Engineering Implementation And Comparison Analysis Of Power Battery SOC Estimation

With the continuous development of electric vehicles in recent years,both in terms of energy consumption and environmental protection have obvious advantages.Breakthrough in key technologies of electric vehicles has become a research hotspot.As a powertrain for an electric vehicle,the battery pack relies on a battery management system for its efficient operation,safe use,and optimization of energy control strategies.Accurate estimation of the state of charge(SOC)of the battery pack is a function of the basic core of the battery management system.Therefore,this paper studies the engineering implementation and comparative analysis of the estimation method of the state of charge(SOC)of the lithium battery.Firstly,the effects of factors such as magnification and temperature on the voltage,capacity and internal resistance of Li-ion battery were analyzed;the SOC-OCV mathematical model was established,and the first-order RC and second-order RC equivalent circuit simulation models were established.The accuracy of the two batteries under different operating conditions is analyzed.The accuracy of the first-order RC model and the second-order RC model is similar,and they have good adaptability to the operating conditions.Considering the degree of complexity of the engineering implementation,the first-order RC model is selected.As an SOC estimation model,it lays the foundation for engineering state estimation.Secondly,the advantages and disadvantages of the SOC estimation method are compared and analyzed.The different algorithm models are built in Matlab and simulated based on the battery condition data.The simulation results of different algorithms are compared to show the PI observer.Kalman filter has higher precision,better robustness and adaptability to "working conditions,and can accurately track true values quickly.Under the same conditions,the accuracy of the PI observer is slightly higher,but the fluctuation is slightly greater when it is affected by noise.In order to solve the problem of low precision at the low end,the PI observer+Ampere Time Integration method and Kalman+Ampere Time Integration method are proposed.The simulation proves the effectiveness of the method.A hardware-in-the-loop verification platform for battery state-of-charge(SOC)estimation algorithms was built.The principle of the acquisition of the physical quantity of the control board power supply,the control board CAN transceiver,current and voltage,and the selection basis are described.The software design,C language implementation of the algorithm,CAN communication program design and real-time monitoring program design are completed.Finally,the physical structure of the experimental platform was achieved.Finally,experimental experiments were carried out on 25Ah Li-ion cells and 25Ah LTO cells under different working conditions,verifies the effectiveness,reliability,and accuracy of estimating State of Charge(SOC)for PI observer and Kalman filter algorithms.The effectiveness of the proposed correction method for PI observer+Ah and Kalman filter+Ah was verified,and the SOC estimation error was controlled within 5%.The advantages and disadvantages of the two algorithms for engineering applications are compared and analyzed.