Research On Electric Vehicle Insulation Detection System Based On Variable Forgetting Factor Recursive Least Square Algorithm

As the energy crisis and environmental pollution problems become more and more serious,people pay more and more attention to electric vehicles.Governments and car companies around the world have planned the development of electric vehicles.Electric vehicles use power batteries as their power source.In order to meet power requirements,their voltages are generally above 300 V.Higher voltage levels also place higher requirements on the insulation performance of the vehicles.However,the complex working environment of electric vehicles easily leads to a decrease in the insulation performance between the power battery and the chassis,posing a threat to the safety of drivers and passengers.Therefore,it is of great significance to accurately detect the insulation performance of electric vehicles.Based on the low-frequency signal injection method and considering the influence of the Y capacitance of the system,this paper proposes a real-time detection method of insulation resistance based on the variable forgetting factor recursive least square algorithm.By comparing the commonly used insulation detection methods,this paper establishes the circuit model when the power battery is connected based on the low-frequency signal injection method,and derives the calculation model of the positive and negative insulation resistance.By analyzing the influence of the Y capacitance of the system on the low-frequency signal injection method to measure the insulation resistance,the schematic diagram of the insulation detection circuit containing the Y capacitance is established,and the reflected wave voltage model affected by the Y capacitance is deduced.The model is linearized using the first-order Taylor expansion method,and a variable forgetting factor and an appropriate stopping criterion are constructed.An insulation resistance detection method based on the variable forgetting factor recursive least squares algorithm is proposed.Based on this method,the hardware and software design of the insulation detection system is carried out.Taking Freescale MC9S08DZ60 as the main control chip,the hardware circuit is divided into six modules:MCU(microcontroller unit)control unit,power conversion unit,injection signal generator,signal sampling module,CAN(Controller Area Network)communication circuit and light alarm circuit.And design the hardware circuit schematic diagram and PCB(Printed Circuit Board).The software is divided into three parts:device driver layer,signal processing layer and application layer.The device driver layer mainly includes timer programs,CAN drivers,ADC(Analog-to-Digital Converter)sampling drivers,etc.;the signal processing layer mainly includes data input programs and data output programs;the application layer programs mainly implement model parameter identification,insulation resistance calculation,and equivalent Y capacitance calculation,insulation fault level judgment and other functions.Finally,the insulation detection system was verified experimentally by building an experimental platform.In the single working condition experiment,the maximum relative error of insulation resistance R_p and R_n is within 3.5%,the maximum relative error of equivalent Y capacitance C_p//C_n is within 5%,and the measurement time is within 2.5 s.In the multi-working condition experiment,in addition to the error near the initial time and the switching point of the operating conditions,the relative errors of R_p,R_n and C_p//C_n in the whole process are less than 2%,and the RMSE(root mean square error)is less than 0.012.The results show that the system has high measurement accuracy and response speed,strong anti-interference ability and robustness.