SOC estimation of Lithium battery by UKF algorithm based on dynamic parameter model

The accurate SOC estimation of lithium ion battery is not only a prerequisite for the effective use of batteries, but also one of the key technologies to be solved in battery management system. Improving the accuracy of SOC estimation for the lithium ion battery is important for prolonging the life span of the battery and improving the utilization rate of the battery. As an important parameter in the process of charging and discharging, SOC will be influenced by the charge discharge rate, charge discharge efficiency, self-discharge rate, charge and discharge cycle number, temperature and other factors, which make it difficult to guarantee the accuracy of SOC estimation. In this paper, we choose the most promising lithium iron phosphate battery as the research object, the charging and discharging characteristics of lithium iron phosphate battery were studied by experiments and the relationship between the SOC and the open circuit voltage was obtained. According to the discharge mechanism of lithium iron phosphate battery and the relationship curve of OCV-SOC, based on equivalent circuit of second-order RC model with on-line identification of the model parameter by limited memory recursive least squares algorithm, the model of the dynamic parameters of the battery is established; In order to overcome the linearization error of EKF algorithm and control calculation, using UKF algorithm to estimate SOC of battery based on the dynamic model. The parameters of the model were identified, and SOC was estimated by using A-H measurement method and UKF algorithm in MATLAB. Simulation results show that, model parameters are time-varying and dynamic parameter model is more in line with the battery discharge, the estimation result by UKF algorithm is roughly the same with the result by AH method. The method has certain application value in engineering for high accuracy.