基于无线传感器网络的co监测系统

Telecommun Syst(2013)53:47–53DOI10.1007/s11235-013-9675-4Design of wireless sensor network node for carbon monoxide monitoringZujue Chen·Zhan Shi·Qiuyue GuoPublished online:20June2013©Springer Science+Business Media New York2013Abstract Aim at protecting environment,a design proposal of Wireless Sensor Network(WSN)node for remote moni-toring Carbon Monoxide is presented.The monitoring node uses a low-power SOC microprocessor C8051F930who acts as a controller,utilizes Carbon Monoxide metal ox-ide(MOX)semiconductor sensor whose detective princi-ple is that MOX semiconductor has redox reactions with CO and its resistance varies with gas concentration when it is heated,perceiving local CO concentration,condition-ing the collected CO signals from the sensor,converting and processing datas,etc.The datas can be transmitted to Sink by high-performance wireless radio frequency chip Si4432 which supports for communication protocol IEEE802.15.4, and Sink uses GPRS to send them remotely to the mobile network,realizing remote monitoring CO.Keywords Carbon monoxide sensor·Microprocessor·Data processing·Wireless radio frequency chip·Remote monitoring1IntroductionCarbon monoxide is inhaled for a long time,which can cause body histogenous hypoxia,finally leading to suffocation. Z.Chen( )·Z.Shi·Q.GuoCollege of Computer Science and Communication Engineering, Jiangsu University,Zhenjiang,Jiangsu212013,Chinae-mail:chenzujue@Z.Shie-mail:shizhan-shelly@Q.Guoe-mail:hanruoxue1986@ It is particularly important to design a real-time monitor-ing system for carbon monoxide to detect carbon monoxide emissions.In thefield of monitoring of harmful gases,domestic and foreign relevant researchers have developed a number of gas sensor detection systems,gas detection technology has matured gradually.In2010Spain’s Prades,etc.designed a gas sensors made of metal oxide nanomaterials which can rapidly and selectively detect carbon monoxide and water vapor[1].In2010U.S.Sazia A Eliza,etc.designed and sim-ulated a highly selective gas sensors which can monitor hy-drogen,carbon monoxide,etc.[2].In2008the researchers of Zhengzhou Aeronautical Industry Management Institute designed a kind of carbon monoxide density distributed on-line detection system using single chip microprocessor tech-nology and CAN bus technology,realizing the concentration of carbon monoxide on-line monitoring and data real-time analysis and processing[3].In2009State Key Laboratory of Fire Science China University of Science and Technol-ogy designed wirelessfire detection system that uses short-range wireless transceiver chip nRF2401and long-distance wireless communication module GPRS,realizing wireless multisensorfire detection[4].In2011College of Mechan-ical Engineering of Taiyuan University of Technology de-signed a Coal Mine Detection Robot for the detection of harmful gases in coal mining areas to enter deeply into the mine environment andfinish explorational tasks instead of the staff[5].For a large number of carbon monoxide gas emissions pollute the environment because of coal spontaneous com-bustion,present a design proposal of carbon monoxide mon-itoring node for wireless sensor network so that the gas mon-itoring node can be with the characteristics of miniaturiza-tion,low power consumption and wireless communication, achieving real-time remote monitoring carbon monoxide.48Z.Chen et al.2Carbon monoxide gas detection principle2.1The selection of gas sensorGas sensors are chemical sensors.According to the sen-sors’sensitive materials and the different mechanisms and effects of the interaction with the gas,the gas sensors can be divided into semiconductor gas sensors,solid electrolyte gas sensors,the contact combustion-type gas sensors,op-tical gas sensors[6],quartz resonant gas sensors,surface acoustic wave gas sensors,solid polymer electrolyte-type gas sensors,electrochemical gas sensors.Semiconductor gas sensors can be divided into metal oxide semiconductor gas sensors and organic semiconductor gas sensors.Metal oxide semiconductor gas sensors can be divided into the resistive and the non-resistive.This system uses the resistive metal oxide semiconductor gas sensor TGS2442.2.2Metal oxide semiconductor gas sensor detectionprincipleCarbon monoxide sensor TGS2442is metal oxide semi-conductor gas sensor,whose main ingredient is high active MOX tin dioxide sintered body.This type of sensor utilizes the physical phenomena of that MOX semiconductor has re-dox reactions in contact with the gas and its conductivity varies with the ambient gas concentration to detect gas con-centrations.The sintered body’s surface of tin dioxide exists many active particles(SnO2−x)∗.When the metal oxide semi-conductor is heated to a certain temperature in the air,the oxygen in the air are adsorbed on the surface of negatively charged semiconductor.Electronics on the surface of semi-conductor will be transferred to the adsorbed oxygen into negative oxygen ions(O−,O2−):12O2+(SnO2−x)∗=O−ad(SnO2−x)(1)O−ad(SnO2−x)signifies the O−is adsorbed on the SnO2−x. At this point a positive cumulative space charge layer which is formed on the semiconductor surface makes the surface barrier increase,impeding theflow of electrons,which leads to the increasement of the material’s resistance.Within the gas sensitive material,free electrons must pass through the micro-grain’s binding sites(grain boundaries)of metal ox-ide semiconductor to form aflow of electrons,barrier gen-erated by oxygen adsorption also exists at grain boundaries and impedes the freeflow of electronics,which the sensor’s resistance is due to.When reducing gas CO which is as the detected gas contacts with the surface of the semiconduc-tor component,CO and adsorbed oxygen negative ions have chemical reaction:CO+O−ad(SnO2−x)=CO2+(SnO2−x)∗(2)Electronics captured by oxygen atoms return to the sur-face of semiconductor oxide,the surface barrier decreases with its ambient CO concentration reducing,then the resis-tance of the gas sensitive material decreases.With the en-hanced absorption of carbon monoxide,the energy band of the metal oxide semi-conductor varies with the conductivity. Combine with Schottky barrier model about grain bound-ary and Freundlich isotherm adsorption principle about solid surface and gas to derive the relationship of tin dioxide’s conductance G and measured CO’s concentration C as fol-lows.Tin dioxide semiconductor generates current under the action of the additional voltage Va:I=AT2exp−q φskTexpqV akT−1(3)among which,the q is electronic charge,the V a is the ad-ditional voltage,the A is the Richardson constant,the φs is the variation of barrier voltage.Because conductance is the current’s derivation with the voltage,get the derivation as follows:G=dIdV aV a=0=AT2exp−q φskTqkT=AT qkexp−q φskT(4)The relationship of the variation of barrier voltage φs and the surface coverage of sensitive materialθis expressed by the formula(5):φs=−qN2sθ22εε0N d(5)among which,the N s is the number of charge per unit area, the N d is the number of donor per unit volume,theε,ε0is the semiconductor’s dielectric constant,the symbol‘−’in-dicates that the barrier reduces.The relationship of the con-ductance G and the surface coverageθcan be deduced by formulas(4)and(5)as follows:G=AT qkexpq2N2sθ22kTεε0N d(6)The formula(6)indicates that the conductance G is pro-portional to the square of the surface coverageθ.Utilize Fre-undlich adsorption principle to derive the relation of the sen-sitive material’s conductance G in the reducing gas and the gas concentration C.Freundlich isotherm adsorption equa-tion is:θ=KC b(7) among which,K=αRT nα’,α =αexp(− E kT),theαand the R are constants,the b is a constant related to tempera-Design of wireless sensor network node for carbon monoxide monitoring49Fig.1The system structure diagram of WSN node for carbon monox-ide monitoringture.Eliminateθby the formula(6)and the formula(7):G=AT qkexpq2N2s K2C2b2kTεε0N d=A expBC d(8)among which,A =AT q k,B=q2N2s K22kTεε0N d ,d=2b.The for-mula(8)gives the quantitative relationship between the sur-face conductance G of the sensitive materials reacting in reducing gas and the gas concentrations C.In the circum-stance of certain gas concentration,as the mutual adsorption effect between tin dioxide and the detected gas is affected by temperature,its electrical conductivity also have a great relationship with the temperature.In the above given con-ditions and proper gas concentration range,the relationship between the sensor’s resistance and carbon monoxide con-centration can be approximately indicated by the following equation:R s=A[C]−α(9) among which,R s:sensor’s resistance;A:constant;C:gas concentration;α:slope of R s curve[7,8].3Hardware design of carbon monoxide WSN node3.1The hardware composition of carbon monoxide WSNnodeCarbon monoxide wireless sensor network node hardware circuit is made up of the signal conditioning circuit collec-tion,data processing unit,the wireless transceiver circuit. The system structure diagram is shown in Fig.1.Select the MCU C8051F930chip as data processing unit while the core of wireless transceiver circuit is the wireless transceiver module Si4432.3.2The design of signal acquisition and condition circuit Sensor’s hardware equivalent circuit is shown in Fig.2.The sensor TGS2442for CO has a heating process at the be-ginning of energization,the heating resistor Heater pack-aged together with it changes sensor’s temperature,making Fig.2Sensor’s hardwareequivalentcircuitit work at the temperature of sensitive to the gas[9].Af-ter the R s is heated,it has adsorption reaction with carbon monoxide in the air,the change of metal oxide semiconduc-tor resistance is reflected by the voltage V out of its series re-sistance R l.The relationship of the output voltage V out and the semiconductor resistance R s is derived from Fig.2,then calculate the gas concentration of CO indirectly by formulas (9)and(10).V c−V outR sR l=V out,R s=R lV c−V outV out(10)The sensitive of the sensor TGS2442is influenced by the surrounding temperature and humidity,the relevant temper-ature and humidity compensation circuit should be consid-ered to eliminate the impact from temperature and humidity on the sensor resistance when the sensor signal detection circuit is designed.In a certain condition of humidity,it will compensate for the changes of sensor’s temperature to use a thermistor.Sensor and thermistor should be in a position that is away from the heat generated by the circuit which can get a good compensation effect[10,11].Signal acquisition and conditioning circuit is made up of power control module,signalfiltering and amplification cir-cuit,transmit the analog voltage signal as the analog input to the I/O port of microcontroller of data processing unit.Power control module uses the voltage regulator TA78M05F to adjust10V into5V.The output signal of the sensor is small when carbon monoxide concentration is low,which needs to be amplified.Conditioning circuit uses the chip TMP47P443VN to generate pulse signal to control heating so that gas sensitive resistance changes in the heat-ing conditions.To reduce drift and error,the load voltage signal is conditioned by second-orderfilter and operational amplifiers and transmitted into the microcontroller’s analog input pin.The diagram of signal acquisition and amplifica-tion circuit is shown in Fig.3.3.3The data processing unitFor the output voltage signal of sensor signal acquisition and conditioning circuit,the data processing unit converts50Z.Chen et al.Fig.3Signal acquisition and amplificationcircuitFig.4The I/O ports configuration ofC8051F930the analogous to the digital,calculates the correspond-ing gas concentration value,establishes a continuous wire-less remote communication with a communication module through the SPI serial port[12].In addition to the above features,data processing unit must also read the datas and control informations from communication module for data fusion control the other modules of the hardware platform, complete the protocol processing of MAC and routing in the process of WSN communication.C8051F930is a low power mixed-signal MCU of SOC.C8051F930has a10-bit A/D converter,which has23 external input channels and a16-bit accumulator for auto-matic averaging with burst mode which can increase A/D converter’s resolution by oversampling.It ensures to collect the datas rapidly and correctly after the sudden rouse.The I/O ports configuration of C8051F930is shown in Fig.4. The voltage output signal of CO gas detection module as analog input connects to the I/O pin P0.7of C8051F930and is converted to the digital by A/D converter.The analog ground signal connects to the I/O pin P0.1of C8051F930. For these analog inputs,select0.5as the programmable gain, use1.65V as the internal high-speed voltage reference,cal-culate the corresponding analog voltage:V out=3300×ADC01023(11)C8051F930can read from and write to the internal reg-ister of the wireless transceiver chip Si4432by the built-in enhanced SPI bus,configuringflexibly various parameters. The microprocessor of C8051F930uses4-wire SPI mode as the master device,including MOSI,MISO,SCK and nSEL. MOSI is used to transmit the serial datas from C8051F930 to Si4432;MISO is used to transmit the serial datas from Si4432to C8051F930;SCK is used to synchronize the serial datas transmission between the C8051F930and the Si4432;Design of wireless sensor network node for carbon monoxide monitoring51nSEL is the chip select signal.Only if the chip select signal is the low level,the operation on the Si4432is valid.3.4The wireless transceiver circuitSi4432highly integrated,low power,multi-band wireless ISM transceiver is part of the EZRadioPROTM family.The Si4432offers advanced radio features including continuous frequency coverage from240to930MHZ and adjustable output power of up to+20dBm.Built-in antenna diversity and support for frequency hopping can be used to further extend range and enhance performance.4Software design of carbon monoxide WSN node4.1The software design of signal acquisitionand processing moduleThe programming of signal acquisition and processing mod-ule reflects in the A/D conversion interrupt function and the main function to calculate the corresponding voltage.The flow chart of ADC interrupt function is shown in Fig.5.Ac-cording to the above-mentioned theory,the code is in Fig.5.4.2The software design of wireless transceiving module The software programming of Wireless transceiver uses the method of modular design,the programming environment is the Silabs IDE V4.02which supports the Keil com-piler and assembler.The major modules of this system are programmed independent functions called by the main function.The modules include:initialization(including C8051F930,SPI,Si4432),ADC interrupt function,wireless sending function and wireless receiving function and so on. Wireless sending function is responsible for transmitting the packet including data load,preamble,sync word,the length of the data payload and the CRC checking byte according to the communication protocol;wireless receiving function is responsible for receiving the packet and checking the CRC checking byte in the packet to ensure the correctness of the received data[13,14].The packet adds a preamble in front of each one,the length of which can be set.To identify the packet’s arrival, the receiver needs the preamble to synchronize the packet. Sync word as a sign of synchronization pattern is after the preamble.The receiver starts to receive datas after the sync word is detected.Si4432integrates internally some func-tions such as modulation/demodulation,coding/decoding, so just set the length of the data load,preamble and sync word at the time of initialization.(1)Initialization function Main(){...while(1){if(adcflg==1;{V out=result;V out=V out∗3300;V out=V out/1023;EA=1;adcflg=0;...}...}INTERRUPT(ADC_ISR,INTERRUPT_ADC0_EOC){static unsigned long accumulator=0;static unsigned int measurements=2048;AD0INT=0;accumulator+=ADC0;measurements--;if(measurements==0){measurements=2048;result=accumulator/2048;accumulator=0;EA=0;adcflg=1;}ADC0H=0;ADC0L=0;}Fig.5Theflow chart of ADC interrupt functionThe initialization function initializes C8051F930,SPI, Si4432internal registors about wireless transceiver fre-quency,working mode,sending rate,etc.(2)Wireless sending functionTheflow chart of wireless sending function is shown in Fig.6.If the datas are transmitted successfully,the LED will be lightened.The code is in Fig.6.52Z.Chen et al.RF_ENUM RFTransmit (uint8∗packet,uint8length){uint8temp8;SpiRfWriteAddressData((REG_WRITE|TransmitPacketLength),length);for (temp8=0;temp8<length;temp8++){SpiRfWriteAddressData((REG_WRITE|FIFOAccess),packet[temp8]);}SpiRfWriteAddressData((REG_WRITE|OperatingFunctionControl1),0x09);SpiRfWriteAddressData((REG_WRITE|InterruptEnable1),0x04);while (RF_NIRQ_PIN ==1);ItStatus1=SpiRfReadRegister (InterruptStatus1);ItStatus2=Sp&&iRfReadRegister (InterruptStatus2);return RF_OK;}Fig.6The flow chart of wireless sending function(3)Wireless receiving functionThe flow chart of wireless receiving function is shown in Fig.7.The data from RX FIFO is send to the LCD by SPI bus to display the result and then receive the next data.The code is in Fig.7.5ConclusionThrough the study on MCU data processing and wireless communication technology,present a design proposal of a low power carbon monoxide monitoring node for wirelessRF_ENUM RFReceive (void){SpiRfWriteAddressData((REG_WRITE|OperatingFunctionControl1),0x05);SpiRfWriteAddressData((REG_WRITE|InterruptEnable1),0x13);SpiRfWriteAddressData((REG_WRITE|InterruptEnable2),0x00);ItStatus1=SpiRfReadRegister (InterruptStatus1);ItStatus2=SpiRfReadRegister (InterruptStatus2);return RF_OK;}RF_ENUM RFPacketReceived (uint8∗packet,uint8∗length){xdata uint8i;if (RF_NIRQ_PIN =0){ItStatus1=SpiRfReadRegister (InterruptStatus1);ItStatus2=SpiRfReadRegister (InterruptStatus2);if ((ItStatus1&0x02)==0x02){∗length =SpiRfReadRegister(ReceivedPacketLength);for (i =0;i <∗length ;i ++){∗packet ++=SpiRfReadRegister (FIFOAccess);}SpiRfWriteAddressData((REG_WRITE|OperatingFunctionControl1),0x01);return RF_PACKET_RECEIVED;}if ((ItStatus1&0x01)=0x01){SpiRfWriteAddressData((REG_WRITE|OperatingFunctionControl1),0x01);return RF_CRC_ERROR;}}return RF_NO_PACKET;}Fig.7The flow chart of wireless receiving functionDesign of wireless sensor network node for carbon monoxide monitoring53sensor network.The carbon monoxide detection range of the monitoring node is from30ppm to1000ppm.The data-transmitting distance can be200m theoretically.This design can be done in future:the data aggregate algorithm,routing algorithm and locating algorithm can be further designed. This monitoring system can be used to realize remote mon-itoring carbon monoxide in coal mine,buildings,etc. Acknowledgements The support of the National High Technology Research and Development Program(863Program)(2006AA10Z258) is gratefully acknowledged.The authors gratefully acknowledge the support of the National High Technology Research and Development Program(863Program)(2006AA10Z258).References1.Prades,J.D.,Hernandez-Ramirez,F.,Fischer,T.,Hoffmann,M.,Muller,R.,Lopez,N.,Mathur,S.,&Morante,J.R.(2010).Quan-titative analysis of CO-humidity gas mixtures with self-heated nanowires operated in pulsed mode.Applied Physics Letters, 24(97),243108.2.Sazia,A.E.,&Dutta,A.K.(2010).Ultra-high sensitivity gas sen-sors based on GaN HEMT structures.In6th international confer-ence on electrical and computer engineering(V ol.12,pp.18–20).3.Wen,X.(2008).Research and design of distributed carbonmonoxide(CO)concentration on-line detection system.In2nd in-ternational symposium on intelligent information technology ap-plication.doi:10.1109/IITA.2008.46.4.Hu,H.,Wang,G.,Zhang,Q.,Wang,J.,Fang,J.,&Zhang,Y.(2009).Design wireless multi-sensorfire detection and alarm sys-tem based on ARM.In9th international conference on electronic measurement and instruments(pp.285–288).5.Stefanovic,N.,&Pavel,L.(2011).A.Lyapunov-Krasovskii sta-bility analysis for game-theoretic based power control in optical links.Telecommunications Systems,47(1–2),19–33.6.Ma,C.,Lee Scott,B.,Pickrell,G.R.,&Wang,A.(2010).Porouscapillary tubing waveguide for high temperature carbon monoxide detection.IEEE Photonics Technology Letters,22(5),March1. 7.Gzara, F.,&Erkut, E.(2011).Telecommunications networkdesign with multiple technologies.Telecommunication Systems, 46(2),February.8.Hackner,A.,Habauzit,A.,Müller,G.,Comini,E.,Faglia,G.,&Sberveglieri,G.(2009).Surface ionization gas detection on plat-inum and metal oxide surfaces.IEEE Sensors Journal,9(12),De-cember.9.Bicelli,S.,Depari,A.,Faglia,G.,Flammini,A.,&Fort,A.(2009).Model and experimental characterization of the dynamic behav-ior of low-power carbon monoxide MOX sensors operated with pulsed temperature profiles.IEEE Transactions on Instrumenta-tion and Measurement,58(5),May.10.Cicioni,M.,Bissi,L.,&Placidi,P.(2009).Interface circuit for anultra low power gas sensor.In International instrumentation and measurement technology conference,Singapore,5–7May.11.Bissi,L.,Cicioni,M.,&Placidi,P.(2011).A programmable inter-face circuit for an ultralow power gas sensor.IEEE Transactions on Instrumentation and Measurement,60(1),January.12.Xing,Z.,Li,D.,&Shan,Y.(2010).The design and implemen-tation of a C8051F-based embedded image acquisition system.In 2nd international workshop on intelligent systems and applica-tions(ISA)(pp.1–3).13.Jan,M.F.,Habib,Q.,Irfan,M.,Murad,M.,Yahya,K.M.,&Has-san,G.M.(2010).Carbon monoxide detection and autonomous countermeasure system for a steel mill using wireless sensor and actuator network.In6th international conference on emerging technologies(ICET)(pp.405–409).14.Kovács,Z.G.,Marosy,G.E.,&Horváth,G.(2010)Case studyof a simple,low power WSN implementation for forest monitor-ing.In12th biennial Baltic electronic conference(BEC)(pp.161–164).Zujue Chen was born in Shang-hai,China.He received B.E.de-gree in Automobile Engineeringfrom Jiangsu College of Technol-ogy,China,in1978.Hefinishedadvanced studies in College of Com-puter Technology from HeFei Uni-versity of Technology and Col-lege of Computer Technology fromBeijing University of Posts andTelecommunications,in1982and1985,respectively.Since then,hehas been on faculty of School ofComputer Science and Telecommu-nication engineering in Jiangsu Uni-versity,China,where he is currently Professor and Dean of Communi-cations Engineering division.His research interests include embedded computing,wireless sensor network,intelligent instruments,monitor-ing and controlling system,remote wireless monitoringsystem.Zhan Shi was born in Zhenjiang,Jiangsu Province,China.Birthday:August,1985.She received B.E.de-gree in Communication Engineer-ing from Jiangsu University.She is a graduate student of Com-puter Science and CommunicationEngineering College,Jiangsu Uni-versity.Her research interests in-clude wireless sensor network,em-bedded remote wireless monitoringand controllingsystem.Qiuyue Guo was born in JiangsuProvince,China.Birthday:July,1986.She received B.E.degree inCommunication Engineering fromJiangxi University of Science andTechnology.She is a graduate student of Com-puter Science and CommunicationEngineering College,Jiangsu Uni-versity.Her research interests in-clude wireless sensor network,re-mote wireless monitoring and con-trolling system.。