SOC Estimation Of Lithium-ion Battery Based On Unscented Kalman Filter

With the increasingly severe global energy and environmental protection situation,electric vehicles have become an important direction for the development of the automotive industry in various countries due to their advantages of low pollution and high energy conversion efficiency.During the use of electric vehicles,accurate estimates of the state of charge of power batteries is important to maintain the smooth running state of power batteries and improve energy conversion efficiency.On the basis of studying the characteristics and parameter identification of lithium-ion power battery,this paper studied the online estimation of lithium-ion battery SOC.By comparing the advantages and disadvantages of lithium-ion battery equivalent circuit models,and based on the characteristics of lithium-ion batteries,a second-order RC equivalent circuit model was selected for modeling.The parameters of the equivalent model were identified by offline identification method,and the accuracy of the model was verified by the method of loading the working current,which showed that the offline identification method had certain limitations.Then,based on the second-order RC equivalent model,this paper used the recursive least squares algorithm with forgetting factor to perform online parameter identification.In order to verify the accuracy of the equivalent model,the current under the working condition was also loaded on the established simulation model,and then the output voltage of the model was compared with the voltage of the simulation module in the software to obtain the accuracy of the equivalent model.The model of identification parameters can well reflect the dynamic characteristics of lithium-ion batteries.Finally,this paper added the innovation adaptive covariance matching process based on the Unscented Kalman Filter algorithm.it speeded up the convergence of thealgorithm compared with the traditional extended Kalman filtering algorithm,and also matched by noise information compared with the unscented Kalman filtering.The proposed method improved the algorithm’s ability to track the actual state of the system,and improved the estimation accuracy by about 3.5%.The results showed that the SOC estimation method proposed in this paper improved the estimation accuracy and contained certain reference value.