Design Of A Capacitive Load Based Photovoltaic Array I-V Characteristic Testing System

Fast and accurate measurement of the photovoltaic(PV)current and voltage(I-V)characteristics curves is of great significance for the accurate modeling,status assessment and fault diagnosis of PV modules and arrays.To address the problems of existing I-V characteristic testing system(such as slow scanning speed,relatively low scanning power,inaccurate parameter estimation,and low reliability),a new I-V characteristic testing system is developed for PV arrays in this thesis,which is based on capacitive load and integrates parameter identification function.The main work of this thesis is as follows.(1)Based on the dynamic capacitor charging method,a new I-V characteristic curve testing scheme is proposed for PV arrays.Firstly,the charging and discharging characteristics of capacitors are analyzed,and then a Matlab/Simulink based simulation is carried out to verify the feasibility of the capacitor charging based I-V curve scanning scheme.Secondly,based on the IGBT and solid-state relay,a new switching circuit is proposed to control the charging and discharging of the capacitors for high-power PV arrays,which can overcome the shortcomings of traditional electromagnetic relays,such as short life and large switching delay.In view of the nonlinear varying of I-V during the charging process,a simple and effective segmental delay based I-V scanning method is proposed to realize relatively uniform sampling of the I-V curve.The Standard Testing Condition(STC)correction formula is used to convert the measured Operating Condition(OPC)curve to the STC curve.Then,the Lagrange parabolic interpolation method is used to interpolate the missing part of the STC curves around the open-circuit point.Finally,a temperature and irradiance sensor based on nRF24L01 wireless transmission is designed,and some other modules are also integrated,such as LCD display,a U disk interface and a rechargeable Lithium-ion battery.Thus,a complete portable photovoltaic I-V characteristic curve tester is developed,(2)An accurate single-diode model for PV array is built on the platform of Matlab/Simulink,and the impact of the internal parameters on I-V characteristic output curves is simulated and analyzed.And then,based on Artificial Bee Colony(ABC)and Nelder-Mead Simplex(NMS)algorithms,a new efficient ABC-NMS hybrid optimization algorithm is proposed to extract the parameters of the PV model by curve fitting,and the superiority of the algorithm is verified by experimental analysis and comparison.At the same time,the algorithm is ported to the embedded DSP processor(TMS320F28335)of the I-V characteristic tester to realize real-time online parameter extraction and thus expand the function of the portable I-V tester.In addition,based on the RMSE of parameter extraction,a thresholding method is proposed to effectively detect the partial shading status.Moreoever,a Matlab/GUIDE based data management software is designed to facilitate the data management and system configuration.(3)The functions and performance of the system are comprehensively tested and verified on a PV array in our laboratory.The experimental results show that the system can scan high-quality I-V characteristic curves from the PV array,and the scanning time for a curve is about 100-400 milliseconds.The computed I-V curves using the extracted model parameters are in good agreement with the measured curves under normal conditions,which validates the high accuracy of the parameter extraction algorithm.The running time of the parameter extraction algorithm on the DSP platform is about 10 seconds for an I-V characteristic curve with 100 data points.In addition,according to the rated data of the main components of the I-V tester,the maximium scanning voltage and current are up to 450V and 20A respectively.