State-of-Charge Estimation Of Lithium Battery For Electric Vehicle

Global automotive industry is facing a huge challenge of energy shortage and environmental pollution, the development of electric vehicles has become an important measure to promote the strategic transformation of the traditional auto industry. Battery as the energy source of electric vehicles, the accurate estimation of SOC is one of the core technology of the development of electric vehicles, and it can theoretically avoid the battery overcharge and over discharge, can bring security for the safety of the driver. The main contents of this paper are as follows:1.Through the analysis and comparison of the various types of commonly used battery’s performance, lithium-ion battery was identified in this paper to realize SOC estimation method in the research, and research status of various estimation methods were analysised,finally the algorithm based on Kalman filter was selected to estimate battery SOC.2.This paper was based on second-order Thevenin equivalent circuit model to complete on lithium-ion battery SOC estimation, through the discharge test of battery modules in Simulink, and based on the least square method as the basic criterion, the parameters in the model were identified, the model establishment and verification were completed.3.Combined with the built second-order Thevenin equivalent circuit model and its parameter identification results, respectively, and the application of extended Kalman filtering algorithm and unscented Kalman filter algorithm of SOC estimation, and for the problem of state noise and observation noise variable, an adaptive covariance matching method was introduced in this paper to adjust error.Application of MATLAB in the simulation experiment, the result shows that, under the same conditions of lithium-ion battery SOC estimate, unscented Kalman filter algorithm is more accurate and with adaptive covariance matching algorithm for the SOC estimation is more effective.4.In the advanced vehicle simulation software ADVISOR (Advanced Vehicle SimulatOR) simulation environment, the choice of electric vehicle model vehicle parameters and simulation parameters were input, parameters and model were configurated, CYC-UDDS and CYC-NYCC condition for road simulation test were choosed in this paper, the data obtained in the test current to verify the extended Kalman filter and unscented Kalman filtering estimation of lithium-ion battery SOC validity, the test results show that the electric vehicle in the actual environment, current conditions, the two algorithms can estimate the SOC better, unscented Kalman filtering algorithm has obvious advantages, the estimation effect of SOC with the adaptive covariance matching algorithm has significantly improved.