26-29.[16]LI Yong,MA Fei,CHEN Shunxin,et al.PMSM simuation for AC drive in mining dump truck[Z].The Ninth International Conference on Information and Management Sciences(IMS2010),Urumchi,2010.[17]KENNEDY J,EBERHART R.Particle swarm optimization[Z].Pro IEEE Int Conf on Neural Networks,Perth,1995.[18]钱昊,赵荣祥.永磁同步电机矢量控制系统[J].农机化研究,2006(2):90-91.。
基于内模控制的永磁同步电机控制
2 2. 1
内模解耦控制系统设计 内模控制基本原理
内模控制结构如图 1 所示。
控制 策 略 起 初 应用 在 耦 合 强、 干 扰 大、 图1
^
非线性的 控制 系统 中。 它 具 有 对 系统 模型准 确 度 要 求 低, 对不可测干 扰 能 够 消 除,能 较 快 的 跟踪 系统 的 输 入, 鲁 棒 性能好等优点。 文章应用内 模 控制方法进行 了 调 节 器 的设计, 并 在 此 基础上对基于内模 控制 理 论 的永 磁 同 步 电 机 调 速 系统 进行 收稿日期: 2012 - 07 - 20 控制器
Study on Control of Permanent Magnetic Synchro Motor Base on Internal Mould Control
WANG Dun - sheng
( Xuzhou China Mining Drivers&Automation Co. Ltd,Jiangsu Xuzhou 221116 ,China)
[1 ]
1
概
述
永磁同步电动机( PMSM) 因其具有体积小、效率高、转 矩惯性比和能量密 度 高等 特 点 ,使 得 永 磁 同 步 电 动 机 驱 动逐渐取代 传 统 的 交 直 流 驱 动 系统, 得 到 了 广 泛 的 应用。 但它又是非 线 性、 多变量、 强 耦 合、 时变 的 系统。 因此 研 究如何获得高性 能的永 磁 同 步 电 机 的 控制 策 略, 实 现 电 机 高性能控制是具有实际意义。 传 统 PI 控制 在永 磁 同 步 电 动 机 矢 量 控制中 被 广 泛 应 用,但由于算法 本 身 对 电 机 本 体 参 数 的 依 赖, 使 得 其 鲁 棒 性较差。 内 模
永磁同步电动机英文翻译
英文原文Research on Voltage Space-vector Control System of Synchronous Motor Vector control of field oriented control, the basic idea is: through coordinate transformation control method for simulation of DC motor to control the permanent magnet synchronous motor. Three-phase symmetrical windings in three-phase AC can produce a rotating magnetic motive force, two phase symmetrical windings into two symmetric alternating current can produce the same rotating magnet ometive force; therefore the three-phase symmetric winding can be replaced with two phase symmetrical windings equivalent independent of each other, equivalent principle is the constant magnetomotive force produced before and after transformation, transformation and total power constant.In oil field, the power factor was reduced and the reactive power consumption was increased because of the usage of the large number of asynchronous motor, and resulting in a huge waste of energy, which reduced the integrated cost-effective of field. The permanent magnet synchronous motor possess all the advantages of synchronous motor and it has high efficiency and higher power factor. For the advantages of permanent magnet synchronous,it will bring good energy saving results if it is used in pumping unit. As a result,the study on permanent magnet synchronous motor control system is important.In this paper the theory of vector control system on PMSM is first deeply studied,and the idea of coordinate transformation is used to build the mathematical model of PMSM. An in-depth theoretical analysis of voltage space vector control algorithm is done. Secondly,based on the mathematical model of permanent magnet synchronous motor and SVPWM theory,the model of PMSM vector control system is established by of Matlab/Simulink. The simulation result shows the possibility of using the control system.In the paper, the software and hardware of PMSM vector control system is designed core-based TI Company’s motor control DSP chip TMS320LF2407A. Hardware ncludes the main circuit,control circuit and its peripheral circuits;software contains the main program and SVPWM interrupt subroutine,it achieves the implementation of the dual closed-loop current. At last,the motor experiments are carried on under the laboratory,the experimental results verify the correctness of the hardware and control program.Permanent magnet synchronous motor with the advantages of simple structure,high efficiency,wide speed range,widely used in machining,aerospace and electric traction fields,this paper introduces the structure,control strategy of permanent magnet synchronous motor and its vector torque control research present situation and direction.Based on space vector principle,the three kind of coordinate systems as well as the transformation of them which usually used in motor’s speed control system areintroduced,then,the mathematic models on different coordinate systems are derived,be based on that,the principle of traditional direct torque control system as well as the direct torque control system based on SVPWM are analyzed detailed,meanwhile,the realization process of SVPWM algorithm is derived.Finally,the simulation model of convientional DTC Control system are established in MATLAB/Simulink.Control of permanent magnet synchronous motor mainly in the following1.1 vector controlThe core idea of vector control of three-phase current,voltage,the flux of the motor by coordinate transformation into the rotor flux oriented phase reference coordinate system, control idea according to DC motor, control motor torque.The advantages of the field oriented vector control is good torque response,precise speed control,zero speed can achieve full load.However,the vector control system needs to determine the rotor flux,to coordinate transformation,a large amount of calculation,but also consider the effect of changes in the rotor of the motor parameters,which makes the system more complex,this is the vector control deficiencies.1.2 direct torque controlIt is based on stator flux orientation,implementation of direct control of stator flux and torque.The control is based on the idea of amplitude real-time detection of motor torque and flux are given,and the torque and flux linkage value comparison,the torque and flux adjusting the appropriate stator voltage space vector selection table switch calculated directly from an offline,power switch and control of inverter state.Direct torque control does not need the vector coordinate transformation complex,the motor model is simplified,no pulse width modulation signal generator,control has the advantages of simple structure,motor parameter changes,can obtain good dynamic performance.But there are also some shortcomings,such as the inverter switching frequency is not fixed,large torque ripple current to realize digital control requires high sampling frequency.1.3 direct torque control based on space vector modulation(SVM-DTC)The SVM-DTC control is the vector control and direct torque control together,its theory foundation and DTC control theory,is based on torque angle control.According to the change of torque angle and flux vector position,get the flux of the next cycle position,which can be the reference voltage vector is required,then the reference voltage vector modulation,PWM wave inverter driving.The SVM-DTC control,the flux changes to determine the next position,so the accurate estimation of flux has great effect on the control system,and the flux estimation depends on motor parameters are stable.In addition,the electromagnetic torque and torque angle is a nonlinear relationship,but in the practical application is approximately linear,using PIregulation,performance so that the PI parameters can also affect the system.1.4 The model reference adaptive control(MRAS)The model system requirements of the control system with a model for the adaptive control,the output response model is ideal,this model is called the reference model.The system always tries to make dynamic consistency can bedynamic reference model and the adjustable model in operation.By comparing the output of reference model and actual process,and through the adaptive controller to adjust some parameters of the adjustable model or generate anauxiliary input,so that the output error between actual output and the reference model as small as possible.In practical application,usually used for speed estimation,to realize the speed sensor less operation.Therefore,the model reference adaptive depends mainly on the accuracy of the adjustable model,the stable operation of the system plays a decisive role in.In addition,the adaptive control law parameters tuning is a difficult problem,the control accuracy of the control system has a great impact.1.5The state observer based controlControl based on state observer is developed based on the modern control theory,observer based on the mathematical model of permanent magnet synchronous motor,used for each observation control system and the state,thus extracting speed control.It is also dependent on the accuracy of the motor model,the appearance of large error will run at low speed or increasing temperature leads to the variation of motor parameters,so as to bring large deviation to control.1.6 intelligent controlThe use of intelligent algorithms,intelligent control of the control system, such as fuzzy control,neural network control,self-tuning parameters and so on,through one or several times after the trial operation, automatic parameter tuning out,to realize the optimization control.Intelligent control has many advantages,especially in the motor is multi variable,nonlinear control system,however,control and its performance depends on the control object,that is to say not every control system can achieve good control,which require sexperience.At the same time,the large amount of computation,but also has certain requirements for the controller.Synchronous Motor because of having power factor higher run – time efficiency higher , stability good, the revolving speed settles to wait a merit, is extensively been applied to industrial production amid. The starting fault that acquaints with synchronous motor, and debugging in time, all have important meaning to the motor and the production systems . By way of energy in time, accurate debugging and transaction fault, have the familiar faultprogress of the synchronous motor in detail analytical!2 Familiar fault2.1 The synchro motor after switching on electricity the incapability startsThe synchro motor after starting the incapability run - time generally has the reason of severals as follows:(1)Power supply voltage over low.Because at the square of voltage, the starting torque direct proportion of synchro motor's the voltage of power supply over make low the starting torque of synchro motor significantly the droop is lower than load troque, can not start thus and want to raise vs this power supply voltage to enlarge the starting torque of dynamo.(2)The fault of motor. Check motor settle, the rotor winding had no short circuit, open circtui, open soldering and link bad etc. fault, these the faults will make the dynamo can not start to create starting of rating of intensity of magnetic field, make thus the dynamo can not start;Checking the motor bearing has already had no failure, the port cap has have no loose, if bearing failure port shroud loose, result in bearing's down sinking, mutually rub with stator iron core, result in thus dynamo's canning not start, vs settle the rotor fault can be shaken table with the low tension, gradually click to check to seek a fault condition and adopt homologous treatment;The countersgaft accepts and carries to shroud a loose condition and all wants a pan car before driving each time and sees motor rotor whether slewing is vivid, if bearing or shaft kiowatt damage and replace in time.(3)The control device breaks down.This kind of faults are mostly the d.c. output voltage of the windings of Li magnetic belt to adjust not appropriate or don't output, result in the stator current of motor over big, cause the motor conduct electricity the run make or the losing of dynamo magnetic belt run - time.Should check whether output voltage current and its waveform that the Li magnetic belt equips is normal at this time, the Rong breaks whether the machine Rong breaks, the contact is bad;Whether circuit board plug-in puts prison or alignment;Check loop resistance, put out whether crystal gate tube of magnet burns out or brokes through.(4)Mechanical trouble. Such as be dragged along a dynamic machinery to block, result in motor incapability's starting, the rotor that moves motor in response to the pan at this time sees whether the slewing is vivid, machinery burden whether existence fault2.2The synchro motor incapability leads long into synchronization.Synchro motor in common use law of nonsynchronous starting,throw in Li magnetic belt when the motor rotor revolving speed hits synchronous revolving speed of 95%, make it leads long into synchronization. The synchro motor incapability leads long into synchronous reason as follows:(1)The Li magnetic belt winding short circuit.Because the winding of Li magnetic belt, existence short circuit breaks down, as a result makes motor able to stabilize run - time but incapability and lead long into synchronization while being lower than synchronous revolving speed. Check to seek the Li magnetic belt winding short circuit, can open into low - tension(about the 30 Vs) in the rotor derivation on - line, put on the magnetic poles surface with a hand work steel saw, pursue inspection magnetic poles, if vibrating is violent, explain the magnetic poles to have no short circuit on steel saw of the magnetic poles' surface, if the vibrating of saw blade micro or don't flap, explain the magnetic poles short circuit. After unloading the magnetic poles, check the fault to click,is short-circuit degree, adopt local to mend or re- round to make.(2) Power supply voltage over low. Power supply voltage over low, result in the strong Li link of the device of Li magnetic belt incapability working, make the motor incapability lead long into synchronization thus, the concrete way is to raise power supply voltage appropriately.(3) The fault of Li magnetic belt device. Such as throw Li over speedy(namely throw in Li magnetic belt, motor rotor revolving speed over low), will make the motor can not lead long into synchronization, should check to throw if the Li link exists fault at this time. If Li magnetic belt device fault, the output's current is lower than a rating value, cause the electricity magnetic troque of dynamo over small but can not lead long into synchronization, at this time in response to scrutiny Li magnetic belt device of throw Li link and phase - shifting link, waveform use oscillo graph to check to throw Li link and phase - shifting link, should also check and put out magnetic belt link and put out crystal gate of magnetic belt whether tube discovers a question as usual, handle in time, if the incapability handles in time, by way of the energy quickly restore capacity, should replace to provide for use circuit board.2.3 Brush and compress tightly spring and gather to give or get an electric shock ring fault.The brush leads short and compresses tightly spring press scarcity and make brush and gather to give or get an electric shock ring of indirectly touch badly, thus generate spark or arc electric, arc electric or spark to on the other hand and easily spark short circuit, will make arc electric burn on the other hand shorter, spark open circtui thus, result in Li magnetic belt device only the Li magnetoelectricity press but have no Li magnetoelectricity streaming;Compress tightly spring ageing lapse, make brush and gather to give or get an electric shock ring of indirectly touch badly, effect the starting of motor thus;Gather to give or get an electric shock a ring surface to there is grease stain and scar or slot scar, will make brush and gather to give or get an electric shock ring of indirectly touch badly, generate spark, spark further burn gather to give or get an electric shock ring, will also make gnd short-circuit, the spark effects the starting of motor thus.For gather to give or get an electric shock ring superficial grease stain, can wipeto clean with the acetone; For thin trace, use many fettle shagging rings of sandpapers surface, is ring surface roughness to hit R1.6 ums, if the slot scar obviously needs to get on the car bed transform, truning, enter amount of knife to take every time 1 mm as proper, in the 1-1.5 ms/s, the truning speed control's roughness hits of the ums of R1.5-1.8 and becomes bad anti to finally polish with the sandpaper 2-3 times over the 0.05 mms.2.4 The damper winding breaks down.The damper winding of synchro motor rotor is provided for synchro motor starting to use and wipe - out run - time at the same time amid spark because of loading to change of out of step osc.Start the damper winding in the process to incise the magnetic field of stator revolution but induced very big starting current in the synchro motor, so the big current by all means will result in damping hair thermal expansion, under the normal condition because of starting time short, the damper winding starting is behind soon will cool off, but block up revolution in the motor, lack phase, start the super - in time to length ways wait a condition down, if don't shut down in time, will result in the damping take off soldering to split etc. condition.The damper winding is weaker link in the synchro motor parts, the damper winding familiar fault has:The damping takes off soldering and split, the damping ring discharges wildfire, damping ring the strain is serious.These faults will effect the starting of synchro motor. The damping takes off soldering and chooses silver actinium welding rod and adopts oxyacetylene welding to weld, the dynamo after taking out the core heats into rotor 200 Celsius degrees set rotor vertical in the oven, after taking out and adopt 750 Celsius degrees to or soly weld temperature, damping and the blind side of of damping ring complete solderings are full, clear a soldering dirt again, ;For split of the damping , after dismantling original damping, choose the material of material homology and adopt the above-mentioned method to weld after packing good damping.Damping ring the wildfire is mainly what damping ring indirectly touches bad or get in touch with area isn't enough to result in. Damping ring the strain seriously is mainly a damping to fix anticoincidence in the slot, the damping plugs into damping ring while welding hole falsely, appear additional stress after welding, at plus damping ring intensity not enough to, treatment is loose open all connectivity bolts of damping rings, vs strain anti big of damping ring, after oxyacetylene welding heating adjust with the exclusive use fixture even, vs strain serious replace a new damping of ring.3 ConclusionWhen the synchro motor appears fault, cautiously analytical possible reason, gradually expel, look into related data when it's necessary, absorb experience, propose corrective actions.Analytical the dynamo fault not only need to have firm theory knowledge and experience of prolific maintenance repairs, but also need to aim at concrete fault, deepconsideration, brave creative, the dynamo after ensuring to break down removal can stabilize run - time over a long period of time.中文翻译永磁同步电动机矢量控制系统(中文对照)矢量控制亦称磁场定向控制,其基本思路是:通过坐标变换实现模拟直流电机的控制方法来对永磁同步电机进行控制。
永磁同步电动机英文翻译
英文原文Research on Voltage Space-vector Control System of Synchronous Motor Vector control of field oriented control, the basic idea is: through coordinate transformation control method for simulation of DC motor to control the permanent magnet synchronous motor. Three-phase symmetrical windings in three-phase AC can produce a rotating magnetic motive force, two phase symmetrical windings into two symmetric alternating current can produce the same rotating magnet ometive force; therefore the three-phase symmetric winding can be replaced with two phase symmetrical windings equivalent independent of each other, equivalent principle is the constant magnetomotive force produced before and after transformation, transformation and total power constant.In oil field, the power factor was reduced and the reactive power consumption was increased because of the usage of the large number of asynchronous motor, and resulting in a huge waste of energy, which reduced the integrated cost-effective of field. The permanent magnet synchronous motor possess all the advantages of synchronous motor and it has high efficiency and higher power factor. For the advantages of permanent magnet synchronous,it will bring good energy saving results if it is used in pumping unit. As a result,the study on permanent magnet synchronous motor control system is important.In this paper the theory of vector control system on PMSM is first deeply studied,and the idea of coordinate transformation is used to build the mathematical model of PMSM. An in-depth theoretical analysis of voltage space vector control algorithm is done. Secondly,based on the mathematical model of permanent magnet synchronous motor and SVPWM theory,the model of PMSM vector control system is established by of Matlab/Simulink. The simulation result shows the possibility of using the control system.In the paper, the software and hardware of PMSM vector control system is designed core-based TI Company’s motor control DSP chip TMS320LF2407A. Hardware ncludes the main circuit,control circuit and its peripheral circuits;software contains the main program and SVPWM interrupt subroutine,it achieves the implementation of the dual closed-loop current. At last,the motor experiments are carried on under the laboratory,the experimental results verify the correctness of the hardware and control program.Permanent magnet synchronous motor with the advantages of simple structure,high efficiency,wide speed range,widely used in machining,aerospace and electric traction fields,this paper introduces the structure,control strategy of permanent magnet synchronous motor and its vector torque control research present situation and direction.Based on space vector principle,the three kind of coordinate systems as well as the transformation of them which usually used in motor’s speed control system areintroduced,then,the mathematic models on different coordinate systems are derived,be based on that,the principle of traditional direct torque control system as well as the direct torque control system based on SVPWM are analyzed detailed,meanwhile,the realization process of SVPWM algorithm is derived.Finally,the simulation model of convientional DTC Control system are established in MATLAB/Simulink.Control of permanent magnet synchronous motor mainly in the following1.1 vector controlThe core idea of vector control of three-phase current,voltage,the flux of the motor by coordinate transformation into the rotor flux oriented phase reference coordinate system, control idea according to DC motor, control motor torque.The advantages of the field oriented vector control is good torque response,precise speed control,zero speed can achieve full load.However,the vector control system needs to determine the rotor flux,to coordinate transformation,a large amount of calculation,but also consider the effect of changes in the rotor of the motor parameters,which makes the system more complex,this is the vector control deficiencies.1.2 direct torque controlIt is based on stator flux orientation,implementation of direct control of stator flux and torque.The control is based on the idea of amplitude real-time detection of motor torque and flux are given,and the torque and flux linkage value comparison,the torque and flux adjusting the appropriate stator voltage space vector selection table switch calculated directly from an offline,power switch and control of inverter state.Direct torque control does not need the vector coordinate transformation complex,the motor model is simplified,no pulse width modulation signal generator,control has the advantages of simple structure,motor parameter changes,can obtain good dynamic performance.But there are also some shortcomings,such as the inverter switching frequency is not fixed,large torque ripple current to realize digital control requires high sampling frequency.1.3 direct torque control based on space vector modulation(SVM-DTC)The SVM-DTC control is the vector control and direct torque control together,its theory foundation and DTC control theory,is based on torque angle control.According to the change of torque angle and flux vector position,get the flux of the next cycle position,which can be the reference voltage vector is required,then the reference voltage vector modulation,PWM wave inverter driving.The SVM-DTC control,the flux changes to determine the next position,so the accurate estimation of flux has great effect on the control system,and the flux estimation depends on motor parameters are stable.In addition,the electromagnetic torque and torque angle is a nonlinear relationship,but in the practical application is approximately linear,using PIregulation,performance so that the PI parameters can also affect the system.The model reference adaptive control(MRAS)The model system requirements of the control system with a model for the adaptive control,the output response model is ideal,this model is called the reference model.The system always tries to make dynamic consistency can bedynamic reference model and the adjustable model in operation.By comparing the output of reference model and actual process,and through the adaptive controller to adjust some parameters of the adjustable model or generate anauxiliary input,so that the output error between actual output and the reference model as small as possible.In practical application,usually used for speed estimation,to realize the speed sensor less operation.Therefore,the model reference adaptive depends mainly on the accuracy of the adjustable model,the stable operation of the system plays a decisive role in.In addition,the adaptive control law parameters tuning is a difficult problem,the control accuracy of the control system has a great impact.1.5The state observer based controlControl based on state observer is developed based on the modern control theory,observer based on the mathematical model of permanent magnet synchronous motor,used for each observation control system and the state,thus extracting speed control.It is also dependent on the accuracy of the motor model,the appearance of large error will run at low speed or increasing temperature leads to the variation of motor parameters,so as to bring large deviation to control.intelligent controlThe use of intelligent algorithms,intelligent control of the control system, such as fuzzy control,neural network control,self-tuning parameters and so on,through one or several times after the trial operation, automatic parameter tuning out,to realize the optimization control.Intelligent control has many advantages,especially in the motor is multi variable,nonlinear control system,however,control and its performance depends on the control object,that is to say not every control system can achieve good control,which require sexperience.At the same time,the large amount of computation,but also has certain requirements for the controller.Synchronous Motor because of having power factor higher run – time efficiency higher , stability good, the revolving speed settles to wait a merit, is extensively been applied to industrial production amid. The starting fault that acquaints with synchronous motor, and debugging in time, all have important meaning to the motor and the production systems . By way of energy in time, accurate debugging and transaction fault, have the familiar faultprogress of the synchronous motor in detail analytical!2 Familiar fault2.1 The synchro motor after switching on electricity the incapability startsThe synchro motor after starting the incapability run - time generally has the reason of severals as follows:(1)Power supply voltage over low.Because at the square of voltage, the starting torque direct proportion of synchro motor's the voltage of power supply over make low the starting torque of synchro motor significantly the droop is lower than load troque, can not start thus and want to raise vs this power supply voltage to enlarge the starting torque of dynamo.(2)The fault of motor. Check motor settle, the rotor winding had no short circuit, open circtui, open soldering and link bad etc. fault, these the faults will make the dynamo can not start to create starting of rating of intensity of magnetic field, make thus the dynamo can not start;Checking the motor bearing has already had no failure, the port cap has have no loose, if bearing failure port shroud loose, result in bearing's down sinking, mutually rub with stator iron core, result in thus dynamo's canning not start, vs settle the rotor fault can be shaken table with the low tension, gradually click to check to seek a fault condition and adopt homologous treatment;The countersgaft accepts and carries to shroud a loose condition and all wants a pan car before driving each time and sees motor rotor whether slewing is vivid, if bearing or shaft kiowatt damage and replace in time.(3)The control device breaks down.This kind of faults are mostly the d.c. output voltage of the windings of Li magnetic belt to adjust not appropriate or don't output, result in the stator current of motor over big, cause the motor conduct electricity the run make or the losing of dynamo magnetic belt run - time.Should check whether output voltage current and its waveform that the Li magnetic belt equips is normal at this time, the Rong breaks whether the machine Rong breaks, the contact is bad;Whether circuit board plug-in puts prison or alignment;Check loop resistance, put out whether crystal gate tube of magnet burns out or brokes through.(4)Mechanical trouble. Such as be dragged along a dynamic machinery to block, result in motor incapability's starting, the rotor that moves motor in response to the pan at this time sees whether the slewing is vivid, machinery burden whether existence fault2.2The synchro motor incapability leads long into synchronization.Synchro motor in common use law of nonsynchronous starting,throw in Li magnetic belt when the motor rotor revolving speed hits synchronous revolving speed of 95%, make it leads long into synchronization. The synchro motor incapability leads long into synchronous reason as follows:(1)The Li magnetic belt winding short circuit.Because the winding of Li magnetic belt, existence short circuit breaks down, as a result makes motor able to stabilize run - time but incapability and lead long into synchronization while being lower than synchronous revolving speed. Check to seek the Li magnetic belt winding short circuit, can open into low - tension(about the 30 Vs) in the rotor derivation on - line, put on the magnetic poles surface with a hand work steel saw, pursue inspection magnetic poles, if vibrating is violent, explain the magnetic poles to have no short circuit on steel saw of the magnetic poles' surface, if the vibrating of saw blade micro or don't flap, explain the magnetic poles short circuit. After unloading the magnetic poles, check the fault to click,is short-circuit degree, adopt local to mend or re- round to make.(2) Power supply voltage over low. Power supply voltage over low, result in the strong Li link of the device of Li magnetic belt incapability working, make the motor incapability lead long into synchronization thus, the concrete way is to raise power supply voltage appropriately.(3) The fault of Li magnetic belt device. Such as throw Li over speedy(namely throw in Li magnetic belt, motor rotor revolving speed over low), will make the motor can not lead long into synchronization, should check to throw if the Li link exists fault at this time. If Li magnetic belt device fault, the output's current is lower than a rating value, cause the electricity magnetic troque of dynamo over small but can not lead long into synchronization, at this time in response to scrutiny Li magnetic belt device of throw Li link and phase - shifting link, waveform use oscillo graph to check to throw Li link and phase - shifting link, should also check and put out magnetic belt link and put out crystal gate of magnetic belt whether tube discovers a question as usual, handle in time, if the incapability handles in time, by way of the energy quickly restore capacity, should replace to provide for use circuit board.2.3 Brush and compress tightly spring and gather to give or get an electric shock ring fault.The brush leads short and compresses tightly spring press scarcity and make brush and gather to give or get an electric shock ring of indirectly touch badly, thus generate spark or arc electric, arc electric or spark to on the other hand and easily spark short circuit, will make arc electric burn on the other hand shorter, spark open circtui thus, result in Li magnetic belt device only the Li magnetoelectricity press but have no Li magnetoelectricity streaming;Compress tightly spring ageing lapse, make brush and gather to give or get an electric shock ring of indirectly touch badly, effect the starting of motor thus;Gather to give or get an electric shock a ring surface to there is grease stain and scar or slot scar, will make brush and gather to give or get an electric shock ring of indirectly touch badly, generate spark, spark further burn gather to give or get an electric shock ring, will also make gnd short-circuit, the spark effects the starting of motor thus.For gather to give or get an electric shock ring superficial grease stain, can wipeto clean with the acetone; For thin trace, use many fettle shagging rings of sandpapers surface, is ring surface roughness to hit R1.6 ums, if the slot scar obviously needs to get on the car bed transform, truning, enter amount of knife to take every time 1 mm as proper, in the 1-1.5 ms/s, the truning speed control's roughness hits of the ums of R1.5-1.8 and becomes bad anti to finally polish with the sandpaper 2-3 times over the 0.05 mms.2.4 The damper winding breaks down.The damper winding of synchro motor rotor is provided for synchro motor starting to use and wipe - out run - time at the same time amid spark because of loading to change of out of step osc.Start the damper winding in the process to incise the magnetic field of stator revolution but induced very big starting current in the synchro motor, so the big current by all means will result in damping hair thermal expansion, under the normal condition because of starting time short, the damper winding starting is behind soon will cool off, but block up revolution in the motor, lack phase, start the super - in time to length ways wait a condition down, if don't shut down in time, will result in the damping take off soldering to split etc. condition.The damper winding is weaker link in the synchro motor parts, the damper winding familiar fault has:The damping takes off soldering and split, the damping ring discharges wildfire, damping ring the strain is serious.These faults will effect the starting of synchro motor. The damping takes off soldering and chooses silver actinium welding rod and adopts oxyacetylene welding to weld, the dynamo after taking out the core heats into rotor 200 Celsius degrees set rotor vertical in the oven, after taking out and adopt 750 Celsius degrees to or soly weld temperature, damping and the blind side of of damping ring complete solderings are full, clear a soldering dirt again, ;For split of the damping , after dismantling original damping, choose the material of material homology and adopt the above-mentioned method to weld after packing good damping.Damping ring the wildfire is mainly what damping ring indirectly touches bad or get in touch with area isn't enough to result in. Damping ring the strain seriously is mainly a damping to fix anticoincidence in the slot, the damping plugs into damping ring while welding hole falsely, appear additional stress after welding, at plus damping ring intensity not enough to, treatment is loose open all connectivity bolts of damping rings, vs strain anti big of damping ring, after oxyacetylene welding heating adjust with the exclusive use fixture even, vs strain serious replace a new damping of ring.3 ConclusionWhen the synchro motor appears fault, cautiously analytical possible reason, gradually expel, look into related data when it's necessary, absorb experience, propose corrective actions.Analytical the dynamo fault not only need to have firm theory knowledge and experience of prolific maintenance repairs, but also need to aim at concrete fault, deepconsideration, brave creative, the dynamo after ensuring to break down removal can stabilize run - time over a long period of time.中文翻译永磁同步电动机矢量控制系统〔中文对照〕矢量控制亦称磁场定向控制,其基本思路是:通过坐标变换实现模拟直流电机的控制方法来对永磁同步电机进行控制。
永磁同步电机模糊滑模无位置传感器控制
第28卷㊀第1期2024年1月㊀电㊀机㊀与㊀控㊀制㊀学㊀报Electri c ㊀Machines ㊀and ㊀Control㊀Vol.28No.1Jan.2024㊀㊀㊀㊀㊀㊀永磁同步电机模糊滑模无位置传感器控制禹聪1,2,㊀康尔良1,2(1.哈尔滨理工大学电气与电子工程学院,黑龙江哈尔滨150080;2.黑龙江省高校直驱系统工程技术创新中心,黑龙江哈尔滨150080)摘㊀要:针对滑模量在滑模面切换以及速度非线性变化而致使的系统抖振问题,提出一种超旋转滑模模糊观测器㊂滑模观测器(SMO )存在的高频抖振会对电机控制系统产生很大的影响,导致电机产生转速波动和稳态误差㊂为了削弱SMO 的抖振问题,首先对滑模动态变量的趋近速度动态变化导致的抖振问题,通过引入模糊逻辑理论使得系统状态量趋动速度智能化,设置模糊规则以达到智能动态化速度,以系统动态变量趋向切换面的距离与状态量动态趋向速度为规则因子,动态智能化趋向速度;其次对系统变换函数导致的系统抖振,进一步采用连续函数F (s )代替不连续的sgn (s )符号函数㊂该方案有效削弱了系统的抖振问题,相较于SMO 控制提高了系统的稳定性㊂关键词:永磁同步电机;无位置传感器控制;滑模观测器;模糊控制;高频抖振;滑模控制DOI :10.15938/j.emc.2024.01.009中图分类号:TM341文献标志码:A文章编号:1007-449X(2024)01-0087-08㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀收稿日期:2022-04-11基金项目:国家科技助力经济2020(Q2020YFF0402198);黑龙江省科技攻关资助项目(GC04A517)作者简介:禹㊀聪(1997 ),男,硕士研究生,研究方向为永磁同步电机及其控制;康尔良(1967 ),男,博士,教授,硕士生导师,研究方向为电机测试与电机控制㊂通信作者:康尔良Fuzzy sliding mode position sensorless control of permanentmagnet synchronous motorYU Cong 1,2,㊀KANG Erliang 1,2(1.School of Electrical and Electronic Engineering,Harbin University of Science and Technology,Harbin 150080,China;2.Engineering Technology Innovation Center of Direct-Drive System in Colleges and Universities in Heilongjiang,Harbin 150080,China)Abstract :A super rotating sliding mode fuzzy observer was proposed to address the system chattering problem caused by the switching of sliding mode variables on the sliding mode surface and nonlinearchanges in velocity.The high-frequency chattering in sliding mode observer (SMO)can have a signifi-cant impact on the motor control system,leading to speed fluctuations and steady-state errors in the mo-tor.In order to weaken the chattering problem of SMO,aiming at the chattering problem caused by dy-namic changes in the approaching velocity of sliding mode dynamic variables,by introducing fuzzy logictheory,the trend speed of system state variables was intelligentized.Fuzzy rules were set to achieve intel-ligent dynamic speed,with the distance between the system dynamic variables towards the switching sur-face and the dynamic trend speed of the state variables as the rule factors,and dynamic intelligent trendspeed was achieved;Secondly,in response to the system chattering caused by the system switching func-tion,a continuous function F (s )was further adopted to replace the discontinuous sgn(s )symbol function.This scheme effectively weakens the chattering problem of the system and improves the stability of the sys-tem compared with SMO control.Keywords :permanent magnet synchronous motor;sensorless control;sliding mode observer;fuzzy con-trol;high frequency chattering;sliding mode control0㊀引㊀言永磁同步电机(permanent magnet synchronous motor,PMSM)由于其体积小㊁效率高等优点在工业领域得到了广泛的应用[1]㊂PMSM控制需要传感器㊁编码器等机械器件来确定转子的位置,但是目前常用的增量式编码器和霍尔传感器使得PMSM的成本增加,体积增大,同时会使得系统的稳定性降低,因此对于无传感控制的研究得到了广泛的关注[2]㊂无传感控制技术是通过检测电机绕组中的电信号来提取转子的位置信息,如定子电压和电流,通过控制算法实现电机转子速度和位置估算,常用的无传感控制方法可以分为两类,包括基于显著性跟踪的高频注入法[3]和基于机器模型的反电动势方法[4]㊂目前应用算法可投入广泛应用的有滑模观测器法[5-7]㊁模型参考自适应控制算法[8]㊁扩展卡尔曼滤波算法[9]等㊂滑模观测器(sliding mode observer,SMO)作为一种强鲁棒性的非线性观测器,以其设定电流与反馈电流为误差控制元素来设计观测器,以此可以得出PMSM转子数据以及反电动势大致数值等数据㊂作为一种典型的反电动势方法,该方式有不敏感于电机参数的优势㊂然而,滑模控制的抖振问题会降低观测器的估算精确度,导致电机产生转速波动㊂在实际应用中为了减小系统的抖振问题,通常会以开关函数和状态量趋近速度为出发点进行优化,通过采用平滑函数来代替切换函数[10-12]来削弱系统抖振㊂文献[13]设计了一种连续幂次函数Fal函数来代替传统的符号函数,有效地减小了抖振问题㊂同时有些人通过对状态量趋近速度进行控制[14-15],文献[16]采用模糊控制原理对滑模切换增益进行智能调节,从而控制状态量的趋近速度,该方式有效削弱了系统的抖振问题㊂本文采用表贴式永磁同步电机作为系统控制对象,通过分析滑模观测器抖振问题,并究其产生的原因进行研究,提出一种超螺旋滑模观测器(fuzzy su-per twisting silding mode observer,FSTSMO)㊂首先,采用F(s)函数代替传统的sgn(s)开关函数㊂其次,对滑模控制的滑模切换增益采用模糊控制方式,使其随着与滑模面距离的变化而变化㊂采取以上方式以期能够削弱系统的抖振问题㊂1㊀传统滑模观测器PMSM的两相旋转电压方程为uαuβéëêêùûúú=R+d d t L dωe(L d-L q)-ωe(L d-L q)R+d d t L qéëêêêêùûúúúúˑiαiβéëêêùûúú+eαeβéëêêùûúú㊂(1)其中eαeβéëêêùûúú=(L d-L q)ωe i d-d d t i q()+ωeψf []-sinθe cosθeéëêêùûúú㊂(2)式中:L d㊁L q为电感;ωe为电角速度;ψf为永磁磁链;θe为转子位置角;uα㊁uβ㊁iα㊁iβ为定子电压和电流;eα㊁eβ为扩展反电动势㊂对于表贴式PMSM而言有L d=L q=L s,由式(1)可知表贴式PMSM在α-β坐标系下的电流方程为dd tiαiβéëêêùûúú=-RL siαiβéëêêùûúú+1L suαuβéëêêùûúú-1L seαeβéëêêùûúú㊂(3)为了得到电机转子的转速和位置,传统SMO的设计如下:dd ti^αi^βéëêêùûúú=-R Lsi^αi^βéëêêùûúú+1Lsuαuβéëêêùûúú-1L sEαEβéëêêùûúú㊂(4)式中i^α㊁i^β为定子电流观测值㊂由式(3)㊁式(4)可得电流误差方程为i~αi~βéëêêùûúú=-R Lsi~αi~βéëêêùûúú+1Lseα-Eαeβ-Eβéëêêùûúú㊂(5)式中i~α㊁i~β为电流观测误差㊂因此滑模面函数S和滑模控制律Eα可定义为:S T=i~αi~β[]=0;(6)EαEβéëêêùûúú=ksgn(i~α)sgn(i~β)éëêêùûúú㊂(7)当满足S T S㊃<0时,SMO进入滑动模态,当状态量到达滑模面时,i~=i~㊃=0,此时E趋近于e㊂根据滑模控制的等效原理可得k sgn(i^α-iα)k sgn(i^β-iβ)[]T=eαeβéëêêùûúú=EαEβéëêêùûúú㊂(8)由式(8)可知,估计得到的反电动势值有高频的切换信号,在转子位置估算时采用反正切函数代入运算会产生抖振现象㊂88电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀2㊀超螺旋滑模观测器对于一个动态系统中的控制器来说,通过设置控制器输入,并通过数据反馈调节使得系统控制状态量在有限的时间内收敛到0㊂本节提出一个动态观测器,该观测器采用超螺旋控制(super-twisting control,STC)算法,根据式(4)可得d d t i ^αi ^βéëêêùûúú=-R L s i ^αi ^βéëêêùûúú+1L s u αu βéëêêùûúú+γαγβéëêêùûúú㊂(9)由式(3)和式(9)作差得d d t i ~αi ~βéëêêùûúú=-R L s i ~αi ~βéëêêùûúú-1L s e αe βéëêêùûúú-γαγβéëêêùûúú㊂(10)以γα㊁γβ作为控制系统的控制器输入,基于STC 算法,结合SMC 控制理论设计超螺旋滑模观测器(super twisting sliding mode control,STSMO)输入如下:γαγβéëêêùûúú=A sgn(i ~α)sgn(i ~β)éëêêùûúú+B i ~αi ~βéëêêùûúú+ʏc j1c j2d t ʏc j3c j4d t éëêêêùûúúú㊂(11)滑模面函数s 定义为s =i ~αi ~β[]T㊂(12)式中:A =χ1|i ~α|12χ2|i ~β|éëêêùûúú;B =χ200χ6éëêêùûúú;c j1=χ3χ4[];c j2=sgn(i ~α)00i ~αéëêêùûúú;c j3=χ7χ8[];c j4=sgn(i ~β)00i ~βéëêêùûúú㊂由上式可知滑模面函数为:i ~α=-χ1|i ~α|12sgn(i ~α)-χ2i ~α+η1+μ1;(13)i ~β=-χ5|i ~β|12sgn(i ~β)-χ6i ~β+η2+μ2㊂(14)式中:㊀㊀μ1=-R L s i ~α-1L s e α;(15)㊀㊀μ2=-R L s i ~β-1L s e β;(16)㊀㊀η1=-χ3sgn(i ~α)-χ4i ~α;(17)㊀㊀η2=-χ7sgn(i ~β)-χ8i ~β㊂(18)对于大于0的常数σ1㊁σ2,使其满足条件:|μ1|ɤσ1|i ~α|12;|μ2|ɤσ2|i ~β|12㊂}(19)当状态量到达滑模面时有s =s ㊃=0,与此同时i ~㊃α=i ~㊃β=0,因此根据等效原理可得γαγβéëêêùûúú=e αe βéëêêùûúú=ωe ψf -sin θe cos θe éëêêùûúú㊂(20)由于滑模观测器中的不连续开关函数会导致系统的抖振问题,因此本文采用连续函数F (s )作为系统的切换函数,其表达式为F (s )=s|s |+ζ㊂(21)为了得到准确的电动势(electromotive force,EMF)估计值,需要对控制量进行滤波处理,滤波截止频率会引发较大的相位延迟,所以在运用过程中需要对位置角的相位进行补偿,故转子位置估计值为θ^e =-arctan(e αe β)+|e β|-e β2|e β|π㊂(22)由式(21)可得电机的转速估计信息为ω^e =e 2α+e 2βψf㊂(23)由Lyapunov 定理可知系统满足ss ㊃<0,系统渐进稳定,即系统状态量具有较短时间稳定优势㊂3㊀模糊控制器为了削弱滑模控制存在的抖振问题,将模糊控制理论引入滑模控制中,采用滑模面作为模糊控制的输入,模糊逻辑设计时对于被控对象的模型并无特别要求,但对专家经验非常依赖,其控制原理是将专家经验融入控制系统来设计模糊规则,随着状态量与滑模面距离的变化对滑模增益进行有效估计㊂定义模糊控制输入量的模糊语言为:负高(NH)㊁负中(NM)㊁负低(NL)㊁零(ZO)㊁正低(PL)㊁正中(PM)㊁正高(PH)㊂定义模糊输出的语言为:负高(NH)㊁负中(NM)㊁负低(NL)㊁零(ZO)㊁正低(PL)㊁正中(PM)㊁正高(PH),设计模糊控制规则表如表1所示㊂由表可知,模糊逻辑理论设计为7个模糊子集并对应7个数据输出,模糊逻辑采用Mam-dani 为其核心算法以及采用重心反模糊化得出可识别输出量㊂其控制逻辑如图1~图3所示㊂98第1期禹㊀聪等:永磁同步电机模糊滑模无位置传感器控制表1㊀控制规则Table 1㊀Control rules s㊃NH NM NL ZO PL PM PH NHPH PH PM PM PM PL ZO NM PHPHPM PM PL PLZO NL PM PM PL PLPLZO ZOZO PM PL PLZO NL NLNM PLZO ZO NL NLNLNM NM PM ZO NL NLNM NM NH NH PHZO NLNM NMNM NHNH 图1㊀输入s 的隶属函数Fig.1㊀Membership function of inputs图2㊀输入s ㊃的隶属函数Fig.2㊀Membership function of input s㊃图3㊀输出P (s )的隶属函数Fig.3㊀Membership function of output P (s )设计控制规则,使得STSMO 系统切换增益随着状态量与切换面的距离自整定㊂系统状态量距离滑模切换面较远时,滑模增益值较大,同时状态量趋近速度很快;当系统状态量与滑模切换面较近时,滑模增益值较小,状态量趋近速度较小,从而削弱系统的抖振㊂4㊀仿真和实验验证搭建Simulink 模型以及搭建平台试验,验证本文所提控制策略,PMSM 参数如表2所示㊂表2㊀PMSM 参数Table 2㊀PMSM parameters㊀㊀参数数值额定功率P /kW 2.6额定电压U /V 220定子电阻R s /Ω0.73磁极对数p 4d /q 轴电感L s /H 0.00245黏滞阻尼F /(N㊃ms)0.005转动惯量J /(kg㊃m 2)0.00194永磁体磁链ψf /Wb0.175PMSM 控制系统框图如图4所示㊂图4㊀PMSM 控制系统框图Fig.4㊀Block diagram of PMSMFSTSMO㊁STSMO 以及SMO 系统仿真波形如图5~图12所示㊂由图5㊁图6可知,STSMO 控制相较于SMO 控制提高了系统转子位置估计精确度㊂由图7可知,将模糊控制理论引入STSMO 中,FSTSMO 相比于STSMO 转子位置估计更精确,系统控制性能更好㊂设置仿真时间为0.1s,给定阶跃转速指令为1000r /min,开关频率为10kHz,突加突减负载为5N㊃m,系统转子速度估计值与实际值仿真波形如9电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀图8~图10所示㊂图5㊀SMO 转子位置估计值与实际值Fig.5㊀Rotor position estimated value and actual valueofSMO图6㊀STSMO 转子位置估计值与实际值Fig.6㊀Rotor position estimated value and actual valueofSTSMO图7㊀FSTSMO 转子位置估计值与实际值Fig.7㊀Rotor position estimated value and actual valueofFSTSMO图8㊀SMO 控制突加突减负载时转子速度波形Fig.8㊀Rotor speed waveform when the SMO controlsuddenly adds and reduces the load由图8㊁图9可知,给定转速为1000r /min,SMO 控制存在较大的转速超调量,STSMO 控制相较于SMO 控制转速超调量较小,同时转速估计更加准确㊂如图10所示,将模糊控制理论引入STSMO 中,可知FSTSMO 相较于STSMO 系统的转速超调量更小,削弱了系统的抖振,实现了更精确的转速估计㊂图9㊀STSMO 控制突加突减负载时转子速度波形Fig.9㊀Rotor speed waveform when the STSMO con-trol suddenly adds and reduces theload图10㊀FSTSMO 控制突加突减负载时转子速度波形Fig.10㊀Rotor speed waveform when the FSTSMOcontrol suddenly adds and reduces the load突加突减负载设置为5N㊃m,如图8㊁图9所示,STSMO 相较于SMO 转速脉动大大减小㊂由图10可知,FSTSMO 相较于STSMO 控制系统的转速脉动更小,控制系统更稳定㊂图11㊀FSTSMO 转子速度估计值与实际值差值Fig.11㊀Difference between FSTSMO rotor speed es-timated value and actual value由图8㊁图9可知,SMO 转子转速估计的波动较大,转速误差在-10~10r /min 之间,STSMO 转速估计误差在-0.95~-0.45r /min 之间㊂由图10㊁图11可知,FSTSMO 转子转速估计值与实际转速的差值在-0.085~-0.065r /min 之间㊂图12为FSTSMO 控制反电动势波形,由波形可知,E α与E β19第1期禹㊀聪等:永磁同步电机模糊滑模无位置传感器控制相差90ʎ相位㊂图12㊀FSTSMO 控制反电动势E α,E β波形Fig.12㊀Waveform of back EMF E α,E βof FSTSMOcontrol由文献[8]可知,传统模型参考自适应转速估计误差在8.1~10.6r /min 采用改进滑模-模型参考自适应方式时,转速误差估计在3.9~4.6r /min㊂由文献[9]可知,扩展卡尔曼滤波转速估计误差值也远大于FSTSMO 控制系统,可知所提出的FSTSMO 控制转速估计更加精准,系统响应更稳定㊂系统搭建控制试验平台如图13所示㊂图13㊀试验平台Fig.13㊀Test platformSMO㊁STSMO 和FSTSMO 的控制速度实验波形如图14~图16所示㊂由图14可知,系统给定转速为1000r /min,SMO 控制存在较大的抖振,会影响系统的运行性能㊂图14㊀SMO 控制速度实验波形Fig.14㊀Waveform of speed experiment of SMOcontrol图15㊀STSMO 控制速度实验波形Fig.15㊀Waveform of speed experiment ofSTSMOcontrol图16㊀FSTSMO 控制速度实验波形Fig.16㊀Waveform of speed experiment ofFSTSMO control由图14~图16可知,STSMO 相较于SMO 控制大大削弱了系统抖振,提高了系统稳定性,FSTSMO 控制相较于STSMO 控制系统抖振更小,系统稳定性更强,以突加突减负载为突加状况时,系统有较短稳定时间优势㊂SMO㊁STSMO 和FSTSMO 控制系统反电动势波形如图17~图19所示㊂由图17㊁图18可知,STSMO 相较于SMO 控制其反电动势估计波形较平滑,提高了系统控制精确度,E α与E β相差90ʎ相位,进一步证明了反电动势估计的正确性㊂由图19可知,将模糊控制理论引入STSMO 控制中,FSTSMO 相较于STSMO 系统控制精确度更高,系统更稳定㊂图17㊀SMO 控制系统反电动势E α,E β波形Fig.17㊀Back EMF E α,E βwaveform of SMO controlsystem29电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀图18㊀STSMO 控制系统反电动势E α,E β波形Fig.18㊀Back EMF E α,E βwaveform of STSMOcontrolsystem图19㊀FSTSMO 控制系统反电动势E α,E β波形Fig.19㊀Back EMF E α,E βwaveform of FSTSMOcontrol system本文提出的FSTSMO 控制相较于SMO 控制有效地削弱了系统的抖振,降低了转速波动,提高了转子位置估计精确度,能够以更短的时间达到系统稳定,提高系统稳定性㊂5㊀结㊀论本文提出了一种FSTSMO 控制方案,将模糊逻辑理论引入STSMO 控制中,设置模糊规则是以系统动态量趋近动态面的距离与趋动速度动态化为规则元素,以此来动态智能化状态量趋动速度,使得状态量趋近速度随着与滑模面的距离动态变化,同时进一步采用了连续函数F (s )代替不连续的sgn(s )符号函数,进一步提高了系统的稳定优势㊂通过仿真和实验表明,FSTSMO 控制大大提高了系统的稳定性,由仿真数据可知,FSTSMO 系统转子位置估计误差为5ˑ10-5rad 左右,转速估计误差在-0.085~-0.065r /min 之间,相较于SMO 控制有更好的抖振控制优势,其得出的转子位置数据精确度和系统稳定性具有更好展现㊂参考文献:[1]㊀WANG B,WANG Y,FENG L,et al.Permanent magnet synchro-nous motor sensorless control using proportional-integral linear ob-server with virtual variables:a comparative study with a sliding mode observer[J].Energies,2019,12(5):1.[2]㊀REN N,FAN L,ZHANG Z.Sensorless PMSM control with slid-ing mode observer based on sigmoid function[J].Journal of Elec-trical Engineering &Technology,2021,16(2):933.[3]㊀LIN T C,ZHU Z Q.Sensorless operation capability of surface-mounted permanent-magnet machine based on high-frequency sig-nal injection methods[J].IEEE Transactions on Industry Applica-tions,2015,51(3):2161.[4]㊀ZHAO L,HUANG J,LIU H,et al.Second-order sliding-modeobserver with online parameter identification for sensorless induc-tion motor drives[J].IEEE Transactions on Industrial Electron-ics,2014,61(10):5280.[5]㊀王春风,赵青青,孟旭,等.直流电机的非奇异快速Terminal 滑模位置控制[J].哈尔滨理工大学学报,2019,24(4):36.WANG Chunfeng,ZHAO Qingqing,MENG Xu,et al.Non singu-lar fast Terminal sliding mode position control for DC motors [J].Journal of Harbin University of Science and Technology,2019,24(4):36.[6]㊀LIANG D,LI J,QU R,et al.Adaptive second-order sliding-modeobserver for PMSM sensorless control considering VSI nonlinearity [J ].IEEE Transactions on Power Electronics,2017,33(10):8994.[7]㊀GONG C,HU Y,GAO J,et al.An improved delay-suppressedsliding mode observer for sensorless vector-controlled PMSM[J].IEEE Transactions on Industrial Electronics,2020,67(7):5913.[8]㊀王栋.基于滑模-模型参考自适应的异步电动机无速度传感器控制[D].西安:西安理工大学,2021.[9]㊀程梁,王维强,余天赐,等.基于无迹卡尔曼滤波的永磁无刷电机转矩脉动抑制研究[J].大电机技术,2021(6):10.CHENG Liang,WANG Weiqiang,YU Tianci,et al.Research on torque ripple suppression of permanent magnet brushless motor based on unscented Kalman filtering [J].Large Electric Machine and Hydraulic Turbine,2021(6):10.[10]㊀祝新阳,曾国辉,黄勃,等.改进滑模观测器的永磁同步电机矢量控制[J].信息与控制,2020,49(6):708.ZHU Xinyang,ZENG Guohui,HUANG Bo,et al.Improving the vector control of a permanent magnet synchronous motor for a slid-ing-mode observer [J ].Information and Control,2020,49(6):708.[11]㊀张伯泽,宗剑,钱平.基于滑模变结构控制的内置式永磁同步电机无位置传感器矢量控制[J].电机与控制应用,2020,47(8):36.ZHANG Boze,ZONG Jian,QIAN Ping.Built in permanent mag-net synchronous motor based on sliding mode variable structure39第1期禹㊀聪等:永磁同步电机模糊滑模无位置传感器控制control[J].Electric Machines and Control Application,2020,47(8):36.[12]㊀张紫君,熊官送,曹东海.基于无位置传感器的永磁同步电机控制技术研究[J].导航定位与授时,2020,7(4):102.ZHANG Zijun,XIONG Guansong,CAO Donghai.Research onthe control technology of permanent magnet synchronous motorbased on no position sensor[J].Navigation,Positioning andTiming,2020,7(4):102.[13]㊀张文宾,缪仲翠,余现飞,等.基于改进型滑模观测器的永磁同步电机分数阶微积分滑模控制[J].电机与控制应用,2018,45(7):8.ZHANG Wenbin,MIAO Zhongcui,YU Xianfei,et al.Fractionalcalculus sliding mode control of a permanent magnet synchronousmotor based on a modified sliding mode observer[J].ElectricMachines and Control Application,2018,45(7):8. [14]㊀陶彩霞,赵凯旋,牛青.考虑滑模抖振的永磁同步电机模糊超螺旋滑模观测器[J].电力系统保护与控制,2019,47(23):11.TAO Caixia,ZHAO Kaixuan,NIU Qing.A fuzzy super helicalsliding mode observer of a permanent magnet synchronous motorconsidering sliding mode shaking[J].Power System Protectionand Control,2019,47(23):11.[15]㊀赵凯旋.基于滑模变结构永磁同步电机控制系统抖振问题的研究[D].兰州:兰州交通大学,2019.[16]㊀徐冬磊.基于滑模变结构的永磁同步电机矢量控制研究[D].武汉:华中科技大学,2019.(编辑:邱赫男)49电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀。
基于改进型ADRC_的PMSM_无位置传感器控制
第28卷㊀第4期2024年4月㊀电㊀机㊀与㊀控㊀制㊀学㊀报Electri c ㊀Machines ㊀and ㊀Control㊀Vol.28No.4Apr.2024㊀㊀㊀㊀㊀㊀基于改进型ADRC 的PMSM 无位置传感器控制肖芳,㊀谢元宇,㊀王明辉,㊀林海波(哈尔滨理工大学大型电机电气与传热技术国家地方联合工程中心,黑龙江哈尔滨150080)摘㊀要:针对永磁同步电机矢量控制中传统速度环PI 调节器存在协调性与抗干扰能力差的问题,引入自抗扰控制(ADRC )替代速度环PI 调节器,考虑到线性ADRC 中PD 控制存在性能较差的缺点,选择滑模控制代替PD 控制,提出改进型ADRC 控制方法;同时采用模型参考自适应系统(MRAS )取代传统机械式位置传感器对电机的转子位置进行检测,实现了永磁同步电机无位置传感器控制㊂通过采用改进型ADRC 与MRAS 相结合的控制方法,电机控制系统具有了更好的动态响应速度和抗干扰能力㊂最后通过永磁同步电机对拖实验平台对电机在转速恒定负载突变和带载启动转速突变两种工况下进行实验,验证了改进型ADRC 与MRAS 相结合的控制策略的有效性㊂关键词:永磁同步电机;无位置传感器;滑模控制;自抗扰控制;模型参考自适应DOI :10.15938/j.emc.2024.04.006中图分类号:TM343文献标志码:A文章编号:1007-449X(2024)04-0050-11㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀收稿日期:2023-08-28基金项目:黑龙江省高校基本科研业务费(2021-KYYWF -0746)作者简介:肖㊀芳(1982 ),女,博士,副教授,研究方向为电机系统的电磁兼容㊁电机驱动控制及其系统;谢元宇(1997 ),男,硕士研究生,研究方向为电机驱动控制;王明辉(1991 ),男,博士,讲师,研究方向为永磁同步电机无位置传感器技术㊁高性能交流电机控制技术;林海波(1997 ),男,硕士研究生,研究方向为电机驱动控制㊂通信作者:肖㊀芳Improved ADRC-based PMSM control without position sensorXIAO Fang,㊀XIE Yuanyu,㊀WANG Minghui,㊀LIN Haibo(National Engineering Research Center of Large Electric Machines and Heat Transfer Technology,HarbinUniversity of Science and Technology,Harbin 150080,China)Abstract :Aiming at the problem of poor coordination and anti-interference ability of the traditional speed loop PI regulator in vector control of permanent magnet synchronous motor,active disturbance rejection control (ADRC)was introduced to replace the speed loop PI regulator,and the sliding mode control wasselected to replace the PD control,considering the disadvantage of poor performance of PD control in the linear ADRC;at the same time,model reference adaptive system (MRAS)was adopted to replace the traditional mechanical position sensor to detect the rotor position of the motor,and the improved ADRC control method was proposed.Considering the poor performance of PD control in linear ADRC,the slid-ing mode control was chosen to replace the PD control,and the improved ADRC control method was pro-posed;meanwhile,the model reference adaptive system (MRAS)was adopted to replace the traditional mechanical position sensor to detect the rotor position of the motor,so as to realize the sensorless control of the permanent magnet synchronous motor.By using the improved ADRC combined with MRAS,the motor control system has better dynamic response speed and anti-interference ability.Finally,through the permanent magnet synchronous motor pair-drag experimental platform,the motor was experimented underthe two working conditions of constant speed and sudden load change and loaded startup speed change,which verifies effectiveness of the control strategy combining the improved ADRC and MRAS.Keywords:permanent magnet synchronous motor;sensorless;sliding mode control;active disturbance rejection control;model reference adaptive system0㊀引㊀言永磁同步电机(permanent magnet synchronous motor,PMSM)具有功率因数高㊁可靠性高㊁体积小等优点[1-2]㊂PMSM作为非线性㊁强耦合系统,其内外扰动都会影响电机的控制性能,因此要求控制策略应具有良好的抗干扰能力㊂目前电机控制系统常采用传统PI调速方式,传统PI控制具有结构简单㊁稳定性好等优点,但在动态响应速度与超调量之间存在矛盾,难以满足较宽速度范围的要求[3-5]㊂因此为了使系统具有更好的控制效果,各种非线性控制方法也被应用到电机控制中,自抗扰控制(active disturbance rejection control,ADRC)技术就是其中之一,其具有更强的鲁棒性,利用反馈控制技术对系统内外扰动进行实时估计和补偿,实现对电机输出的精确控制[6-8]㊂文献[9]对ADRC控制器中的非线性状态反馈控制律进行改进,提出一种基于滑模理论的改进非线性状态误差反馈控制律,但是ADRC控制器中的非线性函数参数较多,难以调节,对ADRC控制器的参数整定,主要依赖于经验㊂高志强教授通过对ADRC进行线性化处理,将ADRC中的非线性函数转换为线性函数[10],提出了线性ADRC(linear AD-RC,LADRC),其可以有效提高控制精度和抗干扰能力,解决传统ADRC参数整定困难的问题[11]㊂文献[12]将速度误差观测器的比例增益通道加入线性状态扩展观测器,以提高对扰动的观测速度,同时,在控制系统中加入在线识别算法,获取实时转动惯量信息,并对ADRC输出增益进行不断校正,从而提高其控制精度和抗干扰性能㊂文献[13]通过利用神经网络非线性函数的拟合能力识别PMSM 的干扰函数,用训练好的神经网络对状态扩张观测器的干扰量进行补偿,从而提高ADRC的观测精度和抗干扰能力㊂在PMSM矢量控制中,机械传感器可以为PMSM控制提供必要的转子位置,但机械传感器的使用不仅增加了生产成本,还受到电机运行条件和外部环境的影响,因此无位置传感器控制技术逐渐成为高性能PMSM控制的核心技术[14]㊂为此,本文首先分析了ADRC结构,根据PMSM的机械运动方程设计速度环一阶线性ADRC控制器,针对线性状态误差反馈控制律(linear state error feedback,LSEF)中PD控制抗干扰能力弱㊁控制性能差的问题,提出了改进型ADRC控制策略,最后,将改进型ADRC与模型参考自适应系统(model reference a-daptive system,MRAS)无位置传感器控制策略相结合㊂实验结果验证了所提控制方法的可行性与有效性㊂1㊀PMSM的数学模型为了研究方便,作出以下假设[15]:1)三相定子绕组空间对称,气隙空间中的磁动势以正弦方式分布;2)忽略磁饱和效应及各种损耗;3)忽略电机运行中温度和频率的影响㊂根据以上假设条件,PMSM在同步旋转坐标系下的电压方程为:u d=R s i d+dψdd t-ωeψq;u q=R s i q+dψqd t+ωeψd㊂üþýïïïï(1)定子磁链方程为:ψd=L d i d+ψf;ψq=L q i q㊂}(2)电磁转矩方程为T e=32p n i q[i d(L d-L q)+ψf]㊂(3)式中:u d㊁u q为定子电压在d㊁q轴的分量;i d㊁i q分别是定子电流在d㊁q轴的分量;ψd㊁ψq为定子磁链在d㊁q轴的分量;ωe为转子电角速度;L d㊁L q分别是d㊁q轴电感分量;ψf是永磁体磁链;p n为PMSM的极对数㊂本文选用表贴式PMSM电机,其L d=L q,则转矩方程可简化为T e=32p n i qψf㊂(4)机械运动方程为Jdωmd t=T e-T L-Bωm㊂(5)式中:ωm为电机的机械角速度;J为转动惯量;B为15第4期肖㊀芳等:基于改进型ADRC的PMSM无位置传感器控制阻尼系数;T L 为负载转矩㊂2㊀改进型ADRC 的数学模型2.1㊀自抗扰控制器分析与设计ADRC 由跟踪微分器(tracking differentiator,TD)㊁扩张状态观测器(extend state observer,ESO)和非线性状态误差反馈控制律(nonlinear state error feedback,NLSEF)三部分组成㊂根据上节电机机械运动方程与转矩方程可得电机转速输出状态方程为d ωm d t =3p n ψf i q 2J -T L J -BωmJ㊂(6)将i q 设为控制输出量u ,则控制输出量增益b =(3p n ψf )/(2J );由式(6)可知,PMSM 转速受到转动惯量J ㊁负载转矩T L 和摩擦系数B 的影响,将系统内外总扰动量用上述三者的变化量a (t )表示㊂因此PMSM 控制中转速环状态方程可改写为d ωmd t=bu +a (t )㊂(7)ADRC 控制器的设计由被控对象阶数决定,式(7)为一阶方程,因此使用一阶ADRC 控制器,下面给出一阶线性ADRC 控制器的离散算法㊂线性微分跟踪器(linear tracking differentiator,LTD)表达式为:e 1=v 1-v ∗;v ㊃1=-re 1㊂}(8)线性扩张观测器(linear extend state observer,LESO)表达式为:e 2=z 1-v ;z ㊃1=-β1e 2+z 2+bu ;z ㊃2=-β2e 2㊂üþýïïïï(9)线性反馈控制律(LSEF)表达式为:e 3=v 1-z 1;u 0=k p e 3+k d e 3;u =(u 0-z 2)/b ㊂üþýïïï(10)式中:v ∗为电机给定转速;v 1为给定转速v ∗经LTD平滑过渡后的输出转速;v 为电机的实际转速;z 1为LSEO 输出的状态变量v 的观测值;z 2为系统所受的内外扰动之和的估计值;β1㊁β2为构建LESO 结构中的各阶误差增益;k p ㊁k d 为线性状态误差反馈控制率增益㊂一阶线性ADRC 控制器结构如图1所示㊂图1㊀一阶线性ADRC 控制器Fig.1㊀First-order linear active disturbance rejectioncontroller2.2㊀滑模变结构分析与设计滑模变结构也称为滑模控制(sliding mode con-trol,SMC),其主要包括滑模切换面s (x )的选择与滑模趋近律u (x )的设计[16]㊂对于非线性系统,状态空间定义为:x ㊃=f (x ),x ɪR n ;状态空间中有一超平面:s (x )=s (x 1,x 2, ,x n )=0,其将状态空间分为s >0和s <0两部分,在切换面上的点被分为3种类型:通常点C ㊁起始点B 和终止点A [17],如图2所示㊂图2㊀滑模切换面上3种运动情况Fig.2㊀Three motion conditions on sliding mode switchingsurface当两侧点均趋近于切换面时,系统的运动轨迹就会沿着s (x )=0运动至理想状态,由此可得系统运动在滑模动态区满足的条件表达式为lim s ң0s ㊃s ɤ0㊂(11)确定滑模面函数为s =s (x ),s ɪR n ㊂(12)构建合适的滑动模态控制律:u =u +(x ),s (x )>0;u-(x ),s (x )<0㊂{(13)式中u +(x )ʂu -(x )㊂由上述可知,滑模运动的实现需要满足以下3个基本条件:1)滑动模态存在;2)满足可达性条件lim s ң0s ㊃s ɤ0;25电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀3)满足稳定性要求㊂2.3㊀改进型ADRC 控制器设计为了提高系统的控制性能,设计一种基于滑模控制的线性状态误差反馈控制系统(sliding mode control-linear state error feedback,SMC-LSEF)㊂由式(10)可知速度环ADRC 控制器中SMC-LSEF 的状态误差e 3为v 1与z 1之差,定义控制输出量为u 0=e ㊃3,则有:e 3=v 1-z 1;u 0=e ㊃3;u =(u 0-z 2)/b ㊂üþýïïïï(14)为方便分析,令x 1=e 3,对式(14)求导可得:x ㊃1=v ㊃1-z ㊃1;x ㊃2=x ㊃㊃1㊂}(15)根据上节滑模变结构理论,首先构建滑模面,考虑抑制高频噪声和消除稳态误差,引入积分量,因此定义积分滑模面为s =x 1+c ʏx 1d t ㊂(16)式中c 为滑模面参数,且c >0㊂为了减少运动点在滑模动态区内的运动时间,保证过渡过程顺畅[18],选择指数趋近律作为系统的滑动模态趋近律,其形式为s ㊃=-ξsign(s )-ks ㊂(17)式中:ξ>0㊁k >0且均可调;sign(㊃)为符号函数㊂根据Lyapunov 稳定性原理,判断滑模控制系统的稳定性㊂根据选取的滑模面,取李雅普诺夫函数为V =12s 2㊂(18)对式(18)进行求导,并结合式(17)可得V ㊃=ss ㊃=s (-ξsign(s )-ks )=-ξ|s |-ks 2㊂(19)由于ξ和k 均大于0,故ss ㊃<0恒成立,根据李雅普诺夫稳定判据得,系统状态运动点到达滑模动态区后,滑模运动趋于稳定㊂由此可得SMC-LSEF 在线性ADRC 中的表达式为:e 3=v 1-z 1;u 0=-ξsign(s )-ks -ce 3;u =(u 0-z 2)/b ㊂üþýïïï(20)通过以上分析可得到改进型滑模ADRC 控制的原理框图如图3所示㊂图3㊀改进型滑模ADRC 控制原理框图Fig.3㊀Block diagram of improved sliding mode ADRCcontrol principle3㊀MRAS 无位置传感器算法传统MRAS 转速辨识以电机参数不变为条件,但在运行电机温度升高时,电机定子电阻将增加;同时,电机磁饱和效应使电机电感值随电流变化,因此传统MRAS 性能较差㊂针对此问题,设计一种基于改进型MRAS 转速观测器,其可在线识别定子电阻和电感,提高速度识别精度㊂3.1㊀参考模型与可调模型的设计PMSM 模型可视为二维方程,当识别参数多于方程维数时,会出现欠秩问题,导致识别结果不收敛㊂针对该问题提出一种固定部分参数并对其他参数分步进行辨识的方法㊂表贴式PMSM 在同步旋转坐标系下的电流方程为:d i d d t =-R s L s i d +ωe i q +ud L s;d i q d t =-R s L s i q -ωe i d +u q L s -ωe ψf L s ㊂üþýïïïï(21)将式(21)改写为状态空间表达式为dd ti =Xi +Y u +Z w ㊂(22)式中:i =i d i q éëêêùûúú;X =-R sL s ωe ωe -R s L s éëêêêêùûúúúú;Y =1L s ;u =u d u q éëêêùûúú;Z =-ψf L s ;w =0ωe[]㊂由式(22)可以引入MRAS 中待估转速ωe 和定子电阻R s ,可得到可调模型状态空间表达式为d d ti ^=X ^1i ^+Y u +Z w ^㊂(23)式中:i ^=i ^d i ^q éëêêùûúú;X ^1=-R ^s L s ω^e ω^e -R^s L s éëêêêêêùûúúúúú;w ^=0ω^e éëêêùûúú;其35第4期肖㊀芳等:基于改进型ADRC 的PMSM 无位置传感器控制余参数同上式一致㊂当ωe ㊁R s 稳定后,将R s 作为固定值进行ωe ㊁L s 的辨识,此时可调模型的状态空间表达式为d d ti ^=X ^2i ^+Y ^u +Z ^w ^㊂(24)式中:X ^2=-R s L ^s ω^e ω^e -R s L ^s éëêêêêêùûúúúúú;Y ^=1L ^s ;Z ^=-ψf L ^s ;w ^=0ω^e éëêêùûúú㊂3.2㊀自适应律的确定定义状态误差e =i -i ^,将参考模型式(22)与可调模型式(23)作差可得d d te d e q éëêêùûúú=(Xi -X ^1i ^)+(Z w -Z w ^)=Xe +(X -X ^1)i ^+Z (w -w ^)㊂(25)式中:e d =i d -i ^d ;e q =i q -i ^q ;X -X ^1=-R s L s+R ^sL s ωe -ω^e ω^e -ωe -R s L s +R^s L s éëêêêêêùûúúúúú;w -w ^=0ωe -ω^e éëêêùûúú㊂根据的Popov 超稳定性理论,令W =-[(X -X ^1)i ^+Z (w -w ^)],则式(25)变为dd te =Xe -W ㊂(26)令V =e ,将式(26)代入Popov 积分不等式,可得η(0,t 1)=ʏt 10V TW d t =-ʏt 10e T -R s L s +R ^s L s ωe -ω^e ω^e -ωe -R s L s +R ^s L s éëêêêêêùûúúúúúi ^d i ^q éëêêùûúú-ψf L s 0ωe -ω^e éëêêùûúúìîíïïïïïüþýïïïïïd t ȡ-γ20㊂(27)为辨识ωe 和R s ,将式(27)拆分为两式:η1(0,t 1)=-ʏt 1e T(-R s L s +R ^s L s )i ^d i ^q éëêêùûúúd t ȡ-γ21;η2(0,t 1)=-ʏt 10e T (ωe -ω^e )i ^q -i ^d -ψf L s éëêêêùûúúúd t ȡ-γ22㊂üþýïïïïïïïï(28)因此,当上式中η1(0,t 1)ȡ-γ21,η2(0,t 1)ȡ-γ22成立时,系统即稳定㊂最后求解参数的自适应律为:ω^e =(K p +K i s )[i d i ^q -i q i ^d -ψf L s(i q -i ^q )];R ^s =-L s (K p +K i s)[(i d -i ^d )i ^d +(i q -i ^q )i ^q ]㊂üþýïïïï(29)当R s 值收敛到稳定值后,将电阻值视为固定值,对ωe 和L s 进行辨识㊂将参考模型式(22)与可调模型式(24)作差可得d d te d e q éëêêùûúú=(Xi -X ^2i ^)+(Y u -Y ^u )+(Z w -Z ^w ^)=Xe +(X -X ^2)i ^+(Y -Y ^)u +Z (w -w ^)+(Z -Z ^)w ^㊂(30)式中:Y -Y ^=1L s -1L ^s ;Z -Z ^=-ψf L s +-ψf L ^s;其余参数同上式一致㊂构造Popov 积分不等式,令W =-[(X -X ^2)i ^+(Y -Y ^)u +Z (w -w ^)+(Z -Z ^)w ^],得到速度和电感的自适应律,其中ωe 的表达式在各条件下与式(29)相同,电感的自适应律为1L ^s=-(K p +K i s )[(i d -i ^d )(R s i ^d -u d )+(i q -i ^q )(R s i ^q -u q +ω^e ψf )]㊂(31)根据上述分析,得到基于改进MRAS 的PMSM 转速辨识框图,如图4所示㊂图4㊀基于改进MRAS 的PMSM 转速辨识结构图Fig.4㊀Structure diagram of PMSM speed identificationbased on improved MRAS45电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀根据图4,分步辨识法的步骤可归纳如下:1)固定L s 值,通过式(29)辨识电机转速和定子电阻;2)电机R s 值在运行过程中不断变化;3)当R s 趋于稳定后,代入式(31)计算L s 的值,最终实现对R s ㊁L s 和ωe 的辨识㊂4㊀仿真分析以三相PMSM 为控制对象,在仿真平台上基于MRAS 的转速环PI 控制和转速环滑模ADRC 控制的仿真模型㊂基于滑模ADRC 的PMSM 无位置传感器控制系统总体控制框图如图5所示㊂图5㊀基于滑模ADRC 的PMSM 无位置传感器控制框图Fig.5㊀Sensorless control block diagram of PMSM based on sliding mode ADRC在速度环采用传统PI 调节器下将改进MRAS 与传统MRAS 在相同工况下进行仿真,通过对比图6与图7可知,改进MRAS 转速误差更小,动态性能更佳㊂由图8可知,电机在1000r /min 转速稳定运行时,改进MRAS 转子位置估计值和实际值的偏差更小㊂通过图4转速辨识结构图进行仿真模型搭建,验证MRAS 参数辨识策略,波形如图9所示㊂电阻初值为0.72Ω,电感初值为3.93mH㊂图9(a)电阻在35ms 处开始收敛至0.73Ω,误差约为1.4%;图9(b)电感辨识在0.1s 处开始收敛至3.86mH 左右,误差约为1.8%㊂因此,可以证明MRAS 参数辨识测策略的有效性㊂下面对所提出的PMSM 无位置传感器控制策略进行仿真,将电机转速设置为600r /min,在0.2s 时将负载增加到2N㊃m,由图10和图11可知:基于PI 控制下电机在启动阶段会产生超调,50ms 后达到给定转速;基于滑模ADRC 控制下转速波形大约30ms 达到给定转速,与PI 控制相比,滑模ADRC 控制的响应速度更快,超调量几乎为0,当负载突变时,滑模ADRC 控制的波动幅度更小,并且在短时间内趋于稳定㊂图6㊀传统MRAS 控制下动态转速波形Fig.6㊀Dynamic speed waveform under traditional MRAScontrol图7㊀改进MRAS 控制下动态转速波形Fig.7㊀Dynamic speed waveform under improved MRAS control55第4期肖㊀芳等:基于改进型ADRC 的PMSM 无位置传感器控制图8㊀传统MRAS 与改进MRAS 转子位置对比波形Fig.8㊀Comparison of rotor position waveforms be-tween traditional MRAS and improvedMRAS图9㊀电机参数辨识波形Fig.9㊀Motor parameter identification waveform图12为2种控制器下转子位置误差对比波形,图12可知,电机稳态运行期间,2种控制器下的转子位置估计值与实际值之间误差均较小,因此实际转子位置和估计转子位置在2个速度控制器下都可实现较好跟踪㊂图10㊀PI 控制的转速波形Fig.10㊀Speed waveform of PIcontrol图11㊀滑模ADRC 控制的转速波形Fig.11㊀Speed waveform of sliding mode ADRCcontrol图12㊀2种控制器转子位置角对比波形图Fig.12㊀Comparison waveform of rotor position angleof two controllers65电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀为了验证不同工况下控制策略的运行状态,将电机转速设定为600r /min,给定负载2N㊃m,电机带载启动,当系统稳定运行到0.2s 时,转速由600r /min 突变到1000r /min,仿真结果如图13所示㊂在带载启动负载突变的情况下,由图13(a)可知,PI 控制下波形转速突变时存在超调,转速误差为200r /min 左右,50ms 后稳定运行且转速误差为-43~43r /min;滑模ADRC 控制下波形在0.2s 时负载突变,超调量几乎为0,转速波动更平稳且转速误差在-30~20r /min 之间㊂图13㊀2种控制器转速对比波形图Fig.13㊀Speed comparison waveform of two controllers图14为2种控制器转子位置误差对比波形,由两组对比图可知,基于PI 控制的改进MRAS 实际转子位置与估计转子位置之间有轻微延迟,在转速变化时转子位置误差略有增大,而基于滑模ADRC 控制的转子角可以在转速波动前后实现良好的跟踪㊂由以上对比分析可知,基于滑模ADRC 的无位置传感器矢量控制在电机启动阶段可以快速且无超调的达到稳态,转速误差波动较小,当负载突变时,使用滑模ADRC 控制可有效地抑制电机转速波动,在带载启动转速突变的情况下,采用滑模ADRC 使转速变化更平稳,保证控制器控制性能的同时提高了抗干扰能力㊂图14㊀2种控制器转子位置误差对比波形图Fig.14㊀Waveform diagram of rotor position error comparison between two controllers5㊀实验验证PMSM 控制平台如图15所示,由以TI 公司DSP芯片TMS320F28075为核心的控制电路㊁逆变电路㊁隔离驱动电路㊁电流采样电路和PMSM 组成㊂5.1㊀滑模ADRC 控制器实验为了验证转速环ADRC 的控制性能,将在电机转速恒定负载突变和电机带载启动转速突变2种工况下进行实验验证㊂1)负载突变情况㊂将转速设定为600r /min,图16和图17分别是滑模ADRC 控制器下转速和转子位置响应曲线㊂在0.34s 突加负载,由图16可知,估计转速大约降低了57r /min,此时转速误差为28r /min,0.6s 左右恢复运行,转速波动时间约为0.1s㊂从图17可知,在滑模ADRC 控制下,估计转子位置波形和实际转子位置波形吻合程度较高㊂由上述对比分析可知,与基于PI 控制的改进MRAS 相比,在滑模ADRC 控制下估计转速仍能较好地跟踪实际转速,且当负载变化时,转速脉动更75第4期肖㊀芳等:基于改进型ADRC 的PMSM 无位置传感器控制小,转速误差波动时间更短㊂因此,在转速恒定负载突变的条件下,滑模ADRC 控制的改进MRAS 具有良好的抗干扰能力㊂图15㊀硬件实验平台Fig.15㊀Hardware experimentplatform图16㊀600r /min 时负载变化前后转速及误差波形Fig.16㊀Speed and error waveform before and afterload change at 600r /min2)转速突变情况㊂在电机达到给定转速300r /min 后,将转速升至500r /min㊂由图18可知,PI 控制下的电机在2.5s 时转速由300r /min 升为500r /min,转速超调约为25r /min;滑模ADRC 控制的电机在2.7s 时转速升高至500r /min,转速超调降低至9r /min㊂由图19可看出,变速过程中PI 控制器和滑模ADRC 都可以精确跟踪转子位置信号㊂图20为转子速度由600r /min 升至1000r /min转速对比波形㊂从图中可看出基于PI 控制的MRAS 在转速突变时转速存在明显超调,超调量达到5%,大约0.9s 后达到稳定状态,基于滑模AD-RC 控制的MRAS 超调较小,约为1.4%,大约0.6s 达到稳态㊂在2种转速环控制器下转速误差波动都很小㊂图17㊀600r /min 时转子位置及误差波形Fig.17㊀Rotor position and error waveform at 600r /min图18㊀转速突变时转速对比及误差波形Fig.18㊀Speed comparison and error waveform duringsudden changes in speed由上述对比分析可知,在转速突变的情况下,虽然PI 控制的MRAS 和滑模ADRC 控制的MRAS 都能理想的跟踪转子位置,但滑模ADRC 控制的MRAS 在保证转速上升无明显超调的同时,也保证了响应速度,且转速响应曲线更加平稳㊂85电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀图19㊀转速突变时转子位置对比及误差波形Fig.19㊀Comparison of rotor position and error wave-form during sudden speedchanges图20㊀转速突变时转速对比及误差波形Fig.20㊀Speed comparison and error waveform duringsudden changes in speed5.2㊀MRAS 观测器实验为了验证改进前后MRAS 观测器的速度估计性能,对传统MRAS 和改进后的MRAS 进行了实验和对比验证,图21给出了传统MRAS 与改进MRAS在负载变化前后转速对比及转速误差波形㊂图21㊀600r /min 时负载变化前后转速对比及误差波形Fig.21㊀Speed comparison and error waveform beforeand after load change at 600r /min从图21可以看出,在负载突变时,传统MRAS 转速误差波动约为-21~29r /min,而改进MRAS 转速误差降低至-14~21r /min,有效抑制了转速波动㊂电机稳定运行时,改进前后的估计速度都可以准确跟踪实际速度,但与传统MRAS 相比,改进的MRAS 具有更小的转速波动和更高的估计精度㊂6㊀结㊀论本文针对ADRC 参数设置复杂的问题,通过对ADRC 线性化并结合滑模变结构理论,设计一种用于电机速度环的滑模ADRC 控制器㊂滑模ADRC 用于取代速度外环的PI 控制,同时结合改进型MRAS 对PMSM 进行无传感器控制,仿真结果表明,滑模ADRC 控制器不仅能在电机启动初期快速无超调的达到给定转速,而且能有效减小电机启动阶段的速度波动,当电机负载发生变化或突然升速时,滑模ADRC 具有更强的抗干扰能力和更快的恢复速度㊂通过设定电机参数运行时发生变化,对改进MRAS 算法进行验证㊂实验结果验证了所提出的改进型ADRC 与MRAS 相结合的无传感器控制策略95第4期肖㊀芳等:基于改进型ADRC 的PMSM 无位置传感器控制的可行性与有效性㊂参考文献:[1]㊀甘醇,曲荣海,石昊晨,等.宽调速永磁同步电机全域高效运行控制策略综述[J].中国电机工程学报,2023,43(7):2496.GAN Chun,QU Ronghai,SHI Haochen,et al.Review of high-ef-ficiency control dtrategy for permanent magnet synchronous motors under wide speed range[J].Proceeding of the CSEE,2023,43(7):2496.[2]㊀王晓远,刘铭鑫,陈学永,等.电动汽车用PMSM带滤波补偿三阶滑模自抗扰控制[J].电机与控制学报,2021,25(11):25.WANG Xiaoyuan,LIU Mingxin,CHEN Xueyong,et al.Third-or-der sliding mode active disturbance rejection control of PMSM with filter compensationfor electric vehicle[J].Electric Machines and Control,2021,25(11):25.[3]㊀左月飞,张捷,刘闯,等.基于自抗扰控制的永磁同步电机位置伺服系统一体化设计[J].电工技术学报,2016,31(11):51.ZUO Yuefei,ZHANG Jie,LIU Chuang,et al.Integrated design for permanent magnet synchronous motor servo systems based on active disturbance rejection control[J].Transactions of China Electrotechnical Society,2016,31(11):51.[4]㊀王江,王静,费向阳.永磁同步电动机的非线性PI速度控制[J].中国电机工程学报,2005,25(7):125.WANG Jiang,WANG Jing,FEI Xiangyang.Nonlinear PI speed control of permanent magnetic synchronous motor[J].Proceed-ings of the CSEE,2005,25(7):125.[5]㊀梁骄雁,胡育文,鲁文其.永磁交流伺服系统的抗扰动自适应控制[J].电工技术学报,2011,26(10):174.LIANG Jiaoyan,HU Yuwen,LU Wenqi.Anti-disturbance adap-tive control of permanent magnet AC servo system[J].Transac-tions of China Electrotechnical Society,2011,26(10):174. [6]㊀卢达,赵光宙,曲轶龙,等.永磁同步电机无参数整定自抗扰控制器[J].电工技术学报,2013,28(3):27.LU Da,ZHAO Guangzhou,QU Yilong,et al.Permanent magnet synchronous motor control system based on no manual tuned active disturbance rejection control[J].Transactions of China Electro-technical Society,2013,28(3):27.[7]㊀李思毅,苏健勇,杨贵杰.基于自抗扰控制的永磁同步电机弱磁控制策略[J].电工技术学报,2022,37(23):6135.LI Siyi,SU Jianyong,YANG Guijie.Flux weakening control strat-egy of permanent magnet synchronous motor based on active dis-turbance rejection control[J].Transactions of China Electrotech-nical Society,2022,37(23):6135.[8]㊀闫光亚,艾武,陈冰,等.永磁直线同步电机ADRC控制系统[J].电工技术学报,2011,26(9):60.YAN Guangya,AI Wu,CHEN Bing,et al.ADRC control system of permanent magnet linear synchronous motor[J].Transactions of China Electrotechnical Society,2011,26(9):60.[9]㊀董家臣,高钦和.永磁直线同步电机电流环新型线性自抗扰控制[J].电机与控制学报,2019,46(1):1.DONG Jiachen,GAO Qinhe.New current loop LADRC design for permanent magnet linear synchronous motor[J].Electric Ma-chines and Control,2019,46(1):1.[10]㊀刘艳红,李朋真,李汇来,等.线性自抗扰控制系统的稳定性分析方法[J].控制工程,2024,31(1):95.LIU Yanhong,LI Pengzhen,LI Huilai,et al.Stability analysismethod of linear active disturbance rejection control system[J].Control Engineering of China,2024,31(1):95. [11]㊀李杰,齐晓慧,万慧,等.自抗扰控制:研究成果总结与展望[J].控制理论与应用,2017,34(3):281.LI Jie,QI Xiaohui,WAN Hui,et al.Active disturbance rejec-tion control:theoretical results summary and future researches[J].Control Theory and Applications,2017,34(3):281. [12]㊀贾红敏,张立广,闫曌.基于改进自抗扰的永磁同步电机位置伺服系统[J].电机与控制应用,2020,47(10):33.JIA Hongmin,ZHANG Liguang,YAN Zhao.Permanent magnetsynchronous position servo system based on improved active dis-turbance rejection control[J].Electric Machines&Control Ap-plication,2020,47(10):33.[13]㊀李寅生,陈永军.BP神经网络补偿ADRC在PMSM控制中的研究[J].微电机,2020,53(2):61.LI Yinsheng,CHEN Yongjun.Research on BP neural networkcompensation ADRC in PMSM control[J].Micromotors,2020,53(2):61.[14]㊀TANG Yirong,XU Wei,LIU Yi,et al.Dynamic performanceenhancement method based on improved model reference adaptivesystem for SPMSM sensorless drives[J].IEEE Access,2021,9:135012.[15]㊀安群涛.永磁交流伺服系统的位置控制及相关技术的研究[D].哈尔滨:哈尔滨工业大学,2006.[16]㊀梁传栋,许刘超,潘志恒,等.一种改进型滑模变结构控制永磁同步电机直接转矩控制[J].电机与控制应用,2022,49(3):1.LIANG Chuandong,XU Liuchao,PAN Zhiheng,et al.An im-proved sliding mode variable structure control direct torque con-trol for permanent magnet synchronous motor[J].Electric Ma-chines&Control Application,2022,49(3):1. [17]㊀李明炎.基于SMO的PMSM的控制方法的研究[D].杭州:杭州电子科技大学,2014.[18]㊀李益敏,陈愚,李云龙.基于滑模自抗扰的永磁同步电机控制系统设计[J].电气传动,2019,49(8):22.LI Yimin,CHEN Yu,LI Yunlong.Design of PMSM control sys-tem based sliding mode active disturbance rejection[J].ElectricDrive,2019,49(8):22.(编辑:刘琳琳)06电㊀机㊀与㊀控㊀制㊀学㊀报㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀㊀第28卷㊀。
低定子频率下消除电流测量误差的磁链观测器
低定子频率下消除电流测量误差的磁链观测器周二磊;符晓;伍小杰;戴鹏【摘要】随着控制性能要求的提高,电流测量误差对电励磁同步电动机控制性能的影响愈加显著。
从电流测量路径看,产生误差不可避免,直流偏移和比例不平衡误差会造成转速周期性地波动。
本文采用一种简单的谐振式观测器对低定子频率下存在的电流测量误差进行补偿,并且为了消除残余误差,纯积分磁链观测器采用了残余误差补偿器以精确观测磁链。
最后,基于Matlab/Simulink对电流两相传感器的电励磁同步电动机调速系统进行仿真,仿真结果证明了该方案的有效性。
%As higher performance of control system is required,current measurement error seriously affected the control performance of electrically excited synchronous motor.The errors generated from the current measurement path are inevitable,such as offset currents and gain deviations,which causes the periodic rotor speed ripples.This paper presents a simple resonant type observer to compensate current measurement errors for the pure-integration-based flux estimation at a low stator frequency.The technique further contains a residual error compensator to eliminate miscellaneous residual error of the integrator.Finally,this paper uses the Matlab/Simulink to simulate excited synchronous motor with two phase current sensors.The simulation results show the effectiveness of the proposed scheme.【期刊名称】《电工技术学报》【年(卷),期】2011(026)006【总页数】6页(P67-72)【关键词】电流测量误差;谐振式观测器;纯积分磁链观测器;电励磁同步电动机【作者】周二磊;符晓;伍小杰;戴鹏【作者单位】中国矿业大学信息与电气工程学院,徐州221008;中国矿业大学信息与电气工程学院,徐州221008;中国矿业大学信息与电气工程学院,徐州221008;中国矿业大学信息与电气工程学院,徐州221008【正文语种】中文【中图分类】TM3411 引言电励磁同步电动机以其效率高、功率因数高且可以调节等优点,在机械传动,特别是在大功率传动中广泛应用[1]。
约束非线性系统理想点多目标安全预测控制
第41卷第2期2024年2月控制理论与应用Control Theory&ApplicationsV ol.41No.2Feb.2024约束非线性系统理想点多目标安全预测控制何德峰†,操佩颐,岑江晖(浙江工业大学信息工程学院,浙江杭州310023)摘要:考虑具有状态和控制约束的仿射非线性系统多目标安全控制问题,本文提出一种保证安全和稳定的多目标安全模型预测控制(MOSMPC)策略.首先通过理想点逼近方法解决多个控制目标的冲突问题.其次,利用控制李雅普诺夫障碍函数(CLBF)参数化局部控制律,并确定系统不安全域.在此基础上,构造非线性系统的参数化双模控制器,减少在线求解模型预测控制(MPC)优化问题的计算量.进一步,应用双模控制原理和CLBF约束,建立MOSMPC策略的递推可行性和闭环系统的渐近稳定性,并保证闭环系统状态避开不安全域.最后,以加热系统的多目标控制为例,验证了本文策略的有效性.关键词:非线性系统;模型预测控制;多目标控制;安全控制;稳定性引用格式:何德峰,操佩颐,岑江晖.约束非线性系统理想点多目标安全预测控制.控制理论与应用,2024,41(2): 355–363DOI:10.7641/CTA.2023.20542Utopia multi-objective safe predictive control ofconstrained nonlinear systemsHE De-feng†,CAO Pei-yi,CEN Jiang-hui(College of Information Engineering,Zhejiang University of Technology,Hangzhou Zhejiang310023,China) Abstract:This paper considers the multi-objective safe control problem of input-affine nonlinear systems subject to the constraints on the state and control.A new multi-objective safe model predictive control(MOSMPC)scheme is proposed for the system with guaranteed safety and stability.First,the utopia-point approximation method is adopted to solve the conflicting problem of multiple control objectives.Second,the control Lyapunov-barrier function(CLBF)is used to parameterize the local control laws of the system and determine the unsafe domains of the system.Then the parameterized dual-mode controller of the nonlinear system is constructed to reduce the computational amount of online solving the MPC optimization problem.Moreover,recursive feasibility of the MOSMPC scheme and asymptotic stability of the closed-loop system are established via the dual-mode control principle and the CLBF constraint,which ensures that the states of the closed-loop system can avoid the unsafe domain.Finally,an example of multi-objective control of a heating system is used to verify the effectiveness of the proposed strategy.Key words:nonlinear systems;model predictive control;multi-objective control;safety control;stabilityCitation:HE Defeng,CAO Peiyi,CEN Jianghui.Utopia multi-objective safe predictive control of constrained nonlinear systems.Control Theory&Applications,2024,41(2):355–3631引言安全与稳定是工业过程正常生产的前提,安全事故一旦发生,轻则造成生产过程严重的经济损失,重则威胁生产者的生命健康[1–2],如何保证安全与稳定运行成为工业生产过程的首要任务.除了工业现场的安全保护装置外,过程控制系统可以通过对控制器的约束设计保证生产过程的安全与稳定运行[3–5].因此,各种约束条件在工业过程控制中普遍存在[3–6].另一方面,工业过程控制问题通常涉及多个控制目标,如能耗、效率和控制速率等要求,这些控制目标通常相互冲突[6–8].因此,控制器的设计需要保证这些相互冲突的目标在运行过程中协调实现.现已证明,模型预收稿日期:2022−06−17;录用日期:2023−06−01.†通信作者.E-mail:**************.cn;Tel.:+86130****0667.本文责任编委:夏元清.国家自然科学基金项目(62173303),浙江省重点研发计划项目(2020C03056)资助.Supported by the National Natural Science Foundation of China(62173303)and the Key Research and Development Program of Zhejiang Province (2020C03056).356控制理论与应用第41卷测控制(model predictive control,MPC)具有在求解最优控制问题的同时对约束和多目标问题进行有效处理的优点,已被广泛应用于各种工业生产过程的最优控制[6–10].近年来,兼顾稳定性与安全性的MPC方法受到了学术界和工业界越来越多的关注.例如,文献[11–12]提出一种基于控制避障函数的安全反馈设计来保证系统的安全运行;文献[13]提出利用安全指标函数作为硬约束定义不安全区域,结合稳定控制和控制安全,结果显示所提出的方法可以实现非线性系统的闭环稳定性和运行安全性;文献[14–16]考虑控制目标和安全约束之间的潜在冲突,提出一种基于控制李雅普诺夫障碍函数(control Lyapunov-barrier function,CLBF)的控制策略,通过加权控制李雅普诺夫函数(control Lyapunov function,CLF)和控制障碍函数(control barrier function,CBF)兼顾闭环系统的稳定性和控制安全性,文献[14–16]同时还给出了存在多个不安全情况下的控制方法;文献[15]提出了一种基于CLBF的模型预测控制策略,解决满足约束和保证安全的非线性系统的稳定问题,保证状态收敛到稳态而不进入一个指定的不安全区域;文献[17]推广了不安全区域的定义,并通过基于CLBF的控制方法得到验证.现有基于CLBF的控制策略侧重系统安全性和稳定性控制目标,但缺少对工业过程更多控制目标的系统性处理.工业过程控制通常存在多个相互冲突的性能指标,本质上是一种多目标优化控制问题[18–19].加权函数法因其使用方便,通常被用来近似求解多目标优化控制问题,但确定适当的加权系数通常很困难,一般需要经过大量的离线实验才能得到,特别是复杂系统或非凸的多目标控制问题[20–21].为此,近年来相关学者提出了一些新的多目标优化MPC方法,如对控制目标进行优先级排序的优先级多目标MPC[22–24]、基于拐点的进化多目标优化MPC[25–26]以及基于多目标理想点逼近的多目标MPC[27–28]等.理想点逼近多目标M-PC策略通过帕累托前沿实现理想点目标跟踪,无需复杂的参数选择就能自动处理各个性能指标的冲突性,并获得令人满意的多目标控制结果,在工业过程多目标控制研究中得到了广泛关注.仿射输入非线性系统是工业过程控制中常用的一类非线性模型,广泛用于描述加热过程[29]、聚合反应过程[15–17]等工业过程的动态特性.本文考虑具有状态和控制约束的仿射输入非线性系统,提出一种具有稳定性和安全性保证的非线性系统多目标安全模型预测控制(multi-objective safe predictive control,MO-SMPC)策略.首先,采用理想点跟踪方法协调多个控制目标的冲突性.再利用约束CLBF的特性设计系统参数化局部控制律,保证闭环系统状态避开不安全区域.同时,构造非线性系统的参数化双模MPC控制器,在优化多目标性能的同时减少在线求解优化问题的计算量,从而可增加预测时域扩大闭环系统的初始可行域.进一步,应用双模控制原理和CLBF约束,建立MOSMPC策略的递推可行性和闭环系统的渐近稳定性,并且保证系统状态在初始可行域能始终避开不安全区域.最后,通过加热系统的多目标控制仿真实验验证本文结论的有效性和优越性.符号说明:I 0表示非负整数的集合;I a:b表示集合{i∈I 0:a i b},其中a∈I 0和b∈I 0;u(t0:t1)表示t∈[t0,t1]的一个连续时间信号u(t);给定初始状态x0,对于输入信号u(0:t),t时刻系统的解x(t)由x(t)=ϕ(t;x0,u(0:t))表示;符号|·|表示向量的2范数;上标T表示向量或矩阵的转置;“\”表示集合差,即A\B={x∈R n|x∈A,x/∈B};∅表示空集;∂D 表示集D⊂R n的边界;L f V是标量函数V(·)沿向量函数f(·)的李导数.2问题描述和预备考虑连续时间仿射输入非线性系统为{˙x(t)=f(x(t))+g(x(t))u(t),t 0,x(0)=x0.(1)其中:x(t)∈X和u(t)∈U是t时刻系统的状态变量和控制变量;x0是初始状态;X⊆R n和U⊆R m分别是状态和控制的约束集;f(x)和g(x)是自变量x的连续函数.不失一般性,令原点为该系统的平衡点,并假设系统状态是可测的.为简化书写,令F(x,u)= f(x)+g(x)u.考虑系统状态和控制约束为(x(t),u(t))∈Z,∀t 0,(2)其中Z⊆X×U是包含原点为内点的紧凸集.考虑性能函数向量L(x,u)=[L1(x,u)···L l(x, u)],其中L j:Z→R,j∈I1:l(l 2),并假设L j(x,u)关于x和u连续有界.则定义系统(1)的多目标稳态优化问题为min(x,u)∈Z{L(x,u):F(x,u)=0},(3)由于各性能函数L j(x,u)相互冲突,无法同时取得各个性能指标的最优性,通常采用帕累托(Pareto)最优性定义问题(3)的最优解,即给定优化问题(3)的可行解(x ps,u ps),如果不存在其他可行解,使得如下不等式成立:L j(x,u) L j(x ps,u ps),j∈I1:l,(4)且不等式组中至少有一个j∈I1:l,使得L j(x,u)<L j(x ps,u ps).考虑系统(1)在采样时间t k的状态x k,即x(t k)= x k,则定义有限预测时域0<T<∞上的目标函数为J j(x k,u(t k:t k+T))=第2期何德峰等:约束非线性系统理想点多目标安全预测控制357∫tk+Tt kL j(x(s),u(s))d s,(5)其中s为预测时间.注意单目标MPC通过最小化单个目标函数J j(x,u)优化镇定系统(1),而多目标MPC除闭环系统的稳定性外,应同时优化多个相互冲突的目标函数,因此,定义系统(1)的多目标有限时域最优控制问题为minu(t k:t k+T)J(x k,u(t k:t k+T))s.t.˙x(s)=F(x(s),u(s)),∀s∈[t k,t k+T], (x(s),u(s))∈X×U,∀s∈[t k,t k+T],x(t k)=x k,(6)其中:u(t k:t k+T)是在采样时间t k预测的未来时段[t k:t k+T]的控制输入;J(x,u)是需要最小化的l个目标函数向量,即J(x,u)=[J1(x,u)J2(x,u)···J l(x,u)],(7)如果多目标最优控制问题(6)存在可行解,则设u∗(t k:t k+T)是问题(6)的一个帕累托最优解.根据滚动时域优化控制原理,多目标MPC控制律定义为u(t)=u∗(t k:t k+1),∀t∈[t k,t k+1),k∈I 0,(8)即u∗(t k:t k+T)作用于系统(1)直到下一个采样时刻t k+1=t k+δ,其中采样周期δ>0.在下一采样时刻t k+1,用更新的状态x k+1重复整个过程.多目标MPC控制律(8)作用下的闭环系统的稳定性无法由优化问题(6)中目标函数的最优性保证.进一步,闭环状态演化过程中存在不安全域,如温度、压力、浓度过高等运行区域,但优化问题(6)无约束条件使闭环状态避开系统不安全域.令开集X d表示系统(1)运行区间的不安全区域,本文目标是寻找一个最优反馈控制u(x)∈U,使闭环系统状态轨迹x(t;x0, u)∈X但始终避开X d,即x(t;x0,u)/∈X d,∀t 0, x0∈X,从而确保系统是最优且安全运行的.对此,本文将引入CLBF概念设计安全约束,采用MPC方法保证闭环系统能安全避开不安全域,并渐近稳定于平衡点.考虑系统(1)的一个连续可微函数W c(x),定义集X e=x∈X\(X d∪0)|W c(x)/∂x=0,X c=x∈X|W c(x) 0和X uc={x∈X|˙W c(x)=L f W c(x)+L g W c(x)u(x)< 0,u(x)∈U}∪0∪X e,其中L f W c(x)和L g W c(x)分别为W c(x)对函数f(x)和g(x)的李导数.定义1[15]考虑系统(1)及不安全域X d⊂X uc,如果函数W c(x)有下界,在原点有最小值,并满足如下:1)W c(x)>0,∀x∈X d;2)|∂W c(x)∂x| r(|x|);3)X c=∅;4)X uc\(X d∪X c)∩¯X d=∅;5)L f W c(x)<0,∀x∈{s∈X uc\(X d∪0∪X e)| L g W c(s)=0},其中r是K类函数,则W c(x)是该系统的一个控制李雅普诺夫障碍函数CLBF.注1在实际中,CLBF可以由控制李雅普诺夫函数和控制障碍函数复合而成.令V(x)和B(x)分别是系统(1)的控制李雅普诺夫函数[30]和控制障碍函数[15],则该系统的一个CLBF为W c(x)=V(x)+λB(x)+ν,其中λ和ν可由V(x)和B(x)的上下界给定,详见文献[15].3多目标安全MPC设计3.1理想点计算考虑优化问题(3)的第j∈I1:l个性能函数L j(x,u),求解对应稳态优化问题为L∗s,j:=min(x,u)∈Z{L j(x,u):F(x,u)=0},(9)得最优解(x∗s,j,u∗s,j),其中最优值L∗s,j=L j(x∗s,j,u∗s,j).由于各性能函数L j(x,u)相互冲突,故各L j(x,u)对应的最优解(x∗s,j,u∗s,j)不同.为此,应用L∗s,j定义目标函数向量J(x,u)的理想点为J∗s=T[L∗s,1L∗s,2···L∗s,l]T,(10)显然,理想点J∗s是目标函数向量J(x,u)的不可达点,但给出了各个目标函数J j(x,u)的理想期望性能.因此,求解与该理想点J∗s最接近性能函数值对应的稳态解(x cs,u cs)为(x cs,u cs)=arg min(x,u)∈Z{∥T L(x,u)−J∗s∥p:F(x,u)=0},(11)其中∥·∥p是向量p范数.稳态解(x cs,u cs)又称为多目标优化问题(6)的折衷稳态点.因此,本文多目标MPC控制器的设计遵循使J(x,u)逐渐逼近J∗s并使闭环系统稳定于x cs的原则实现各个目标函数的最优化.3.2基于CLBF的控制器设计为应用CLBF概念设计MPC控制器,首先将系统(1)的折衷稳态点(x c s,u c s)平移至原点.令坐标转换z=x−x cs和v=u−u cs,则系统(1)可变换为˙z=f(z(t)+x s)+g(z(t)+x s)(v(t)+u s):=¯F(z,v),(12)令Z d=X d−x c s及W c(z)为系统(12)的一个CLBF,则有以下结论.引理1考虑系统(12)及其不安全域Z d,并给定358控制理论与应用第41卷实数D 1>0和D 2>0,则存在非空集S T 及其反馈控制律v (z )=h (z,µ)=−p (z,µ)β(z )T ,(13)其中参数µ=(µ1,µ2)∈D =[0,D 1]×(0,D 2],增益为p (x,µ)=α(z )+µ1√α(z )2+µ2|β(z )|4|β(z )|2,β(z )=0,0,β(z )=0,(14)其中α(z )=L f W c (z )和β(z )=L g W c (z ),使其闭环系统满足约束(2),并在不变集S T 中渐近稳定,同时使闭环状态避开不安全域Z d .证令S T ⊂X 为W c (z )的最大水平集,则由CL-BF 定义可知,集S T 非空.当z ∈S T \Z d ,由CLBF 定义可得W c (z ) 0.对W c (z )沿闭环系统状态轨迹求导得˙Wc (z )=α(z )+β(z )h (z,µ)=−µ1√|α(z )|2+µ2|β(z )|4,(15)当β(z )=0时,由CLBF 定义可得˙W c (x )=α(z )<0;当β(z )=0时,˙Wc (x )=−µ1√|α(z )|2+µ2|β(z )|4,即W c (z (t ))<W c (z (0))<0,∀t 0,z (t )/∈Z d .则应用定理1[15]的证明思路可得,闭环系统状态轨迹z 在控制律h (z,µ)作用下保持在域S T \Z d 内,且在不变集S T 内渐近稳定.注2不变集S T 的大小与参数µ取值相关.文献[31]给出了一种不变集S T 的选取方法,如下:先定义集Z h ={z ∈R n |∃µ∈D s .t .h (z,µ)+u c s ∈U }和S T (r )={z ∈R n|W c (z ) r },再将r 从0逐渐增加,直到集(X −x c s )∩Z h无法包含S (r ),从而得到与r max 相关的最大不变集S Tmax ,令S Tmax ={z ∈R n |W (z ) r max }⊆X ,则每时刻都至少存在一个µf ∈D ,使具有控制器(13)的闭环系统(12)满足约束(2),并且闭环状态轨迹在避开不安全域Z d 的前提下渐近收敛到原点.由引理1和注2可知,闭环系统(12)–(13)存在一个不变集S max ,使得闭环系统渐近稳定到平衡点,并且满足对状态量和控制量的约束.为简化书写,令S T (x c s )={x ∈X :x =z +x cs ,∀z ∈S max },其中至少存在一个可行的µf ∈D ,使相应的控制器u (x )=h (x −x c s ,µf )+u cs 满足系统状态和控制约束(2).为设计约束系统(1)–(2)的多目标安全MPC,定义参数化双模控制律如下[30]:u DM (x )={h (x −x c s ,µf )+u cs +c,x /∈S T (x c s ),h (x −x c s ,µ)+u cs ,x ∈S T (x c s ),(16)其中:c ∈R 是求解多目标最优控制问题的修正项;µ是控制器参数向量.在k ∈I 0的每个采样时间t k ,如果状态x (t k )/∈S T (x c s ),则通过在线求解优化问题J (x,u )得到c (t k );如果x (t k )∈S T (x c s ),则在线求得µ(t k ).由此将得到系统(1)满足约束(2)的稳定多目标安全MPC.注3传统双模控制方法[32–35]仅通过在线计算修正项求解最优控制问题,而在终端域S T (x c s )中定义的终端控制律通常是通过系统(1)的线性化模型离线确定的.在本文策略中,当闭环系统状态进入终端域时,通过CLBF 得到一个带可变参数的状态反馈控制律,整体在线更新.这种修改一方面将终端域内外的计算统一到一个代价函数,有利于解决需要在线调整成本函数的控制要求.另一方面,通过引入可变参数,最大程度上弱化控制器和终端域的耦合性,通过在终端控制律(16)中选择一些可行参数µf ,可以离线计算S T (x c s ),降低多目标MPC 在线优化时的计算量.3.3双模多目标安全MPC 算法考虑约束系统(1)–(2),稳态折衷点(11)和双模控制器(16),定义折衷性能指标函数为ˆJ(x (t k ),u (t k :t k +T ))=∥J (x (t k ),u (t k :t k +T ))−J ∗s ∥p ,(17)则定义系统(1)的多目标安全有限时域最优控制问题分别为minc (t k t k +T )ˆJ(x (t k ),u (t k :t k +T ))s.t.˙x (s )=F (x (s ),u (s )),(x (s ),u (s ))∈Z ,u (s )=h (x (s )−x c s ,µf )+u cs +c (s ),x (t k +T )∈S T (x c s ),x (t k )=x k ,∀s ∈[t k ,t k +T ](18)和min µ(t k )∈DˆJ(x k (t k ),u (t k :t k +T ))s.t.˙x (s )=F (x (s ),u (s )),(x (s ),u (s ))∈Z ,u (s )=h (x (s )−x c s ,µ(t k ))+u c s ,x (t k )=x k ,∀s ∈[t k ,t k +T ],(19)其中:目标函数向量J (x,u )和其稳态理想点J ∗s分别由式(7)–(10)给定;c (t k :t k +T )为采样时刻t k 的预测范围[t k t k +T ]内的预测修正.由此本文提出的考虑安全性的双模多目标安全MPC 算法归纳如下:步骤1设定采样周期δ>0、预测时域T =T 0>δ、多个性能指标函数L j 和参数域D ;步骤2离线计算理想点J ∗s ,折衷解(x c s ,u c s )和带参数µf ∈D 的终端不变集S T (x c s );设k =0和t k =0;步骤3在采样时刻t k ,测量当前状态x k ,令x (t k )=x k ;第2期何德峰等:约束非线性系统理想点多目标安全预测控制359步骤4如果状态x (t k )/∈S T (x c s ),在线解决优化问题(18)得到修正项c ∗(t k ),转入步骤5;否则,求解优化问题(19)得µ∗(t k ),转入步骤6,令T =T 0;步骤5将u DM(t )=h (x (t )−x c s ,µf )+u c s +c ∗(t k:t k +1)应用到系统(1),直到下一个采样时刻;令k =k +1和T =T −δ;返回步骤3;步骤6将u DM(t )=h (x (t )−x c s ,µ∗(t k ))+u cs 应用于系统(1),直到下一个采样时刻;令k =k +1并返回步骤3.定义2考虑约束系统(1)–(2),若系统的某一初始状态x (t 0)/∈S T (x c s ),且在此初始时刻优化问(18)存在可行解满足其约束条件,则称x (t 0)为闭环系统的初始可行状态,所有满足条件的初始可行状态值构成的集合称为初始可行域,记为X f0.定理1考虑约束系统(1)及其多目标安全有限时域最优控制问题(18)–(19),则在充分长的预测时域T 内该优化问题是递推可行的.证由注2–3可知,当状态x (t k )∈S T (x c s )时,至少存在一个可行的参数向量µf 满足约束(19);当状态x (t k )/∈S T (x c s )时,由双模控制策略可知,闭环状态可以在一个有限时域T 内被驱动到终端域S T (x c s )内,一旦状态进入S T (x c s ),则回到上一种情况.因此可得算法1中的最优控制问题在容许状态集X f0中是递推可行的.证毕.定理2考虑约束系统(1)-(2)及其控制李雅普诺夫避障函数W c (x )和不安全区域X d ,则对任意初始状态x 0∈X f0,算法1作用下的闭环系统稳定到稳态x c s ,且始终不会进入不安全区域X d .证已知对于初始状态x (t 0)/∈S T (x c s ),在双模控制器作用下,闭环系统状态可在有限时域T 内进入终端域S T (x c s ),记为x T ∈S T (x cs ).假设x T ∈S T (x c s )X d,则由CLBF 定义可得˙W c (x (t ))<0,∀x (t )∈S T (x c s)\(X d ∪{0}),(20)即W c (x (t ))<W c (x (0))<∞,∀t 0.根据控制李雅普诺夫避障函数W c (x )的特性可知,闭环系统的状态轨迹是有界的.这意味着由x 的极限点构成的集合Ω(x )是非空的连通紧集,且lim t →∞d (x (t ),Ω(x ))=0,其中d 为状态和集的距离[36].由注1或文献[15]可知,函数W c (x )=V (x )+λB (x )+ν的控制李雅普诺夫函数V (x )为正定函数,参数和存在下界,则W c (x )存在下界.又不等式(20)意味着W c (x )是单调递减函数,故W c (x )必收敛.考虑对于由x 的极限点构成的集合t n →∞时x (t n )→ξ.由W c (x )的连续性可知,lim t n →∞W c (x (t n ))=W c (ξ),因此,可得Ω(x )={ξ∈X |W c (ξ)=lim n →∞W c (x (t n ))}⊂S T (x c s ),则W c (x )收敛于集Ω(x ),且沿着此时状态轨迹有u =0.当u (t )=0,t 0时,可得d (x (0),Ω(x ))=0,则d (x (t ),Ω(x ))=0,t 0.因此在集Ω(x )中,W c (x )=W c (0)=0,得到Ω(x )={0},即lim t →∞|x (t )|=0.注4尽管优化问题(18)–(19)计算得到的参数c ∗和μ∗是开环解,但由于参数化双模控制律(16)结构特点,算法1得到的多目标MPC 控制器是闭环控制律.进一步,在算法1的预测时间窗口[t k ,t k +T ]内的参数µ的值是不变的,从而,通过参数化压缩了优化问题的决策变量维数,因此,这将有助于减轻算法1的在线优化计算量.注5由于算法1的双模形式,MPC 的控制律关于x 通常是不连续的.然而,对于任何初始条件x 0∈X f0,状态x 都可以被驱动到x ∈S T (x c s ).一旦x 进入S T (x cs ),则可以通过基于CLBF 的解析控制律(16)实现连续条件下的渐近稳定控制.4实例仿真考虑一个多输入多输出非线性加热系统,如图1所示.图中加热系统由一个外部加热装置和一个内部可拆卸传热容器组成,内部容器中可放置需要加热的对象.加热过程目标是通过调节外部加热装置的温度T h 和内部容器的温度T n ,从而对容器中的对象进行加热.这是由加热装置的两个加热器共同控制的,可用的控制输入分别是加热器提供的两个电源W h1和W h2.此外,内部管道通过水温对内部容器进行温度调节实现热量交换,而外部温度通过引起环境的辐射冷却来干扰加热系统.该系统可以表征中药等加热过程,将需要处理的中药放置在内部容器中进行加热处理.图1加热系统示意图Fig.1Schematic representation of the heating system考虑图1所示加热系统,应用传热学和能量守恒原理,加热系统的动力学表示如下[29]: m n c n ˙T n (t )=W n0+A 1h 1(T h (t )−T n (t )),m h c h ˙Th (t )=W h1(t )+W h2(t )−A 1h 1(T h (t )−T n (t ))+A 2h 2(T 4ext(t )−T 4h (t )),(21)其中:T h 是内部可拆卸容器温度;T n 是加热对象温度;T ext 是加热系统的外部温度;m n 是内部容器质量;m h 为加热装置质量;c n 为内部容器比热;c h 为加热装置比热;h 1为内部容器对流系数;h 2为加热装置对流360控制理论与应用第41卷系数;A 1为内部容器面积;A 2为加热装置面积;W n0为每个采样间隔内部管道带来的热量;W h1和W h2为加热器电源功率.该加热系统模型参数值:T ext =283K;m n =3.0kg;m h =20.0kg;c n =300J /kg ·K;c h =4000/kg ·K;h 1=200kg /K ·s 3;h 2=9.203×10−7kg /K ·s 3;A 1=0.3m 2;A 2=0.8m 2;W n0=250W .分别选择内部容器温度T n 和外部加热装置温度T h 为该加热系统的状态变量x 1和x 2,选择加热器W h1和W h2分别为控制输入u 1和u 2,即系统的状态和控制输入x =[T n T h ]T 和u =[W h1W h2]T ,且满足以下约束:200K T n 400K ,0W Wh112000W ,200K T h 400K ,0W W h240400W .(22)对加热系统而言,通常希望在使用过程中降低系统能耗,即最小化性能函数L 1(x,u )=W h1+W h2,(23)同时希望温度变化率尽可能大,使系统在保证安全的前提下尽快达到指定温度,即最小化性能函数L 2(x,u )=−(˙T 2n +˙T 2h ),(24)并且跟踪设定温度373K,即最小化性能函数L 3(x,u )=|T h −373|+|T n −373|,(25)因此,在设计加热系统控制器时应同时满足上述3个性能要求.定义性能函数向量L (x,u )=[L 1(x,u )L 2(x,u )L 3(x,u )]T .根据稳态优化问题(9),L (x,u )的理想稳态点计算为L ∗s =[004.167]T,分别对应稳态解O 1(x ∗s ,u ∗s )=(300.485,296.18,0,0),O 2(x ∗s ,u ∗s )=(300.531,296.364,2.101,1.451)和O 3(x ∗s ,u ∗s )=(373.742,69.576,713.323,7345.673),上述3组稳态最优解不一致,表明L 1(x,u ),L 2(x,u ),L 3(x,u )之间具有冲突性,即加热系统控制是一个多目标冲突的优化控制问题.由式(11)计算2范数下的折衷稳态解O (x c s ,u cs )=(300.534,296.368,1.898,1.898),则加热系统的控制目标是同时最小化性能函数(23)–(25),使系统在满足约束的前提下安全渐近稳定到折衷稳态点.本文算法通过平衡点O (x c s ,u cs )定义移位状态向量z =x −x c s 和控制变量v =u −u cs .为此,选择移位系统的CLF 为定义控制李雅普诺夫函数V (z )=z T P z ,其中P =[29.744.844.8195.8].不安全域Z d 定义为终端区域内的一个开集,状态域Z d 中外部加热装置和内部容器的温差较大,包含在实际加热过程中会产生安全威胁的不安全状态,其范围表示为Z d :={z ∈R 2|F (z )=(z 1+4.2)2+(z 2−1)2/10<0.06}.又定义H :={z ∈R 2|F (z )<0.07},可由文献[15]设计控制避障函数为B (z )=F (z )e F (z )−0.07−e −6,z ∈H ,−e −6,z /∈H ,(26)其中,对于所有状态z ∈H ,B (z )>0.按照注1构造控制李雅普诺夫避障函数W c (x )=V (x )+λB (x )+ν,其中参数c 1=15,c 2=210,c 3=max z ∈∂H|z |2=29.99,c 4=min z ∈∂Z d|z |2=10.62,λ=2.1×106,ν=−c 1c 4=−159.3.终端域S T (x c s )={z ∈R 2:W c (z ) −4764.68},通过离线试错得参数向量µ的范围为[0.0110]×[0.0110].在仿真中,取采样周期为2s,预测步长为10,仿真总步长为500.采用MATLAB2021A 软件中的fminc-on 函数优化计算最优控制问题.选取系统初始状态点A (−5.51.9),B (−61.7)和C (−51.4).图2给出了闭环系统从3个不同初始状态点到稳态点的状态移动轨迹,图中结果显示,所有闭环状态能避开不安全区域,并能最终收敛到稳态点.加热系统从初始状态A 开始,比较本文方法DM-MOSMPC 和基于CLF 的MOMPC [28]控制下的状态轨迹,结果如图3所示,其中,实线表示基于CLBF 的安全MOMPC 控制下的状态轨迹,虚线表示基于CLF 的MOMPC 控制下的状态轨迹.结果显示,本文基于CLBF 的双模安全多目标MPC 将闭环系统的状态保持在稳定安全区域内并驱动到稳态点,而基于CLF 的MOMPC 无法对状态空间中的不安全域进行躲避.因此,DM-MOSMPC 在状态约束下优于基于CLF 的MOMPC,同时,保证系统的安全性和闭环稳定性.图2闭环系统不同初始点的状态轨迹Fig.2The closed-loop state trajectories of different initialpoints为比较本文DM-MOSMPC 和CLBF-MOMPC 方第2期何德峰等:约束非线性系统理想点多目标安全预测控制361法,选择系统从初始状态B 开始,系统状态轨迹如图4所示,其中:实线表示本文控制方法,虚线表示CL-BF-MOMPC 方法[16].系统的状态曲线如图5所示,结果表明本文方法更快更准确地趋近稳态点.3个性能函数的优化过程如图6所示,其中,点实线是稳态性能,实线表示DM-MOSMPC 方法,虚线表示CLBF-MO-MPC 方法.曲线表明,多个冲突经济目标通过控制器的优化迅速下降,直到稳定在稳态性能附近.另外,在计算量方面也可以看到本文方法的优势.图7显示本文方法DM-MOSMPC(蓝色柱状图)和CL-BF-MOMPC 方法(红色柱状图)在不同预测时域进行一次在线优化时的计算CPU 时间的比较.结果显示,本文方法在线优化所用的计算CPU 时间在各个不同的预测时域比所选取的对比方法时间短.图3DM-MOSMPC 和DM-MOMPC 方法的闭环系统状态轨迹Fig.3The closed-loop state trajectories under DM-MOSMPCandDM-MOMPC图4DM-MOSMPC 和CLBF-MOMPC 方法的闭环系统状态轨迹Fig.4The closed-loop state trajectories under DM-MOSMPCandCLBF-MOMPC图5DM-MOSMPC 和CLBF-MOMPC 方法的状态曲线Fig.5The state profiles under DM-MOSMPCandCLBF-MOMPC图6DM-MOSMPC 和CLBF-MOMPC 方法的性能函数曲线Fig.6The Performance function profiles underDM-MOSMPC and CLBF-MOMPC362控制理论与应用第41卷图7DM-MOSMPC和CLBF-MOMPC方法的在线优化计算的CPU时间Fig.7The average computational CPU time for online optimi-zation of DM-MOSMPC and CLBF-MOMPC为比较不同的多目标处理方法,表1给出理想点和加权两种方法对应的系统状态量的稳态误差.在加权方法中,3个目标函数(23)–(25)被转化为一个加权和形式的单目标函数L w(x,u)=w1L1(x,u)+w2L2(x,u)+w3L3(x,u),(27)其中标量w i>0为权重因子,i=1,2,3.在加权法中分别采用3种权重,对应的(w1,w2,w3)分别为a(100,0.1,0.1),权重b(10,50,0.1)和权重c(0.1,10, 100).表1理想点和加权方法的稳态误差Table1Steady-state error of the Utopia point andweighting methods控制策略E1=x1−x c s1E2=x2−x c s2本文方法0.05470.221加权法(a)0.06420.376加权法(b)0.06460.353加权法(c)0.07360.795可以看出,本文方法得到的稳态误差相对较小,在加权方法中对不同权值的选择会得到不同的误差.这意味着在实际中适当的权值是难以选择的,缺乏系统性的权重调整规则,并且会增加控制器实现的复杂性.需要指出的是,此仿真对比并不是为了说明加权方法在控制性能上不如理想点方法,而是为了说明前者过分依赖手工调优和设计者的经验,而本文方法则提供一种更为系统的方法来处理多目标控制问题.通过上述仿真结果和讨论,可以得到对于系统(21),双模MOSMPC算法能够更有效实现加热控制目标.5总结本文考虑连续时间仿射输入非线性系统,提出一种多目标安全模型预测控制(MOSMPC)策略.该策略首先利用理想点逼近方法解决多控制目标冲突问题.其次,引入满足约束的CLBF设计参数化局部控制律,并由约束保证闭环系统的安全性.进一步,应用双模控制原理和CLBF约束,建立MOSMPC策略的递推可行性和闭环系统的渐近稳定性,并且策略保证系统状态在初始可行域能始终避开不安全区域.最后,通过对一加热系统的多目标控制仿真对比实验,验证了本文策略的有效性和优越性.在此基础上,未来的工作方向包括对多目标安全非线性模型预测控制(nonli-near model predictive control,NMPC)的鲁棒性和多目标优化算法的研究.参考文献:[1]WANG rge property damage losses in the hydrocarbon-chemical industries.Petroleum Planning&Engineering,1993,4(3): 59–60.[2]JIANG Chunming,LI Qi.The development and application of pro-cess safety management and technology.China Petroleum and Chem-ical Standard and Quality,2007,27(5):45–49.(姜春明,李奇.过程安全管理与技术的发展与应用.中国石油和化工标准与质量,2007,27(5):45–49.)[3]KLATT K,MARQUARDT W.Perspectives for process systemsengineering-personal views from academia and puters and Chemical Engineering,2009,33:536–550.[4]Liu Derong.The advanced control of complex industrial process.Ac-ta Automatica Sinica,2014,40(9):1841–1842.(刘德荣.复杂工业过程的先进控制.自动化学报,2014,40(9): 1841–1842.)[5]KRETCHIK JOE T.Process safety management of highly hazardouschemicals.Chemical Health&Safety,2000,7(5):44.[6]AGRAWAL N,RANGAIAH G P,RAY A,et al.Multi-objective op-timization of the operation of an industrial low-density polyethylene tubular reactor using genetic algorithm and its jumping gene adap-tations.Industrial&Engineering Chemistry Research,2006,45(9): 3182–3199.[7]FLORES-TLACUAHUAC A,MORALES P,RIVERA-TOLEDO M.Multiobjective nonlinear model predictive control of a class of chem-ical reactors.Industrial&Engineering Chemistry Research,2012, 51(17):5891–5899.[8]WOJSZNIS W,MEHTA A,WOJSZNIS P,et al.Multi-objective opti-mization for model predictive control.ISA Tansactions,2007,46(3): 351–361.[9]ANGELI D,AMRIT M,RAWLINGS J B.On average performanceand stability of economic model predictive control.IEEE Transac-tions on Automatic Control,2012,57(7):1615–1626.[10]MULLER M A,ALLGOWER F.Economic and distributed modelpredictive control:Recent developments in optimization-based con-trol.SICE Journal of Control,Measurement,and System Integration, 2017,10(2):39–52.[11]WIELAND P,ALLGOWER F.Constructive safety using control bar-rier functions.IFAC Proceedings Volumes,2007,40(12):462–467.[12]AMES A D,XU X,GRIZZLE J W,et al.Control barrier functionbased quadratic programs for safety critical systems.IEEE Transac-tions on Automatic Control,2017,62(8):3861–3876.[13]ALBALAWI F,DURAND H,CHRISTOFIDES PD.Process opera-tional safety using model predictive control based on a process Safe-ness puters and Chemical Engineering,2017,104:76–88.。
基于改进锁相环的电动汽车用PMSM转子位置辨识方法
基于改进锁相环的电动汽车用PMSM转子位置辨识方法作者:徐达杨绿吴怀超袁焱何锋来源:《贵州大学学报(自然科学版)》2019年第04期摘要:为提高电动汽车用永磁同步电机无位置传感器转子位置辨识性能,改进设计基于滑模观测器的转子位置辨识方法,将低通滤波器输出量引入滑模观测器,用于观测电机反电动势,缓解低通滤波器造成的信号相位滞后问题。
在转子位置辨识环节中,改进的锁相环较传统锁相环具有更强的鲁棒性,且能够实现无转动方向差别的转子位置辨识功能。
MATLAB/Simulink仿真结果显示,基于改进锁相环的PMSM转子位置辨识方法具有良好的辨识精度、动态响应性和正反向运行适应能力,该方法易于实现,鲁棒性强,简化转子位置辨识系统的设计。
关键词:永磁同步电机;无位置传感器控制;滑模观测器;锁相环中图分类号:TM341文献标识码: A永磁同步电机(Permanent Magnet Synchronous Motor, PMSM)具有高功率密度、高效率、弱磁升速能力强等特点,是电动汽车的主要动力源之一。
PMSM的矢量控制需要准確的转子位置信息作为参考,编码器和旋转变压器的使用会导致电机体积增大和可靠性降低等问题,但无位置传感器技术可用于替代物理传感器的使用,为上述问题提出解决方案。
在无位置传感器转子位置辨识方法中,电机反电动势因含有转子位置信息被广泛应用于转子位置辨识。
滑模观测器具有结构简单、易于实现、鲁棒性强,且对被控对象参数变化不敏感等特点,被广泛应用于永磁同步电机无位置传感器控制[1]。
在传统的滑模观测器设计中,将sign函数作为滑模切换函数,因其为分段函数,会导致抖振现象。
为消除传统滑模观测器的抖振效应,可重新设计滑模观测器前端的低通滤波器,文献[2]在低通滤波器的基础上引入了可调节相差的卡尔曼滤波器,文献[3]在转子位置辨识环节引入了自适应陷波滤波器来替代低通滤波器;或设计平滑的开关函数[4-5];或应用更加准确的控制系统模型,如文献[6]建立了全阶离散滑模观测器,文献[7]则对负载转矩扰动进行估计和补偿;文献[8]根据低通滤波器的截止频率和电机电角速度设计了相位补偿方法。
基于灰狼优化的永磁同步电机自适应反推鲁棒控制策略
俞沛宙 1,王 澍 2,杨继辉 1,张宇华 3,薛 花 3,黄 珂 3
(1.国网浙江宁波市鄞州区供电有限公司,浙江 宁波 315100;2.国网浙江省电力有限公司双创中心, 浙江 杭州 310051;3.上海电力大学电气工程学院,上海 200090)
摘要:针对永磁同步电机存在非线性项的不确定性,以及在运行过程中参数摄动与负载扰动对机械角频率、转速 以及系统控制产生的影响,为适应高精度、高稳定的调速应用要求,提出一种基于灰狼优化的永磁同步电机自适 应反推鲁棒控制策略。应用非线性反推控制解决电机的非线性问题,并将灰狼优化应用于自适应反推算法中,实 现电机转速对期望值的自适应调节跟踪,从而获得较好的动态响应及鲁棒性。首先,建立包含参数摄动以及输出 侧负荷扰动的永磁同步电机数学模型,设计出满足李雅普诺夫稳定的反推控制器。其次,用灰狼算法对反推控制 器参数进行优化,进一步提高系统的稳定性和鲁棒性。仿真结果验证了该控制器具有良好的速度跟踪与抗干扰调 节效果,表明了该方法的有效性。 关键词:永磁同步电机(PMSM);自适应反推控制;李雅普诺夫;灰狼优化;鲁棒性
This work is supported by the Shanghai Science and Technology Innovation Action Plan (No. 19DZ2204700) and the Innovation and Entrepreneurship Incubation Fund Project of Zhejiang Electric Power Company. Key words: permanent magnet synchronous motor (PMSM); adaptive backstepping control; Lyapunov; grey wolf optimization; robustness
不确定分数阶PMSM混沌系统自适应滑模控制
不确定分数阶PMSM混沌系统自适应滑模控制林飞飞;曾喆昭【摘要】针对参数不确定分数阶永磁同步电机混沌系统,研究在含非线性不确定项和外部扰动情况下的控制问题.结合自适应控制理论和滑模控制理论,通过选取一种具有较强鲁棒性的分数阶积分滑模面,设计自适应滑模控制器.该控制器可在参数不确定项、非线性不确定项和外部扰动的上界未知的情况下,实现局部渐进稳定.数值仿真结果验证了该控制器的有效性.【期刊名称】《电力科学与技术学报》【年(卷),期】2018(033)002【总页数】7页(P66-72)【关键词】分数阶;永磁同步电机;分数阶积分滑模面;自适应滑模控制【作者】林飞飞;曾喆昭【作者单位】长沙理工大学电气与信息工程学院 ,湖南长沙 ,410004;长沙理工大学电气与信息工程学院 ,湖南长沙 ,410004【正文语种】中文【中图分类】TM341;TP273永磁同步电机(Permanent Magnet Synchronous Motor, PMSM)具有结构简单、体积小、运行可靠、能量转换效率高和响应快等优点[1-2],因此,在航空航天、家用电器和工业自动化设备等领域得到了广泛的应用。
PMSM是典型的多变量、强耦合非线性系统,在满足一定参数和变量条件下会呈现混沌现象[3],主要表现为转速和转矩的间歇振荡、系统不规则的电磁噪声和控制性能不稳定等。
这些不规则运动严重影响到系统的稳定性和运行质量,因此,控制PMSM混沌系统对PMSM的稳定运行具有重要的实际意义。
关于PMSM混沌系统控制人们提出大量的控制方法,主要有滑模控制[4]、非线性比例控制[5]、自适应反步控制[6]等。
分数阶微积分和整数阶微积分几乎同时出现,距今已有300多年的历史。
随着计算机技术的发展和有效的计算方法的出现,分数阶微积分被广泛地应用到物理学、工程实际等领域。
其中,分数阶混沌系统成为人们研究的热点,例如分数阶Lorenz系统[7]、分数阶Duffing系统[8]、分数阶Chen系统[9]、分数阶Chua系统[10]。