Research On State-of-charge And Health Estimation Of Lithium-ion Batteries

Under the background of environmental pollution and energy crisis,the development of new energy electric vehicles has become one of the effective measures to solve the problems,and high-performance lithium-ion batteries,as the main power source of electric vehicles,have also been widely researched and developed.In this paper,for a NCM811 / graphite soft-pack battery,the research and analysis are carried out in the aspects of experimental characteristics analysis,model and parameter identification,online state of charge(SOC)and state of health(SOH)joint estimation,and offline SOH estimation.The main research contents and results are as follows:First,the performance test of lithium-ion battery was carried out,the main performance indicators and influencing factors of the battery were analyzed,and the open-circuit voltage test,capacity test,internal resistance test and aging cycle test were carried out at different temperatures.The changing curve of the various parameters of the battery with temperature and SOC,as well as the capacity decay curve in the process of battery aging,and the working characteristics of the battery are mastered.Secondly,the commonly used battery models are analyzed,and the Thevenin equivalent circuit model is selected to establish the mathematical model considering the model accuracy and calculation complexity.Based on the battery discharge data,the parameters of the model are identified,and the curve of the model parameters with temperature and SOC is obtained.Thirdly,the basic principle of the extended Kalman filter algorithm is introduced,and a dual extended Kalman filter algorithm based on multiple time scales is proposed to realize the joint estimation of SOC and SOH.The SOC estimation adopts the micro scale and the SOH estimation adopts the macro scale,which greatly reduces the algorithm The amount of calculation.The algorithm has been verified under constant temperature discharge conditions and dynamic conditions of 1C and 0.33 C at different temperatures.The SOC estimation error remains within 3%,and the capacity estimation error remains within 1 Ah.The performance of the algorithm is analyzed under the condition of deviation in the initial state,and the robustness of the algorithm is verified.Finally,according to the capacity increment analysis method,the aging attenuation mechanism of the battery is studied,and the characteristic factors that can characterize the battery decay state are extracted from the battery capacity increment curve.A SOH offline estimation algorithm based on Gaussian process regression is proposed.The characteristic factor in the IC curve is used as the input of the model,and the maximum available capacity of the battery is used as the output of the model,combined with a large amount of historical battery operating data to train the model,the algorithm is verified at different temperatures,and the capacity estimation error under different training conditions Both can be kept within 1.5%.Re-select the type ? battery for verification.The experimental results show that the model is also universally applicable to non-linear capacity decay batteries.