Efficiency of Micro Grid with Storage Battery in Reliability

合集下载

The future of energy Microgrids

The future of energy Microgrids

The future of energy Microgrids The future of energy microgrids holds great promise for addressing the challenges of our current energy landscape. As we face increasing concerns about climate change, energy security, and the need for more efficient and sustainable energy systems, microgrids offer a potential solution that can revolutionize the way we generate, distribute, and consume energy. This technology has the potential to empower communities, reduce reliance on centralized power grids, and pave the way for a more resilient and sustainable energy future. One of the key benefits of energy microgrids is their ability to enhance energy resilience. In the face of natural disasters, cyber-attacks, or other disruptions to the centralized grid, microgrids can continue to supply power to critical facilities such as hospitals, emergency response centers, and essential infrastructure. This capability is especially important as we witness an increase in extreme weather events and other threats to the stability of traditional power grids. By decentralizing energy generation and distribution, microgrids can help communities maintain essential services during times of crisis, ultimately saving lives and minimizing the impact of disasters. Furthermore, energy microgrids have the potential to promote energy independence and security. By enabling communities to generate their own power through renewable sources such as solar, wind, and hydroelectric energy, microgrids reduce reliance on imported fossil fuels and centralized power plants. This not only enhances energy security but also reduces the carbon footprint of communities, contributing to global efforts to mitigate climate change. Additionally, microgrids can enable more efficient energy use and storage, further enhancing the sustainability and reliability of local energy systems. From an economic perspective, energy microgrids offer the potential for cost savings and new business opportunities. By integrating advanced technologies such as smart meters, energy storage systems, and demand response capabilities, microgrids can optimize energy use, reduce waste, and lower energy costs for consumers. Moreover, the development and implementation of microgrid technologies can create new jobs and stimulate economic growth in the renewable energy sector. As the demand for clean energy solutions continues to rise, microgrids present an opportunity for innovation, entrepreneurship, and investment in the green economy. Despite thesepromising benefits, the widespread adoption of energy microgrids still faces several challenges. One of the primary barriers is the need for supportivepolicies and regulations that facilitate the integration of microgrid systems into existing energy infrastructure. Many regions lack clear guidelines for the deployment of microgrids, which can create uncertainty for investors and developers. Additionally, the upfront costs of implementing microgrid technologies can be prohibitive for some communities, especially in developing countries or underserved areas. Access to financing and incentives for microgrid projects is essential to ensure equitable access to the benefits of this technology. Moreover, the technical complexity of microgrid systems and the need for interoperability with existing energy infrastructure pose significant challenges for widespread adoption. Integrating diverse energy sources, managing fluctuating demand, and ensuring the stability and reliability of microgrid operations require advanced technical expertise and sophisticated control systems. Furthermore, cybersecurity concerns are paramount, as the interconnected nature of microgrid technologies introduces new vulnerabilities that must be addressed to safeguard againstpotential threats. In conclusion, the future of energy microgrids holds great potential to revolutionize the way we generate, distribute, and consume energy. By enhancing resilience, promoting energy independence, and creating economic opportunities, microgrids offer a compelling solution to the challenges of our current energy landscape. However, realizing this potential will require concerted efforts to address policy and regulatory barriers, improve access to financing,and overcome technical challenges. As we navigate the transition to a more sustainable and resilient energy future, energy microgrids have a crucial role to play in empowering communities and advancing global efforts to combat climate change.。

基于混合储能的孤岛微网VSG控制策略

基于混合储能的孤岛微网VSG控制策略
1 孤岛微网数学模型
1.1 孤岛微网结构 基于 VSG 控制的光储柴微网结构如图 1 所示。
光伏阵列、超级电容器和蓄电池分别经 Boost 变换 器和双向 DC/DC 变换器汇入直流母线,再经逆变 器与滤波器接入交流母线。三相逆变器采用 VSG 控制,使光储发电单元呈现出同步发电机特性,稳 定微网频率。光储发电单元与柴油发电机并联,通 过交流母线组成独立微网,向负载进行供电。
本文以光-储-柴组建的孤岛微网为例,提出一 种基于混合储能且不需改变惯性系数的 VSG 惯性 强化策略。首先,利用一阶高通滤波器提取母线负 荷与光伏功率差值的高频分量,并将其作为 VSG 额定功率的附加量;其次在 VSG 直流端接入混合 储能,利用蓄电池提供下垂功率,超级电容器提供 惯性功率和额外惯性功率提升微网惯性;然后引入 与超级电容器荷电状态和额外惯性功率相关的出力 系数,优化超级电容器在荷电状态不佳时的出力, 实现自适应控制。最后在不同负荷、光照强度和超 级电容器荷电状态下进行仿真,验证了本文控制策 略的正确性与有效性。
2. College of Electrical and Information Engineering, Hunan University, Changsha 410082, China)
Abstract: Changing the inertia coefficient in the Virtual Synchronous Generator (VSG) control strategy will cause system instability. Thus a hybrid energy storage-based control strategy without changing the inertia coefficient is proposed. First, the load or new energy fluctuation is regarded as the additional quantity of VSG rated power after passing through the first-order high pass filter as with a conventional VSG control strategy. Secondly, the battery connected to the DC terminal of the VSG is used to provide droop power, and the super capacitor provides virtual inertial power and additional inertial power. This improves the frequency dynamic characteristics of the microgrid. Then, the output coefficient related to the state of charge of the supercapacitor and the additional inertial power is introduced to realize the adaptive control of the supercapacitor and optimize the state of charge of the supercapacitor in the inertia strengthening process. Finally, an islanded microgrid with VSG as the inverter control interface is built in Matlab/Simulink. The effectiveness of the proposed strategy is verified by simulation under different loads, light intensities and supercapacitor states of charge.

家庭微网储能分组能量管理优化策略

家庭微网储能分组能量管理优化策略

理论算法2021.07家庭微网储能分组能量管理优化策略胡陈壮(上海电力大学电子信息与工程学院,上海,200090)摘要:在家庭能量管理系统中,可再生能源的发电功率具有不确定性和间断性,成为影响家庭能量优化调度的因素。

储能系统在优化过程中过多充放电次数也会增加储能折旧费用。

针对上述问题,文中提出一种储能分组能量管理优化策略。

根据可再生能源出力不确定部分和确定部分为储能系统配置充电部分和调度部分。

首先建立风力发电系统、光伏发电系统和储能系统模型,然后在此基础上搭建以每日用电费用最小为目标的家庭能量管理优化调度模型。

最后以上海市一住宅用电为例,通过改进遗传算法对模型求解。

仿真算例分析表明所提策略降低用电费用的同时可以减小可再生能源发电不确定性对能量优化调度的影响,具有一定的有效性和参考价值。

关键词:家庭能量管理;优化调度;储能分组;遗传算法Energy management optimization strategy of grouped energy storage system in household microgridHu Chenzhuang(School of Electrical Information and Engineering,Shanghai University of Elec t rie Power,Shanghai,200090) Abstract:In the home energy management system,the generation power of renewable energy has uncertainty and discontinuity,which has become a factor affecting the optimal scheduling of household energy.Excessive charging and discharging times in the optimization process of energy storage system will also increase the depreciation cost of energy storage.To solve the above problems,this paper proposes an energy management optimization strategy of energy storage group.According to the uncertain part and determined part of renewable energy output,the charging part and dispatching part are configured for the energy storage system.Firstly,the models of wind power generation system, photovoltaic power generation system and energy storage system are established,and then the optimal scheduling model of household energy managemerrt with the objective of minimizing daily electrieity cost is built.Finally,taking a residential power consumption in Shanghai as an example,the improved genetic algorithm is used to solve the model.Simulation resuIts show that the proposed strategy can reduce the impact of renewable energy generation uncertainty on energy optimal scheduling while reducing the power consumption cost,and has certain effectiveness and reference value・Keywords:Home energy management;Optimal scheduling;Energy storage grouping;Genetic algorithm0引言随着能源环境问题的日益加重和用电需求的不断增长,太阳能、风能等可再生能源具有清洁、丰富等特点,因而得到人们的充分关注与应用,成为如今能源发展必然趋势[1]。

211223194_孤岛微电网下储能与SVG的电压频率协调控制策略

211223194_孤岛微电网下储能与SVG的电压频率协调控制策略

电气传动2023年第53卷第5期ELECTRIC DRIVE 2023Vol.53No.5摘要:孤岛交流微电网中可再生能源的渗透率较高,其系统惯性较低,运行时一次调频主要依靠柴油机和储能等设备。

利用储能和静止无功发生器(SVG )的快速调节特性,提出了一种基于电压调节的孤岛微电网调频控制策略。

当孤岛微电网频率发生波动时,通过储能与SVG 的协调控制改变负荷节点电压,利用负荷电压静态特性快速调节系统频率。

在该控制策略中,储能和SVG 均采用下垂控制,不需要远程通信,只需检测负荷节点电压和频率信号。

该控制可减少微电网储能容量的配置,提高系统稳定性。

通过DIgSILENT/PowerFactory 仿真研究,对控制策略进行了验证。

关键词:微电网;下垂控制;SVG 控制;一次调频中图分类号:TM76文献标识码:ADOI :10.19457/j.1001-2095.dqcd23067Voltage -frequency Coordinated Control Strategy of Energy Storage and SVG in Island MicrogridZHAO Jingjing ,CHEN Linghan(School of Electrical Engineering ,Shanghai University of Electric Power ,Shanghai 200090,China )Abstract:The penetration rate of renewable energy in the island AC microgrid was relatively high ,and its system inertia was relatively low.The primary frequency regulation during island operation mainly relies on equipment such as diesel engines and energy ing the fast adjustment characteristics of energy storage and static var generator (SVG ),a frequency coordinated control strategy for island microgrid based on voltage adjustment was proposed.When the frequency of the island microgrid fluctuates ,the load node voltage was changed through the coordinated control of energy storage and SVG ,and the static characteristics of the load voltage was used to quickly adjust the system frequency.In this control strategy ,both energy storage and SVG adopt droop control ,no remote communication was required ,and only load node voltage and frequency signals need to be detected.This control can reduce the configuration of the energy storage capacity of the microgrid and improve system stability.Through DIgSILENT/PowerFactory simulation research ,the control strategy was verified.Key words:microgrid ;droop control ;SVG control ;primary frequency regulation作者简介:赵晶晶(1980—),女,博士,副教授,Email :*****************孤岛微电网下储能与SVG 的电压频率协调控制策略赵晶晶,陈凌汉(上海电力大学电气工程学院,上海200090)当今世界能源的需求量不断增加,传统化石能源的过度消耗,使可再生能源的发展变得更为重要[1]。

风光储互补微电网系统的设计及应用

风光储互补微电网系统的设计及应用

风光储互补微电网系统的设计及应用于慧(广西鑫盟工程咨询有限公司 广西南宁 530000)摘要:随着社会的高速发展与时代的不断更替,我国新能源技术得到了很大的发展空间。

为了切合社会发展的需求,该文主要对微电网系统进行了设计,可用于生产、教学、实验等。

系统包括有5个子系统,分别为光伏系统、风力系统、蓄电池储能系统、模拟输电线路系统、微电网监控调度系统。

除此之外,系统可能会受到外部环境因素的影响,如气候条件、当地降水量等,所以还配置了双馈异步风力发电系统与永磁同步发电的系统,能够对系统中参数进行实时的监控,实现程序运行的自由组合,从而产生更好的效果。

基于此,该文主要对电网中子系统与控系统进行阐述,能够帮助部分地区的电网进行合理规划,为扩大市场规模提供了相关经验。

关键词:风力发电 储能系统 监控 线路模拟 光伏发电中图分类号:TM92文献标识码:A 文章编号:1672-3791(2023)16-0094-04 Design and Application of the Wind-Photovotaic-StorageComplementary Microgrid SystemYU Hui(Guangxi Xinmeng Engineering Consulting Co., Ltd., Nanning, Guangxi Zhuang Autonomous Region,530000 China)Abstract:With the rapid development of society and the continuous replacement of the times, China's new energy technology has gained great development space. In order to meet the needs of social development, this paper mainly designs the microgrid system, which can be used in production, teaching, experiments, etc. There are five subsys‐tems in the system, namely the photovoltaic system, wind power system, battery energy storage system, analog transmission line system and microgrid monitoring and dispatching system. In addition, the system may be affected by external environmental factors, including climate conditions, local precipitation, etc., so it is also equipped with the doubly-fed asynchronous wind power generation system and the permanent magnet synchronous power gen‐eration system, which can monitor the parameters in the system in real time, realize the free combination of program operation, and then produce better results. Based on this, this paper mainly expounds the subsystem and monitoring system of the power grid, which can help the power grid in some regions to carry out reasonable planning, and also provide relevant experience for expanding the market scale.Key Words: Wind power generation; Energy storage system; Monitor; Line simulation; Photovoltaic power随着新能源发电规模的逐步扩大,许多电网的系统也变得更加完善,其中的监控系统则是对相关参数进行检测[1]。

[理学]微电网论文:微电网PQ控制Vf控制运行模式切换超级电容器储能

[理学]微电网论文:微电网PQ控制Vf控制运行模式切换超级电容器储能

微电网论文:基于储能装置的微网运行特性研究【中文摘要】由于科学技术的发展和环境保护的要求,包括微型燃气轮机、光伏电池、燃料电池、和风力发电等各种分布式能源开始进入人们的视野。

然而分布式能源尽管优势明显但也暴露出许多问题,例如单机接入成本高、控制困难等问题。

微电网技术的出现很好的解决了这一问题。

微电网将分布式电源、储能装置、控制装置、以及负荷作为一个整体,既可作为一个特殊负荷与电网相连,也可以看作是一个小型电网孤网运行。

微电网作为超高压、远距离、大电网供电模式的补充,代表着电力系统新的发展方向。

微电网进入孤岛方式时仍要保证负荷稳定运行。

同时由于微网内不同微源的特点,因此需要采用不同的控制方法。

并网运行时,微电源只要满足功率平衡,大电网负责电压和频率的平衡;当出现故障时,微网由并网方式转换成孤岛模式,由微汽轮机做主控制单元负责电压和频率的调节,光伏电池仍然采用PQ控制。

由于光伏发电具有波动性,对并网和孤岛运行都有影响,因此采用光伏电池和超级电容器混合搭配的方式来抑制电压波动和功率波动。

最后,通过对微电网在孤网和并网运行模式之间的切换以及孤网模式下负荷突变、短路故障等情况进行动态仿真与分析,验证了对不同微源所采用的综合控制策略可以较好地保证微电网系统平...【英文摘要】Due to the technology development and environment protection, some distributed energy resources(DER), such as micro turbines, photovoltaic, fuel cells and wind power, have entered people’s field of vision. However, application of individual distributed generators can cause as many problems as it may solve, which including problem of high access costs and single DER control. The micro-grid is the most effective way to make e use of DER. Micro-grid is a small-sized grid that integrates a number of micro-sour...【关键词】微电网 PQ控制 V/f控制运行模式切换超级电容器储能【英文关键词】micro-grid PQ control V/f control Operation mode transform super capacitor energy storage【索购全文】联系Q1:138113721 Q2:139938848【目录】基于储能装置的微网运行特性研究摘要5-6Abstract6 1 绪论9-18 1.1 课题的研究背景及意义9-11 1.2 微网国内外研究现状11-13 1.2.1 北美的微网研究11-12 1.2.2 欧洲的微网研究12 1.2.3 日本的微网研究12 1.2.4 国内的微网研究12-13 1.3 超级电容器的发展概况13-14 1.4 超级电容器在微网中的应用14-16 1.4.1 提供短时供电15 1.4.2 作为能量缓冲装置15 1.4.3 改善微电网的电能质量15-16 1.4.4 优化微电源的运行16 1.5 论文的主要研究内容16-18 2 微电网概念及组成18-29 2.1微电网概念18-19 2.2 微电网中微源构成19-25 2.2.1 风力发电20-21 2.2.2 光伏电池21-23 2.2.3 微型燃气轮机23-24 2.2.4 燃料电池24-25 2.3 微网储能设备25-29 2.3.1 飞轮储能25-26 2.3.2 超级电容器储能26-27 2.3.3 蓄电池储能27-29 3 微电网运行与控制29-42 3.1 微电网运行特点29-30 3.2 微电网控制策略分析30-31 3.3 微电网控制策略31-35 3.3.1 主从控制法31-33 3.3.2 对等控制法33-35 3.4 微源逆变器的控制方法35-38 3.4.1 恒功率控制36-37 3.4.2 恒压恒频控制37 3.4.3 下垂控制37-38 3.5 逆变器的电路结构38-39 3.5.1 逆变器电路38 3.5.2 Park 变换38-39 3.6 超级电容器储能系统的结构与工作原理39-42 4 微电网模型的建立42-51 4.1 仿真软件介绍42-43 4.2 微型燃气轮机模型的建立43-44 4.3 光伏电池仿真模型44-46 4.4 微电网运行仿真及分析46-51结论51-52参考文献52-56在学研究成果56-57致谢57。

压缩二氧化碳储能英语

压缩二氧化碳储能英语

压缩二氧化碳储能英语一、单词1.pression [kəmˈpreʃn]- 释义:n. 压缩;压榨;压抑- 用法:常与of连用,表示“……的压缩”,如thepression of gas (气体的压缩)。

- 例句:Thepression of the spring stores potential energy.(弹簧的压缩储存了势能。

)2. carbon dioxide [ˈkɑːbən daɪˈɒksaɪd]- 释义:二氧化碳- 用法:可作主语、宾语等,在科学、环境等话题中常用。

- 例句:Carbon dioxide is a greenhouse gas.(二氧化碳是一种温室气体。

)3. energy storage [ˈenədʒi ˈstɔːrɪdʒ]- 释义:储能- 用法:可作名词短语,如The development of energy storage technology is very important.(储能技术的发展非常重要。

) - 例句:Energy storage can help to balance the power grid.(储能有助于平衡电网。

)二、短语1.pressed carbon dioxide- 释义:压缩二氧化碳- 用法:可作主语、宾语等,例如:Compressed carbon dioxide can be used for various purposes.(压缩二氧化碳可用于多种用途。

) - 例句:The container is filled withpressed carbon dioxide.(容器里装满了压缩二氧化碳。

)2. carbon dioxidepression system- 释义:二氧化碳压缩系统- 用法:在描述二氧化碳压缩相关设备或技术时使用,如The carbon dioxidepression system needs regular maintenance.(二氧化碳压缩系统需要定期维护。

新型电网-微电网(Microgrid)研究综述 - 副本

新型电网-微电网(Microgrid)研究综述 - 副本

一76一继电器术(DistributedStorage),通过储能装置储存分布式电源的多余的能量,如超导线圈、储能电容器及储能能力巨大的超级电容器和飞轮等等。

通过对上述技术展开研究得到的系列成果,并结合电力系统用户对电能质量(PowerQuality)的要求和电力系统发展趋势,逐步形成了将上述技术综合在一起而形成的特殊电网形式一微型电网(MicroGrid)【6】。

目前,国际上对微型电网的定义各不相同【6 ̄1…。

美国电气可靠性技术解决方案联合会(CERTS—ConsortiumfbrE1ectricReliabilityTechn0109ySolutions)给出的定义为:微电网是一种由负荷和微型电源共同组成系统,它可同时提供电能和热量;微电网内部的电源主要由电力电子器件负责能量的转换,并提供必需的控制;微电网相对于外部大电网表现为单一的受控单元,并可同时满足用户对电能质量和供电安全等的要求。

欧盟微电网项目(EuropeaIlCo舢【11issionProjectMicro—grids)给出的定义是:利用一次能源;使用微型电源,分为不可控、部分可控和全控三种,并可冷、热、电三联供;配有储能装置;使用电力电子装置进行能量调节。

美国威斯康辛麦迪逊分校(UIliversitvofWisconsin—Madison)的R.H.hsseter给出的概念是:微电网是一个由负载和微型电源组成的独立可控系统,对当地提供电能和热能。

这种概念提供了一个新的模型来描述微电网的操作;微电网可被看作在电网中一个可控的单元,它可以在数秒钟内反应来满足外部输配电网络的需求;对用户来说,微电网可以满足他们特定的需求:增加本地可靠性,降低馈线损耗,保持本地电压,通过利用余热提供更高的效率,保证电压降的修正或者提供不问断电源。

图1是威斯康新大学新能源实验室的微电网结构图。

综合以上,并结合我国电网的实际状况,本文可以得出微电网的定义:能量来源主要为可再生能源;发电系统类型可为微型燃气轮机(Micro—Turbine)、内燃机(GasEngine)、燃料电池(FuelCeU)、太阳能电池(PVP甜lel)、风力发电机(WindGenerator)、生物质能(BiomassEnergy)等;系统容量为20kw~10Mw;网内的用户配电电压等级为380V,或者包括10.5kV;如与外部电网进行能量交换,电压等级由微电网的具体应用等情况而定。

虚拟同步发电机并机黑启动技术研究

虚拟同步发电机并机黑启动技术研究

ELECTRIC DRIVE 2024Vol.54No.2电气传动2024年第54卷第2期基金项目:陕西省教育厅一般专项科研计划项目(22JK0366)作者简介:杨帆(1992—),男,硕士,工程师,主要研究方向为储能控制算法,Email :虚拟同步发电机并机黑启动技术研究杨帆1,王换民2(1.上海弘正新能源科技有限公司,上海201400;2.商洛学院电子信息与电气工程学院,陕西商洛726000)摘要:随着电力电子技术的发展,具有虚拟同步发电机(VSG )功能的微网型储能逆变器也得到了广泛应用。

在虚拟同步发电机技术中由于惯性调频环及调压环的存在,逆变器在并网时可自主实现电网调频调压功能,离网时也可实现无通信并机。

但是在微网系统中由于负载和变压器的接入,逆变器离网时并机黑启动技术是亟待解决的问题之一。

通过理论仿真分析,提出了一种工频同步的软启动方法,可实现平滑的母线电压建立。

在软启动过程中两台VSG 并机功率均衡,实现了整个系统黑启动的完成。

最后通过实验分析,验证了VSG 并机黑启动控制策略的有效性。

关键词:工频同步;黑启动;虚拟同步发电机中图分类号:TM464文献标识码:ADOI :10.19457/j.1001-2095.dqcd24673Implementation of Parallel Black Start of Virtual Synchronous GeneratorYANG Fan 1,WANG Huanmin 2(1.Shanghai Hongzheng New Energy Technology Corporation ,Shanghai 201400,China ;2.School of Electronic Information and Electrical Engineering ,Shangluo University ,Shangluo 726000,Shaanxi ,China )Abstract :With the development of power electronics technology ,microgrid type energy storage inverters with virtual synchronous generator (VSG )function have also been widely used.In VSG technology ,due to the presence of inertial frequency regulation loop and voltage regulation loop ,the inverter can realize the grid frequency regulation and voltage regulation function independently when it is connected to the grid ,and it can realize the parallel power without communication when it is off-grid.However ,in the microgrid system ,due to the access of load and transformer ,the parallel black start technology is one of the urgent problems to be solved when the inverter is off-grid.Through theoretical simulation analysis ,an industrial frequency synchronous soft start method was proposed ,which can realize smooth bus voltage establishment.The parallel power equalization of two VSGs during the soft start process achieved the completion of the black start of the whole system.Finally ,the effectiveness of the parallel VSG black start control strategy was verified through experimental analysis.Key words :frequency synchronization ;black start ;virtual synchronous generator (VSG )虚拟同步发电机(VSG )控制思想[1-2]的发展,使得风电、光伏、储能等电力电子变换器呈现出同步发电机的运行特性,随着新能源入网占比的逐年递增,虚拟同步发电机技术是提高电力系统稳定性的一种有效方法。

基于微服务架构的众包图像数据集标注系统

基于微服务架构的众包图像数据集标注系统

基于微服务架构的众包图像数据集标注系统①袁晓晨1, 张卫山1, 高绍姝1, 时 斌2, 赵永俊2, 王 冶3, 安云云41(中国石油大学(华东) 计算机科学与技术学院, 青岛 266580)2(青岛海尔空调电子有限公司 电控模块开发部, 青岛 266101)3(解放军9144部队, 青岛 266102)4(国网山东省电力公司 青岛市黄岛区供电公司, 青岛 266500)通讯作者: 张卫山摘 要: 深度学习在图像识别领域凸显出了优势, 而在深度学习图像识别模型训练的准备阶段, 制备图像数据集需要人工将图片上的信息进行标注. 这一准备过程往往需要耗费大量人力成本与时间成本. 为了提升数据制备阶段的工作效率, 从而加速深度学习模型的生成与迭代, 提出了一种基于微服务架构的多人协作众包式图像数据集标注系统. 通过将繁重的标注任务划分为不同的小任务, 使更多的人能够参与并协同完成数据标定. 通过引入对象存储机制并采用微服务架构, 提升了系统性能, 在开发阶段使用了基于Gitlab 的持续集成与持续部署, 实现了系统的快速迭代与部署, 提升了微服务系统在开发过程中的集成效率.关键词: 微服务; Spring Cloud; 持续集成; 持续部署; 图像标注引用格式: 袁晓晨,张卫山,高绍姝,时斌,赵永俊,王冶,安云云.基于微服务架构的众包图像数据集标注系统.计算机系统应用,2021,30(5):83–91./1003-3254/7900.htmlImage Dataset Annotation System in Crowdsourcing Based on Microservice ArchitectureYUAN Xiao-Chen 1, ZHANG Wei-Shan 1, GAO Shao-Shu 1, SHI Bin 2, ZHAO Yong-Jun 2, WANG Ye 3, AN Yun-Yun 41(College of Computer Science and Technology, China University of Petroleum, Qingdao 266580, China)2(Electronic Control Module Development Department, Haier Air Conditioning Electronic Ltd. (Qingdao), Qingdao 266101, China)3(No. 9144 Troops of PLA, Qingdao 266102, China)4(Qingdao Huangdao District Power Supply Company, State Grid Shandong Electric Power Company, Qingdao 266500, China)Abstract : Deep learning has shown visible advantages in the artificial intelligence-based image classification. It usually costs plenty of time on manual information annotation for preparing image datasets. Then this study proposes an online collaborative system for image dataset annotation based on microservice architecture to improve the efficiency of annotating datasets and thus to accelerate the generation and iteration of deep learning models and applications. More users can join for image annotation after the heavy annotation task is divided into smaller ones. Besides, the system performance has been improved by introducing an object storage system and microservice architecture, and the integration efficiency of the system in the development progress has been enhanced by continuous integration and deployment.Key words : microservice; Spring Cloud; continuous integration; continuous deployment; image annotation计算机系统应用 ISSN 1003-3254, CODEN CSAOBNE-mail: Computer Systems & Applications,2021,30(5):83−91 [doi: 10.15888/ki.csa.007900] ©中国科学院软件研究所版权所有.Tel: +86-10-62661041① 基金项目: 国家自然科学基金(62072469); 国家重点科研计划(2018YFE0116700); 山东省自然科学基金(ZR2019MF049); 中央高校基础研究基金(2015020031); 西海岸人工智能技术创新中心建设专项(2019-1-5, 2019-1-6); 上海可信工业控制平台开放项目(TICPSH202003015-ZC)Foundation item: National Natural Science Foundation of China (62072469); National Key R&D Program (2018yfe0116700); Shandong Natural Science Foundation (ZR2019MF049); Basic Research Fund of Central University (2015020031); West Coast Artificial Intelligence Technology Innovation Center (2019-1-5, 2019-1-6); the Opening Project of Shanghai Trusted Industrial Control Platform (TICPSH202003015-ZC)收稿时间: 2020-09-10; 修改时间: 2020-10-09; 采用时间: 2020-10-13; csa 在线出版时间: 2021-04-28计算机视觉图像识别是人工智能的重要应用, 广泛应用在工业、医学、军事、教育、商业、体育、安防检测等行业与领域中. 机器学习, 尤其是深度学习展现出了针对图像识别领域优秀的识别性能. 而机器学习本身需要建立在大量的带有指导意义的既有数据集基础之上. 在进行深度学习模型训练流程中, 往往需要针对海量图片进行人工数据标注, 繁重的图像标注任务增添了大量时间成本.传统的图像标注工具, 如表 1所示, 大多以单机运行的传统单体式系统架构为主, 运行在单机之上, 同一时间同一系统运行实例上只允许一个用户对本地资源进行图像标定. 当使用传统图像标注工具进行协作标定时, 需要用户手动进行图像集的分组并拷贝至协作组员的工作站. 协作组员各自完成任务后仍需要用户自行合并制作数据集. 由于缺少集中化的图像与工作流, 导致传统单机标注工具在协作场景下存在大量的文件与数据传输, 在海量图像文件的压缩、打包与传输过程中产生了大量的时间与人工成本.表1 传统图像标注工具对比图像标注工具实例系统架构兼容系统开源与付费模式协作模式支持的图像标注类型导出格式网址LabelImg单机(PyQt5)全平台开源免费可独立部署手动资源分组打包后独立运行协作标注矩形PASCAL、VOC、ImageNethttps:///tzutalin/labelImgLabelMe单机(PyQt5)全平台开源免费多边形、矩形、圆形、线标注、关键点COCOhttps:///wkentaro/labelmeVIA-VGG Image Annotator Web本地静态页面全平台开源免费多边形、矩形JSON、CSV、COCOhttps:///PixelAnnotaion单机(Qt5)Windows、MacOS、Linux(自行编译)开源免费填充区域JSONhttps:///abreheret/PixelAnnotationToolRectLabel单机MacOS 不开源付费$55多边形、填充区域、矩形PASCAL、VOC、YOLO、COCO、CSVhttps:///OpenCV/CVAT Web本地部署(Django)全平台开源免费单实例不支持直接协作, 需要手动对资源分组打包独立部署多实例图像分类、多边形工具预定义XML格式https:///opencv/cvatVoTT Web本地部署(NodeJs)全平台开源免费矩形Azure自有格式、CNTK、Pascal、VOC、TFRecords、VOTT(JSON)、CSVhttps:///microsoft/VoTTLabelBox纯前端库全平台不开源需要引用公网库资源需要额外系统支持图像分类、多边形、矩形CSV、JSON(由官方后端提供)https:///Labelbox/Labelbox)为了解决机器学习图像识别训练的各种前置准备工作费时费力的问题, 本文提出并设计实现了一种基于Spring Cloud、面向机器学习模型训练的协作式图像数据管理与标注平台, 通过众包任务的方式, 优化图像标注流程, 设计实现了面向海量图像的存储、标定集管理、图像标定任务管理等功能模块. 系统采用微服务架构, 将系统各个部分进行解耦[1], 实现服务注册与发现、负载均衡、容错处理, 提升了系统的高可用性、可维护性与可扩展性. 基于GitLab-CICD实现了微服务的增量更新、持续部署与灰度更新. 通过将海量图像标定数据进行分组划分, 简化数据标注过程中的操作, 降低了数据标注人员的时间成本, 提升了数据标注工作效率.1 概述1.1 众包任务模式众包任务模式指的是将本应当由单一机构或个人执行的工作内容在公开或非公开网络上以公开的方式外包给特定或非特定用户[2]. 通过采用众包任务的模式,将庞大的图像标注任务切分为小的任务, 能够有效提计算机系统应用2021 年 第 30 卷 第 5 期升标定的速度与质量, 从而加速深度学习模型训练流程.1.2 微服务架构微服务[3]是一系列功能简单、互相之间采用轻量级通信协议协同工作的功能区块. 每个小的功能区块具有高内聚、低耦合的特点, 能够独立自主地运行.微服务架构是一种架构模式, 它是SOA架构(Service Of Application)的进一步发扬, 通过将庞大的系统根据业务边界细粒度地拆分为小的服务模块. 每个小的服务模块具有高内聚、低耦合的特点, 均能够独立地运行. 每个模块可以由不同的团队进行开发[4].微服务架构拥有着许多传统单体式应用所不具备的优势[5]:① 复杂度可控: 通过将复杂的系统细粒度地拆分为小体积、业务简单的模块, 降低了每个模块的开发难度与复杂度. 各个模块间互相解耦[6], 当需求变更时,只需要修改对应的模块即可实现系统整体功能上的变更, 无需考虑其他模块可能因为接口实现的变更而无法正常工作的问题.② 团队协作效率高: 当确定系统各个微服务接口定义后即可分组并行开发不同的微服务模块, 从而提升了整体开发效率. 同时不同微服务模块无需集成为一个整体, 只需要能够互相调用即可保证系统的正常运行, 避免了单体式应用集成过程中出现的组件源码不兼容, 降低了集成成本.③ 独立部署: 每个小模块都能够独立地运行, 不会因为其他服务故障而受到影响. 通过微服务调度系统的支持, 能够实现快速部署在单机或集群上, 从而充分利用计算性能.④ 维护难度低: 配合持续集成与持续部署(CI/CD),即可实现灰度发布与服务热更新, 无需系统全部停机或重启即可完成功能更新.⑤ 多种技术允许共存: 微服务独立部署使得微服务间只需要约定技术、语言无关的RPC协议即可实现不同技术、语言实现的微服务共存并配合工作[7].⑥ 系统稳定性高: 微服务架构的系统拥有熔断机制, 当系统部分组件失效时能够及时阻止故障船体, 从而避免系统发生雪崩式功能失效, 提升了系统的稳定性.1.3 对象存储系统基于对象的云存储即对象存储(object storage)是近几年逐渐流行的一个新兴且切实可行的大规模存储方案[8], 使用较为简便的方法实现数据存储, 即充分利用已有的存储组件、网络技术和处理技术, 使系统拥有较好的可扩展性以及高吞吐量[9]. 对象存储思想的核心为对象, 每个对象都有唯一的标识[10]. 对象存储将文件划分为一个个对象, 为用户提供了统一的存储空间,从而能更好的对文件进行访问控制和存储管理[11]. 这些对象被分布在整个集群之中, 为保障数据安全、防止数据丢失将每一个对象多重备份复制到多个设备上.对象存储系统将数据块列表映射为对象列表, 将各类数据块简化成为一个个对象来进行管理[12], 极大改善了系统的可伸缩性, 可以轻易实现海量数据的管理.2 系统设计2.1 需求分析传统的数据集标注往往在单终端中进行, 大量的图像需要由一个部门或者个人完成. 使用传统的标注工具进行多人协作式图像标注时, 在图像整理上浪费了大量时间. 协作前需要人工将大量图像打包分组, 协作后需要将图像与标定数据回收合并为同一个数据集合. 常见的文件系统针对海量小文件的处理性能往往不及少量大文件, 导致人工打包也是一项耗时的工作.针对以上痛点, 图像标注系统主要有以下需求:① 图像数据托管: 为避免文件系统直接频繁操作大量小文件, 从而节省任务分配数据打包的时间, 图像数据应当由专门的机制进行托管, 必要时以图像为基本单位提供图像及其元数据检索服务.② 任务划分与管理: 为实现众包模式的图像标注,需要将包含大量图像的任务拆分为小任务, 交由不同的个人进行标注.③ 支持多种标注模式: 系统应当支持机器视觉训练常用的关键点、矩形、多边形标注模式.④ 支持多种导出格式: 针对主流深度学习源码所接受的格式, 系统应当具备良好的扩展性以支持更多的深度学习框架.⑤ 数据统计: 任务发起者应当能够看到子任务的标注进度、标签数量等统计信息, 从而有针对性地决定是否需要增加或者调整图像组成.⑥ 支撑模块需求: 除此之外, 系统应当具有基本的用户与权限控制功能.2.2 微服务划分与系统架构设计① 系统微服务划分根据系统需求分析结果, 将系统划分为4个功能2021 年 第 30 卷 第 5 期计算机系统应用性微服务和3个架构支撑服务: 图像对象存储服务、图像集合管理服务、标定集合管理服务、数据输入输出服务、用户及授权管理和服务网关.图像对象存储服务用于面向海量图片的对象存储,将图片统一以相同规则生成的不重复定位符作为索引,避免图片重复存储与图片文件名重复冲突的问题.图像集合管理服务用于将离散的图片在逻辑上组成一个集合, 作为系统中图片操作的基本单位.标定集合管理服务用于管理和存储图像集对应的标定数据, 同时提供标定任务的划分与分配.数据导入导出服务主要用于图像、标签数据的解包导入与打包导出. 提供常见的压缩与视频格式的解析与常见数据集格式的导出等功能.② 系统架构设计系统整体采用分层式结构, 如图 1所示. 为了开发过程中能够更明确的分工, 其中服务层按照微服务的思想进行拆分. 各层主要包含内容如下:持久化层: 为了方便使用微服务编排框架进行部署, 所有有状态的服务均从整体架构中分离整合在持久化层中. 其中包括用于对象数据存储的NoSQL 数据库; 用于关系型信息数据存储的SQL 数据库; 用于全局数据缓存的内存NoSQL 数据库; 用于全局消息同步的消息队列中间件.微服务层: 包含了业务逻辑微服务群和架构支撑服务群两部分. 架构支撑服务群中包含了微服务架构必需的注册发现中心、日志监控、配置中心等基础微服务. 业务逻辑微服务群提供了实现业务逻辑的相关微服务, 包括图像对象存储微服务、图像集管理微服务、标定管理微服务和数据导入导出微服务.网关层: 网关层包含了由Spring Gateway 实现的API 网关微服务、前端站点微服务和Nginx 总代理微服务, 提供了系统接入入口.CI/CD 支撑部分: 系统源代码采用了Gitlab 私服进行托管, Gitlab 也提供了对持续集成的支持, 故直接采用Gitlab-CICD 实现系统开发过程中的持续集成与部署.系统采用现有完整的微服务方案, 使用服务发现实现松散的服务间耦合, 采用声明式RPC 客户端实现微服务间的互相调用, 使用统一网关代理微服务作为系统入口, 添加负载均衡机制以扩展系统负载容量, 使用OAuth2开放认证协议作为认证机制.系统部署业务逻辑服务数据导入导出微服务图像对象存储微服务持久化层SSDB 数据库MySQL 数据库Redis 缓存RabbitMQ 消息队列图像集管理微服务标定管理微服务数据导入导出微服务微服务支撑前端微服务Nginx 总代理Spring gateway API 网关Eureka 注册中心Config server 配置中心日志&监控自动化部署CI/CD 支撑Gitlab 代码仓库Gitlab-runner 宿主机代码提交开发团队图1 图像标注系统架构3 持续集成、持续部署与微服务架构实现3.1 微服务注册与发现在微服务架构的系统中, 为了实现微服务间既有松散的耦合度, 又能够互相访问, 微服务注册与发现机制是一种常用方法[13]. 通过由微服务自行将自身的信息主动注册至注册中心的方式, 使得其他微服务可以计算机系统应用2021 年 第 30 卷 第 5 期通过查询注册中心注册记录的方式间接地发现其存在并获得访问相关参数. 注册中心是注册与发现中最重要的一部分, 在Spring Cloud 微服务套件中, 提供了注册中心的一种原生实现和两种接入实现:① Eureka Server: 由Netflix 开发的基于原生Java 的开源注册中心实现, Spring Cloud 套件给出了基于Spring Boot 快速构建注册中心的方案.② Spring Cloud Consul: Consul 是由HashiCorp 公司开发的一种服务网格解决方案, 提供具有服务发现,配置和分段功能的全功能控制平面[14]. 通过添加org.springframework.cloud.spring-cloud-starter-consul-all 依赖, 同样可以实现类似的注册发现功能.Spring Cloud Zookeeper: Apache ZooKeeper 是一项集中式服务, 用于维护配置信息, 命名, 提供分布式同步和提供组服务[15]. 基于Spring Cloud 的微服务可以通过添加org.springframework.cloud.spring-cloud-starter-zookeeper-all 依赖实现基于ZooKeeper 的注册发现.由于后两种注册发现的方案需要单独部署第三方的应用实例来支撑微服务注册与发现的功能, 故选择可自行构建的Eureka Server 方案.Eureka 的注册与发现方案整体流程如图 2所示,主要分为两部分.Eureka 注册中心微服务1微服务2DCFeign 3. 调用服务1. 注册2. 获取实例列表DCDC: Discovery clientFeign图2 基于Eureka 的注册发现实现流程第一部分为Eureka 注册中心. Spring 提供了一套项目生成工具Spring Initializer [16], 允许开发人员直接通过可视化配置直接生成Spring Boot 项目, 这里借用Spring Initializer 可以直接生成Eureka Server 的项目.如图 3所示, 在Spring Initializer 中输入项目信息并选中Eureka Server 依赖, 点击Generate 即可得到初始项目模板. 在Spring Boot 启动类上添加@EnableEurekaServer 参数即可开启项目依赖中的Eureka Server.图3 使用Spring Initializer 生成Eureka Server 实例第二部分即业务微服务中的Discovery Client. 在需要通过服务发现感知其他服务实例的微服务中, 增加org.springframework.cloud.spring-cloud-starter-netflix-eureka-client 依赖, 并在项目引导类上添加@Enable-DiscoveryClient 注解, 即可为Feign 客户端添加服务发现的功能.完成其他业务微服务开发后, 同时启动各个微服务, 可通过登录Eureka 监控页面看到各个微服务的注册情况, 如图4所示.图4 通过Eureka Server 查看服务发现注册状态3.2 微服务接口网关与负载均衡相较于传统的单体式应用, 微服务架构的应用更接近于单体式应用站群. 微服务架构应用为了实现从外部表现为与单体式应用类似的形式, 需要一个微服务作为接口网关, 同时需要提供负载均衡的特性.对于API 网关, Spring Cloud 提供了两套方案:① Zuul: Zuul 来源于Netflix 开源的微服务架构套件. Zuul 底层采用了Tomcat Embeded 版本作为HTTP 支撑层.② Spring Cloud Gateway: Spring Cloud Gateway 是由Spring 项目组基于其自研Web 框架WebFlux 实现的API 网关. Spring Cloud Gateway 是Spring Cloud 去Netflix 进程中非常重要的一个项目, 旨在替代Zuul 成为未来的Spring Cloud 框架下的API 网关组件.WebFlux 底层为直接使用Netty 等高性能非阻塞2021 年 第 30 卷 第 5 期计算机系统应用服务器, 相较于采用Servlet架构的Tomcat(Embeded)性能上略胜一筹 [17]. 考虑到未来Spring项目组开发的方向与性能预期, 采用Spring Cloud Gateway作为系统API网关.Spring Cloud Gateway同样可以使用Spring Initializer 进行项目初始化, 初始化完成后可以通过在application. yml中的spring.cloud.gateway.route字段中配置接口路径与微服务的映射.代码1. Spring Cloud Gateway路由配置spring: application: name: application-gateway cloud: gateway: routes: # 用户服务 - id: user-service uri: lb://user-service predicates: - > Path= /oauth/**, /user/**, /role/** # 其他服务 # …对于负载均衡, Spring Cloud框架提供了Ribbon 组件. Ribbon是一款客户端侧负载均衡器, 它可以自动从Discovery Client中获取微服务实力列表, 应用常见的负载均衡算法实现在同一服务的多个冗余实例上的负载均衡[18]. 通过添加org.springframework.cloud. spring-cloud-starter-netflix-ribbon依赖即可在服务调用侧引入负载均衡组件.3.3 微服务熔断机制微服务之间往往避免不了互相调用对方的服务.当进行同步接口调用时, 被调用方宕机、接口调用超时往往会引起调用方的异常. 为了避免因微服务之间的依赖关系而出现大面积故障, 调用方服务应当感知被调用方的异常并作出防御性动作, 防止故障继续的蔓延.在Spring Cloud框架中, 提供了Hystrix组件. Hystrix能够在被调用微服务出现异常时及时熔断, 触发调用方的异常处理流程, 防止调用方产生异常. Hystrix 组件可以通过在微服务中添加org.springframework.cloud.spring-cloud-starter-netflix-hystrix依赖来添加至微服务中. 在微服务启动类上添加@EnableHystrix注解即可启用熔断器.3.4 基于SSDB的图像对象存储系统在深度学习图像识别模型训练过程中往往需要准备大量图像, 而实现存储并快速索引图像数据就是标注系统性能提升的关键点.SSDB是一个高性能的支持丰富数据结构的NoSQL 数据库, 其底层实现为Google的高性能键值对数据库LevelDB.首先是一个高性能SSDB集群, SSDB采用的是与Redis相同的网络通信实现, 故可以采用Redis集群常用的TwemProxy代理实现如图 5所示的对象存储架构.微服务集群SSDB 实例群SSDB 实例SSDB 实例SSDB 实例SSDB 实例SSDB 实例SSDB 实例SSDB 实例SSDB 实例图像存储微服务图像存储微服务图像存储微服务图像存储微服务TwemProxy TwemProxy TwemProxy···图5 SSDB集群架构图5上部分是图像对象存储微服务. 图像存储微服务使用图像原始数据经哈希运算得到的索引号作为图像在对象系统中的唯一索引. 每一张图像采用一个哈希表进行存储, 在哈希表中额外增加有关图像内容相关的元信息, 对于重复上传的内容采取增加元信息中记录的索引数量而不重复存储.3.5 基于Gitlab-CICD的持续集成与部署在传统的单体式应用中, 只需要编译一次即可得到可运行的产物, 而在微服务架构应用中, 因为划分为了多个实例, 这种类似于“站群”的系统往往需要编译多个“单体式应用”并封装为Docker镜像进行部署. 为了减少编译部署阶段的工作量, 使用Gitlab-CICD实现全自动化的编译、测试与部署.首先是准备一个Gitlab实例, 可以使用Gitlab官方站点或者建立开源的Gitlab-CE实例, 本文不再赘述计算机系统应用2021 年 第 30 卷 第 5 期有关Gitlab 实例搭建的内容.系统采用了Docker Swarm 作为微服务编排框架,使用Harbor 作为Docker Swarm 集群的私有镜像源.Gitlab-CICD 中另一个重要组成部分便是Gitlab-Runner. Gitlab-Runner 是实际执行持续集成任务的主体. Gitlab-Runner 支持多种部署方式, 本项目选择了基于裸机系统的编译环境, 方便做一些特殊的环境配置.Gitlab-CICD 执行流程如图 6所示, 主要环节包括:① 代码提交: 在各个微服务中建立Git 仓库, 并将仓库托管至Gitlab. 当微服务代码得到更新并推送提交至Gitlab 时, Gitlab 会自动检查仓库目标分支中的.gitlab-ci.yml 文件中的配置, 并在流水线(pipeline)中添加持续集成的任务.② CICD 任务的执行: Gitlab-Runner 定时向Gitlab 的流水线队列请求任务, 当Gitlab-Runner 得到任务后在其宿主机上执行配置文件中定义的脚本. 脚本中定义的动作完成微服务的编译、测试、镜像打包与提交、灰度更新.Docker swarm集群Harbor镜像仓库⑤ 拉取更新后的镜像④ 触发灰度更新② 获取任务开发主机① 提交代码到Git 仓库③ 推送镜像GitlabGitlab-Runner图6 Gitlab-CICD 执行流程以图像管理服务为例, 编写了如代码2所示的gitlab-CICD.yml 配置文件.代码2. Gitlab 持续集成与持续部署配置stages: - 测试 - 构建 - 部署测试: stage: 测试 only: - master tags: - dlp script: - mvn test 构建: stage: 构建 only: - master tags: - dlp script: - mvn package -Dmaven.test.skip=true - docker build -t ${DOCKER_REGISTRY}/dlp/${SERVICE_NAME}:latest . - docker push ${DOCKER_REGISTRY}/dlp/${SERVICE_NAME}:latest 部署: stage: 部署 only: - master tags: - dlp script: - docker -H ${DOCKER_SWARM} service update --image ${DOCKER_REGISTRY}/dlp/${SERVICE_NAME}:latest${DOCKER_SWARM_ STACK_NAME}_${SERVICE_NAME}如代码2所示, CICD 一次任务将分为3个阶段:构建、测试与部署. 其中测试阶段由Maven 进行编译并执行系统中的单元测试; 构建阶段直接使用Maven 进行服务器端编译, 随后使用Gitlab-Runner 宿主机上的docker 构建镜像并推送至Harbor; 部署阶段通过暴露Docker Swarm 中的Manager 节点上dockerd 的2375端口, 实现从Gitlab-Runner 宿主机直接控制集群并触发微服务镜像更新. 运行效果如图 7所示.通过在各个微服务实例中分别修改相同逻辑后分别以手动部署与持续集成在多节点集群上部署并统计用时, 得到逻辑更新至生产环境的耗时如表 2所示.图7 Gitlab-CICD 执行效果表2 手动部署与CICD 部署效率对比微服务实例手动部署耗时CICD 耗时图像存储(5实例)30 min 12 s 3 min 8 s 图像集管理(3实例)21 min 05 s 1 min 26 s 标定集管理(3实例)20 min 40 s 1 min 6 s 数据导入导出(1实例)13 min 23 s 6 min 42 s 其他微服务支撑服务16 min 41 s1 min 5 s2021 年 第 30 卷 第 5 期计算机系统应用由表 2可知, CICD 有效减少了系统集成过程中的集成与部署时间成本, 提升了系统逻辑变更同步至生产环境的效率.4 系统测试4.1 系统后端接口测试项目采用了微服务系统常见的前后端分离结构,前端在接口确定后采用桩服务器(mock-server)与后端并行开发. 后端系统测试过程中使用了postman 进行接口的测试, 如图8所示.Postman [19]是一个用于API 接口开发的协作平台.它提供了一套API 接口开发工具, 包括API 客户端、文档生成、自动化测试、API 接口监控、API 设计与桩服务器与接口文档协作.Postman 内嵌了OAuth2的认证模型, 通过配置即可实现一系列共用同认证服务的API 同时获得登录认证的功能.图8 使用postman 对后端接口进行测试4.2 系统功能测试以深度学习图像数据制备流程为例, 对系统中图像导入、任务分配与标定、数据导出流程进行功能测试.首先是图像导入. 图像标注系统支持图像、视频、PDF 、OpenDocument 等格式混合打包上传, 上传界面如图 9所示. 上传完成数据传输后转入异步解压处理流程, 异步解压完成后前端显示实际图像集大小.在图像集中创建出标定集后即可向系统中其他用户分配标定任务进行图像标定的团队协作. 分配任务过程如图 10所示.切换至任务执行用户, 可以在“我的任务”页面中看到由任务发起用户分配的任务, 点击“开始”按顺序对图像进行标定. 进入标定流程后, 如图 11所示,通过在图像上拖拽创建图像区域, 完成对图像内容的标记.图9 创建图像集图10 分配标定任务标注过程中发起任务的用户能够看到各个任务执行用户的标定进度与当前已标定图像的标签统计信息.当图像集标定完成后, 任务发起者可以对图像进行最终修正并导出图像集合与标定集合开展深度学习模型训练. 以导出的VOC XML 格式为例, 最终导出的标签集如图 12所示.图11 图像标注过程图12 标签集合导出计算机系统应用2021 年 第 30 卷 第 5 期。

一种多点定位采水系统设计构想

一种多点定位采水系统设计构想

技术平台一种多点定位采水系统设计构想翟伟东(西北工业大学计算机学院,陕西 西安 710072)摘 要:本设计构想旨在设计一种多点定位采水系统,更具体的是一种能够用来多点定位采集水样并进行分瓶存储的采水系统。

关键词:多点定位;倾斜式排阀0 引言水是生命之源,人类的生产生活活动都离不开水,水质的好坏在很大程度上影响着工业品质量和生活质量。

在水库、公园、湿地这些水资源丰富并且其水质与大众身心健康息息相关的地方,对水质进行定期定点监测尤为重要。

对水质进行定期定点监测是指针对某一水域,根据事先规划好的路径或是事先标记好的地点,按照一定的时间规律(比如每周一的上午)进行多点水质采样,通过水质采样并对所采集的样本进行专业仪器的分析与检测,从而实现定期监测水质的功效。

目前大多数的湖泊、公园、湿地等水域都采取了这样的检查方式。

按照河流型湿地公园、湖库型湿地公园、沼泽型湿地公园的分类,每种类型湿地公园水环境监测采样点应能覆盖所需的生态检测和评价范围,除特殊需要(因地形、水深和监测目标所限制)外,所有采样点应在检测范围内均匀布设,可采用网格式、断面或梅花式等布设方式进行采样点布置。

并且为保证采样的连续性和周期性,通常水环境中部分自然指标如水温、PH、DO等数据需要进行实时连续的监测,其他指标的监测时间设置为丰水期、平水期、枯水期各采样一到两次。

为执行上述任务,现在的湿地公园或是湖泊管理部门都采用了人工操作的人工取样装置,所使用的取样设备十分简陋,一般采用惰性材料(如不锈钢、聚四氟乙烯等)制作成瓶状的采样器,而且操作方式为人工吊放,一方面仅就操作手法而言缺乏规范操作,容易由于人体与水样接触而影响采样结果的准确性,另一方面则是由于所有的过程都是由人来手动控制,在造成效率低下的同时可能会造成管理上的疏忽导致进行多点采样时出现混乱或是错误。

因此,亟需一种能够提升采集效率,降低人为行动对所采水样产生影响,并且能够按照预设采水地点自行工作、自行记录采水地点的一种自动化采水系统。

基于后悔理论-VIKOR法的微电网规划评价方法

基于后悔理论-VIKOR法的微电网规划评价方法

ELECTRIC DRIVE 2024Vol.54No.2电气传动2024年第54卷第2期基金项目:国家自然科学基金(51877201)作者简介:李睿哲(1998—),男,硕士研究生,Email :通讯作者:吴鸣(1981—),男,博士,教授级高级工程师,Email :基于后悔理论-VIKOR 法的微电网规划评价方法李睿哲1,吴鸣1,2,刘诗语1(1.上海电力大学电子与信息工程学院,上海200090;2.中国电力科学研究院有限公司,北京100192)摘要:针对现有微电网规划评价方法在评价过程中未能充分考虑系统与环境中不确定因素对于评价结果影响的问题,提出一种基于后悔理论-多准则妥协解排序法(VIKOR )的微电网规划评价方法。

首先,考虑微电网规划阶段的需求,构建了包含系统经济性、环保性与技术性等在内的微电网规划评价指标体系;而后,顾及专家主观经验与指标客观数据所包含的信息,采用层次分析法(AHP )-相关性定权法(CRITIC )对指标体系进行主客观组合赋权;最后,考虑不确定因素对评价结果的影响,基于后悔理论在评价过程中引入欣喜-后悔函数与状态变量对VIKOR 法进行改进,建立了基于后悔理论-VIKOR 法的微电网规划评价模型。

算例分析表明,所提评价方法考虑系统与环境中不确定因素以及决策者心理预期对于规划方案评价结果的影响,贴近于实际的微电网的规划方案决策需求。

关键词:微电网;综合评价;后悔理论;VIKOR 法中图分类号:TM715文献标识码:ADOI :10.19457/j.1001-2095.dqcd24651Evaluation Method for Microgrid Planning Based on Regret Theory VIKOR MethodLI Ruizhe 1,WU Ming 1,2,LIU Shiyu 1(1.School of Electronics and Information Engineering ,Shanghai University of Electric Power ,Shanghai 200090,China ;2.China Electric Power Research Institute ,Beijing 100192,China )Abstract :To address the issue that existing microgrid planning evaluation methods do not adequately account for the influence of uncertainties in the system and environment on the evaluation results ,a microgrid planning evaluation method based on regret theory -vlsekriterijumska optimizacija i kompromisno resenje (VIKOR )was proposed.Firstly ,a microgrid planning evaluation index system was constructed ,which includes system economics ,environmental protection and technology ,taking into account the needs of the microgrid planning stage.Then ,the analytic hierarchy process (AHP )-criteria importance though intercrieria correlation (CRITIC )method was used to assign weights to the indicators ,taking into account the subjective experience of experts and the information contained in the objective data of the indicators.Finally ,considering the influence of uncertainties on the evaluation results ,the VIKOR method was improved by introducing the euphoria-regret function and state variables in the evaluation process based on regret theory ,and a microgrid planning evaluation model based on regret theory-VIKOR method was established.The analysis of the example shows that the proposed evaluation method consider the influence of uncertainties in the system and environment as well as the psychological expectations of decision-makers on the evaluation results of the planning scheme ,and is close to the actual microgrid planning scheme decision needs.Key words :microgrid ;comprehensive evaluation ;regret theory ;VIKOR method在日趋严峻的能源环境形势下,国家大力推动“双碳”战略的实施以及以新能源为主体的新型电力系统的建设来完善能源绿色低碳转型体制机制,微电网技术作为实现大规模可再生能源就地消纳的关键技术而日益普及。

新能源电力系统中的智能微电网运行与优化研究

新能源电力系统中的智能微电网运行与优化研究

51 电力技术应用光伏太阳能板直交变流器直交变流器电池储能系统负载主网P PG (t )P ESS (t )P lood (t )P grid (t )图1 基于光伏的微电网结构1.2 智能化与自动化新能源电站电网接入下的智能微电网将朝着智能化与自动化方向发展。

智能微电网在运行过程中会产生大量的信息数据,如何有效整合并分析这些数据非常重要。

通过应用人工智能技术能够有效提高数据的分析效率和预测算法的精准性,同时电力企业应不断加强对智能微电网技术的创新和优化。

例如,将储DC-DC 双向变换器蓄电池负载n 负载2负载1微源变换器负载 变换器负载 变换器负载 变换器图2 微电网群观群控技术应用架构 2024年3月25日第41卷第6期53 Telecom Power TechnologyMar. 25, 2024, Vol.41 No.6王 启,等:新能源电力系统中的智能微电网运行与优化研究3 新能源电站电网接入下的智能微电网运行优化策略3.1 能源管理策略新能源电站电网接入下的智能微电网运行的核心是能源管理策略,即将风能、太阳能等新能源发电系统结合到一起,对所有的电力能源进行统一发协调管理,以充分利用可再生能源。

能源管理策略是确保新能源发电系统与传统电力供给系统有效结合的关键[4]。

一方面,利用能源管理策略将不同地区的用电需求量与供电量进行相互关联,并根据用电需求对能源进行合理调配。

例如,先利用可再生能源进行供电,当供电不足时,再通过火力发电来满足剩余的电力需求。

另一方面,通过能源管理策略可以动态调整电力能源的存储量和释放量。

例如,在用电需求低峰期存储多余的电力能源,在用电需求高峰期释放存储的电力能源,以提高能源的供需平衡水平,提高智能微电网的运行质量和效率。

3.2 系统安全性优化在智能微电网运行过程中,应不断加强对系统的安全性优化。

因为新能源电站电网接入下的智能微电网在日常运行过程中会产生大量的信息数据,如果这些信息数据出现泄露、丢失等问题,不仅会对智能微电网的正常运行造成严重影响,还严重威胁着用户的个人隐私安全[5]。

家用电器低碳化技术——网络篇

家用电器低碳化技术——网络篇

技术前沿·CUTTING-EDGE TECHNOLOGY48电器 2012/9黄逊青信息交换技术的应用使家用电器的功能和性能发生了巨大飞跃,同时也令家电运行方式发生变革。

多年来,具备信息交换能力被视为智能化家电的特征之一。

信息交换不仅拓展了家电控制系统的信息输入类型和数量,也促进了家电系统运行管理趋于完善,还延伸了家电的应用范围。

其中,家电的互联和自主运行能力在信息交换技术推动下取得了令人瞩目的发展。

家用电器应用信息交换技术由来已久,但是早期采用的专用数据传输方式在技术和经济性上受到诸多限制,应用范围有限。

20世纪末,利用公共有线或无线电话网络对家电进行远程监控的产品出现,人们期待这类信息交换方式能够给家电行业带来变革。

然而,随着通讯网络技术的发展,利用局域网和互联网等通讯平台实现信息交换的方式,逐步取代了传统电话网络方式。

目前家电的信息交换方式可以用网络化来概括。

网络化的信息交换方式必须基于标准化的通讯协议。

对于能耗较大的白色家电而言,信息交换的基本目的是实现产品的自主优化运行,包括与其他家电的协调运行,减少住宅能源支出。

住宅能源管理系统(HEMS)、网络通讯协议、信息传输方式都是住宅能源管理网络技术的重要组成部分。

智能能源网2002年,美国CERTS 对微型电网(micro grid)做出定义。

微型电网是一种由负荷和微型电源共同组成的系统,可同时提供电能和热能;微型电网的内部电源主要由电力电子器件负责能量转换,并提供必要的控制;相对于外部电网,微型电网为单一的受控单元,可同时满足用户对电能质量和供电安全的要求。

业内普遍认为,这是迄今为止最权威的微型电网定义。

从字面来看,grid 并非特指电网,在CERTS 给出的定义中,微型电网提供的能量并非仅为电能,因此将micro grid 翻译为微型能源网更为准确。

在现有技术条件下,在住宅或住宅小区中构建微型能源网必须依靠信息家用电器低碳化技术——网络篇Copyright ©博看网. All Rights Reserved.49电器 2012/9技术的支撑,具备这些特征的网络可以称为智能互动微型能源网。

基于PSO-BP神经网络的储能装置实时容量识别与实现

基于PSO-BP神经网络的储能装置实时容量识别与实现

现代电子技术Modern Electronics Technique2020年6月15日第43卷第12期Jun.2020Vol.43No.120引言新能源作为未来发电的重要部分,在减少环境污染方面具有重大影响,而储能装置作为新能源发电必不可少的组成部分,在提高电能质量、维护系统稳定性、电源应急备用、提高经济效益等方面发挥了不可或缺的作用[1]。

对储能装置进行容量识别,对微电网的能量管理与调度极具实质意义,鉴于这一技术的重要作用,储能装置的容量识别得到了广泛的关注。

因BP 神经网络的结构简单,可调的参数多,可操作性好,故采用BP 神经网络构建储能装置的实时容量识别模型是一种简便可行的方法。

文中识别的储能装置是基于磷酸铁锂电池组和铅炭电池组的,由于电池组的容量与电池组的充电和放电电流、两端电压、工作温度和电池组循环次数有关,而且电池内部复杂的物理和化学机制使得难以获得精确的数学模型,这就造成了难以准确识别储能装置容量值的一个现实问题。

为此,文中介绍了一种以神经网络为代表的智能算法,用于识别储能装置的容量[2],并建立了基于优化粒子群算法的改进BP 神经网络识别模型,通过实现对储能装置容量的识别、对提高微电网的基于PSO⁃BP 神经网络的储能装置实时容量识别与实现吕磊,王红蕾(贵州大学电气工程学院,贵州贵阳550025)摘要:微电网下新能源的使用促使了人们对能量优化调度的研究,储能装置是其中的关键一部分,准确识别其存储容量是实现微电网电力优化调度的任务之一。

为实现对微电网下储能装置的实时容量在线识别,利用BP 神经网络结构建立了储能装置容量识别模型,并引入了优化的粒子群算法PSO ,实现了储能装置的实时容量在线识别。

通过对比传统的BP 神经网络识别结果,采用PSO⁃BP 神经网络识别模型的容量误差在0.3%~2.4%之间,传统的BP 神经网络误差范围为1.0%~21%,表明采用PSO⁃BP 神经网络识别模型明显优于传统的BP 神经网络识别模型。

风光柴储微电网系统的研究-毕业论文

风光柴储微电网系统的研究-毕业论文

---文档均为word文档,下载后可直接编辑使用亦可打印---摘要伴随着人类社会的发展,人类对大自然的破坏越来越严重,过于依靠煤和石油等化石能源来供给能量,各种环境问题也随之产生慢慢凸显,人们也开始更加关注环境问题,风能、光能等可再生能源的发电受到人们越来越大的重视以及应用。

微电网是一种小型发配电系统,它的构成主要由分布式电源模块、储能模块、负荷模块以及监控保护模块组成,它同时也被叫做微网。

由于资源的缺乏和空气污染等原因,利用风能和太阳能等无污染能源发电的新能源发电技术已成为科学研究的目标。

为了实现分布式电源的灵活高效以及解决各种各样各不相同的分布式电源并网问题,我们提出了风光柴储微电网系统。

风光柴储微电网系统里面包括了太阳能发电模块、风力发电模块、柴油发电机发电模块、储能装置模块、负荷模块和连接他们的微电网。

为了提高促进分布式电源发电与可再生新能源的紧密联系以及实现对各式的负载电源供电可靠性,我们对分布式电源进行开发设计和扩展,找到一种合理的维护方式。

微电网将传统电网连接到了智能电网。

风光柴储微电网系统能够有效解决在普通电网无法到达的远距离供电这一问题,虽然它有很多好处,但是它不支持大电网的备用储能,同时风能和太阳能的供电并不稳定,具有间歇性波动性,受到自然环境因素的影响比较大,所以对于微电网的合理配置也成了研究的重点。

微电网的合理配置降低了建设成本和发电成本,提高了供电的可靠性,降低了负载间的电损,为微电网的优化起到至关重要的作用。

最开始,研究了微电网系统的目的以及意义,应用在哪些领域,在国内外的发展趋势以及前景,还有本文要解决的主要问题。

其次,针对风力发电、光伏发电、柴油机发电的原理进行了探讨,随之讨论了风光柴储微电网系统的结构组成,之后再对各模块的建模进行研究。

最后,对系统进行仿真,解决拟提出的问题,得出结论。

关键词:环境问题;微电网;风光柴储微电网;可再生能源;智能电网;间断性Research on Wind, Solar, Diesel and Storage Microgrid SystemAbstractWith the development of human society, the destruction of human beings to nature is more and more serious, relying too much on fossil energy such as coal and oil to supply energy, various environmental problems are gradually emerging, and people are beginning to pay more attention to environmental problems. The power generation of renewable energy such as wind energy and light energy is getting more and more attention and application.Micro-grid is a small-scale power generation and distribution system, which is composed of distributed power module, energy storage module, load module and monitoring and protection module. It is also called micro-grid. Due to the lack of resources and air pollution, the new energy power generation technology using wind energy and solar energy has become the goal of scientific research. In order to realize the flexibility and efficiency of the distributed power and solve various different problems of the grid connection of the distributed power, we propose a wind and diesel micro grid system. The wind and diesel micro grid system includes solar power module, wind power module, diesel generator power module, energy storage device module, load module and micro grid connecting them. In order to promote the close connection between distributed power generation and renewable new energy and realize the reliability of various load power supply, we develop, design and expand the distributed power supply to find a reasonable maintenance mode. Micro-grid connects the traditional grid to the smart grid.The wind and diesel micro grid system can effectively solve the problem of long-distance power supply which can not be reached by the ordinary grid. Although it has many advantages, it does not support the backup energy storage of the large grid. At the same time, the power supply of wind and solar energy is not stable, with intermittent volatility, which is greatly affected by the natural environment factors. Therefore, the reasonable configuration of the micro grid has also been studied Emphasis. The reasonable configuration of micro-grid reduces the construction cost and generation cost, improves the reliability of power supply, reduces the power loss between loads, and plays an important role in the optimization of micro-grid.At the beginning, the purpose and significance of micro-grid system, its application fields, development trends and prospects at home and abroad, and the main problems to be solved in this paper are studied.Secondly, the principles of wind power generation, photovoltaic power generation and diesel engine power generation are discussed, and then the structure of the micro grid system is discussed, and then the modeling of each module is studied.Finally, the system is simulated to solve the problems to be proposed, and a conclusion is drawn.Keywords: environmental issues; micro-grid; wind and diesel storage micro-grid; renewable energy; smart grid; intermittent; volatility; optimization目录1前言 (4)1.1研究的目的及意义 (5)1.2技术应用 (5)1.3发展前景 (5)1.4研究拟解决的关键问题 (6)2风光柴储微电网系统的研究与设计 (6)2.1设计原理 (6)2.2风光柴储微电网系统的结构 (8)3建模 (9)3.1风力发电机组建模 (9)3.2光伏电池发电仿真建模 (10)3.3储能电池建模 (11)3.4柴油发电机建模 (12)3.5 负荷建模 (13)4风光柴储微电网系统建模仿真 (13)4.1柴油发电机(电压,电流,视在功率,有功功率,无功功率) (14)4.2风电场(电压,电流,视在功率,有功功率,无功功率,风速) (15)4.3光伏发电(电压,电流,视在功率,有功功率,无功功率) (16)4.4储能系统(电压,电流,视在功率,有功功率,无功功率) (17)4.5负荷(电压,电流,视在功率,有功功率,无功功率) (18)4.6电动机投入(电压,电流,视在功率,有功功率,无功功率) (19)5总结 (20)参考文献 (21)谢辞 (21)1前言随着科技的高速发展,人们的日常生活也在随之变化,但同时社会中的人口也在变多。

储能行业术语

储能行业术语

储能行业术语储能行业涉及到许多专业术语,这些术语涵盖了电力储能的各个方面。

以下是一些储能行业常见的中文术语:1.蓄电池(Battery):储能系统中用于存储电能的设备,通常是可充电电池。

2.储能系统(Energy Storage System,ESS):用于储存和释放电能的设备和系统,包括蓄电池、逆变器等组件。

3.逆变器(Inverter):将直流电转换为交流电或将交流电转换为直流电的设备,用于连接储能系统与电力网络。

4.功率电子设备(Power Electronics):包括逆变器、整流器等设备,用于控制电能的转换和流动。

5.充放电效率(Charge-Discharge Efficiency):储能系统在充电和放电过程中的能量损失比率。

6.容量(Capacity):储能系统能够存储的电能总量,通常以千瓦时(kWh)或兆瓦时(MWh)为单位。

7.循环寿命(Cycle Life):储能系统可循环充放电的次数,是评估其寿命的重要指标。

8.储能管理系统(Energy Management System,EMS):用于监控、控制和优化储能系统运行的软件和硬件系统。

9.负荷平滑(Load Leveling):利用储能系统调整电能的供应,以平滑电力系统的负荷曲线。

10.容量储能(Capacity Storage):将电能储存在系统中,以备不时之需。

11.功率储能(Power Storage):储能系统能够瞬时释放大量电能,用于处理瞬时的高负荷需求。

12.微电网(Microgrid):由分布式能源资源、储能系统和智能控制组成的小型电力系统。

这些术语涵盖了储能行业的关键概念,帮助描述和理解电力储能技术及其应用。

Smart Grid and Energy Storage

Smart Grid and Energy Storage

Smart Grid and Energy Storage Smart Grid and Energy Storage: Revolutionizing the Future of Energy The world is facing an unprecedented energy challenge. As population growth and urbanization continue to accelerate, the demand for electricity is skyrocketing. At the same time, the need to reduce greenhouse gas emissions and combat climate change is becoming increasingly urgent. In this context, the smart grid and energy storage technologies are emerging as game-changers that have the potential torevolutionize the future of energy. The smart grid is a modernized electricalgrid that uses advanced digital technology to monitor and control the flow of electricity from power plants to consumers. Unlike traditional grids, whichoperate in a one-way flow of electricity, the smart grid enables two-way communication between the utility and the consumer. This bidirectional flow of information allows for real-time monitoring and optimization of energy consumption, leading to increased efficiency and reliability. One of the key advantages of the smart grid is its ability to integrate renewable energy sources into the grid. As the world shifts towards a more sustainable energy future, the smart grid can facilitate the integration of intermittent renewable sources, such as solar and wind, by balancing supply and demand in real-time. This is achieved through theuse of advanced sensors, communication networks, and intelligent algorithms that enable the grid to anticipate and respond to fluctuations in renewable energy generation. Energy storage, on the other hand, plays a crucial role in ensuring the stability and reliability of the smart grid. As renewable energy sources are inherently intermittent, energy storage technologies can store excess energyduring periods of high generation and release it during periods of low generation. This not only helps to balance supply and demand but also improves the overall efficiency of the grid by reducing the need for conventional backup power plants. Furthermore, energy storage can also provide a range of other benefits. For instance, it can enhance the resilience of the grid by providing backup power during blackouts or natural disasters. It can also enable the integration of distributed energy resources, such as rooftop solar panels and electric vehicles, by providing a means to store and utilize the energy locally. This can help to reduce transmission losses and alleviate stress on the grid. From anenvironmental perspective, the smart grid and energy storage technologies offer significant advantages. By enabling the integration of renewable energy sources and optimizing energy consumption, they can help to reduce greenhouse gas emissions and mitigate the impacts of climate change. Moreover, energy storage technologies, such as batteries, are typically more environmentally friendly than conventional backup power plants, which often rely on fossil fuels. However, the widespread deployment of smart grid and energy storage technologies is not without challenges. One of the main barriers is the high cost of implementation. Building a smart grid infrastructure and deploying energy storage systems requires significant upfront investments. While the costs have been steadily declining in recent years, they still pose a significant challenge, particularly for developing countries with limited financial resources. Another challenge is the need for regulatory and policy frameworks that support the integration of these technologies. The traditional regulatory models were designed for centralized,one-way electricity systems and may need to be updated to accommodate the bidirectional flow of electricity and the new business models that emerge with the smart grid. Additionally, policies that incentivize the adoption of energy storage technologies, such as feed-in tariffs or tax credits, can help to overcome the barriers to deployment. In conclusion, the smart grid and energy storage technologies have the potential to revolutionize the future of energy. By enabling the integration of renewable energy sources, optimizing energy consumption, and improving the overall efficiency and reliability of the grid, they can help to address the energy challenges of the 21st century. However, their widespread deployment requires overcoming various challenges, including high costs and the need for supportive regulatory and policy frameworks. With the right investments and policies in place, the smart grid and energy storage can pave the way for a more sustainable and resilient energy future.。

微电网英语

微电网英语

Operation and Control of Microgrid System Integration and Hierarchical Power Management Strategy for a Solid-State TransformerInterfaced Microgrid SystemSchool of Electrical EngineeringElectrical EngineeringContents1 Detail Abstract (1)2 Core techniques of the paper (1)2.1 Hardware integration of SST and dc microgrid (1)2.2 Hierarchical power management strategy (3)3 Potential problems and my opinion on future work (5)1 Detail AbstractThe existing dc microgrid can only interface with the distribution system by using a heavy and bulky line frequency transformer plus rectifier, and the passive transformer cannot provide functions such as Var compensation or harmonic filtering. Therefore, developing a more compact and active grid interface to enable a more intelligent dc microgrid system is a research focus.Under the situation, this paper investigates, and for the first time presents, the system integration of a novel solid-state transformer (SST) interfaced microgrid system. Accordingly, a hierarchical power management strategy is proposed for this system to enable islanding mode operation, SST enabled operation, and the seamless transfer between two modes.To begin with, the paper review the past achievements in the microgrid field and introduce the different types of microgrid from the architecture and the power management point of view. Then the author presents the detailed structure of the proposed system, as is shown in Fig. 1.Fig 1. SST-enabled dc microgrid diagram.Furthermore, the researcher depicts the hierarchical control frame and analyzes the hierarchical control for dc microgrid in islanding mode and SST-enabled mode. The hierarchical power management strategy includes three control levels: primary control for the local controller; secondary control for the dc microgrid bus voltage recovery; and tertiary control to manage the battery state of charge.Finally, a lab test bed is constructed to verify the system performance, and several typical case studies are carried out. The experimental results verify the proposed system and distributed power management strategy.Keywords: DC microgrid, hierarchical power management, islanding mode, solid-state transformer (SST)-enabled mode.2 Core techniques of the paper2.1 Hardware integration of SST and dc microgridAs previously mentioned, the conventional transformer interfaced with the dc microgrid is heavy and bulky, which takes too much space. The core technique of thispaper is, for the first time, presenting the hardware integration of SST and dc microgrid to demonstrate the feasibility of this novel concept. Compared to the conventional microgrid architecture, the presented microgrid system interfaces with the distribution system by an active grid interface with smaller size and less weight. The SST interfaced microgrid is, therefore, a more compact system.The SST is a power electronic device that replaces the traditional 50/60 Hz power transformer by means of high-frequency transformer isolated ac–ac conversion technique, and its topology is represented in Fig. 2, where a cascaded seven level rectifier is adopted as the font-end. Three dual active bridge (DAB) converters are connected to the floating dc links of the rectifier with the secondary side connected in parallelFig. 2. Topology of presented SST.The basic operation of the SST is first to change the 50/60 Hz ac voltage to a high-frequency voltage, then this high-frequency voltage is stepped up or down by a high frequency transformer, and finally shaped back into the desired 50/60 Hz voltage to feed the load. For a high-frequency transformer has significantly decreased volume and weight, the first advantage that the SST may offer is reduced volume and weight compared with traditional transformers. A laboratory prototype of a single-phase, three-stage SST was built, as shown in Fig. 3.Fig.3. Cascaded type three-stage SST prototype.It is further seen from the topology and the configuration of the SST that some other potential functionalities that are not available to traditional transformer may be obtained. A functional diagram of SST is illustrated in Fig. 4. The SST acts as a smart plug-and play interface for transforming and distributing electric energy from these various different subsystems, some via the ac port and others via the dc port. The use of SST separates the grid side parameters (voltage, frequency) from the DRER (Distributed renewable electric resource) and DESD (Distributed energy storage device) side. This is a very important capability of the proposed system that strengthens the system stability because the low-voltage side is strongly decoupled from the grid side by the SST.Fig. 4. Solid state transformer functional configuration.2.2 Hierarchical power management strategyFrom the power management point of view, the presented microgrid system need to ensure proper and optimal operation under different conditions. Some major challenges for SST-enabled dc microgrid include:1) how to make the dc microgrid more reliable in islanding mode;2) how to achieve seamless transfer between islanding mode andSST-enabled mode for the dc microgrid;3) how to manage individual modules in the dc microgrid considering thecharacter differences when system is in SST-enabled mode.To address these challenges, another core technique of this paper is a corresponding hierarchical power management strategy, which combines the advantages of centralized control and distributed control.Its basic structure is shown in Fig. 5.Fig 5. Hierarchical control frame.The primary control is the distributed control which ensures that the microgrid system can operate without communication. Therefore, the primary control usually takes effect at the microsecond level, which is basically the same level as the converter control. All the local information, including the voltage, current, SOC (State of charge), etc., are sent to the upper controller, which implements the tertiary control and secondary control through a bidirectional communication link. Here, the dc microgrid is enabled by the SST, and therefore the SST controller is used as the upper controller. The objective of secondary control is to recovery the microgrid bus voltage to achieve seamless transfer as the system switches from islanding mode to SST-enabled mode. The time scale for the secondary control is on the order of milliseconds to seconds. The objective of tertiary control is to ensure that battery operates in a reliable SOC range. Thus, the tertiary control is used to charge and discharge the battery in the dc microgrid based on battery’s SOC instead of controlling the point of common coupling power flow. Detailed structure of the hierarchical control strategy is depicted in Fig. 6.Fig. 6. DC microgrid operation diagram.In summary, the proposed SST-enabled dc microgrid can:1)interface with distribution system via SST;2)operate in islanding mode with distributed control;3)seamlessly transfer between islanding mode and SST enabled mode;4)enable battery management in SST-enabled mode.3 Potential problems and my opinion on future workAs is shown in Fig. 6, the SOC monitoring is implemented in the tertiary control, in which battery management is achieved by suitable charging and discharging, but the monitoring method and its hardware design are not mentioned in this paper.In my opinion, accurate monitoring of the energy storage system performance is the fundamental basis to achieve the hierarchical control. However, the current monitoring technology is still not mature enough, of which SOC cannot be directly measured. A pressing matter of the moment, therefore, is to establish an accurate estimation model of the SOC monitoring.In addition, the experimental results in this paper only demonstrate the feasibility of the SST interfaced microgrid system in island operation. But when it is parallel with Power Grid, the power electronic devices of SST may be so sensitive (usually take effects at the microsecond level) that the traditional grid devices as circuit breaker and isolating switch can’t respond synchronously. This may impact the stability of Power System.My opinion is to solve the problem from the following aspects:1) Taking the interaction between the SST and the distributed system intoconsideration to guide the design of the proposed microgrid system;2) Constructing the architecture of multi-layer-cross distributed renewable energymanagement based on the advantages of agent technology;3) Using artificial intelligence technology to control several nodes in themicrogrid and cooperatively;4) Connecting more PV and battery modules to the dc bus to verify the robustoperation of the system.[文档可能无法思考全面,请浏览后下载,另外祝您生活愉快,工作顺利,万事如意!]。

  1. 1、下载文档前请自行甄别文档内容的完整性,平台不提供额外的编辑、内容补充、找答案等附加服务。
  2. 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
  3. 3、如文档侵犯您的权益,请联系客服反馈,我们会尽快为您处理(人工客服工作时间:9:00-18:30)。
相关文档
最新文档