关键词: 电动汽车, 锂离子电池, 荷电状态, 参数识别, 扩展卡尔曼滤波

Abstract:

Proposed in this paper is a SOC ( State of Charge) estimation method to correct the initial error of SOC of the lithium-ion battery pack used in electric vehicles. The new method employs an algorithm combining the extended Kalman filtering with the Ampere-hour integral method. In the investigation,first,the advantages and disadvantages of several equivalent battery circuit models were analyzed,and the improved PNGV model with double RC parallel networks was selected to perform the parameter identification of a lithium-ion battery. Then,a nonlinear dynamic model of the battery in the state-space form was constructed with the help of the improved PNGV model,and the relationship between the open circuit voltage and the corresponding SOC of the battery was analyzed based on the polynomial fitting. Finally,some experiments in the conditions of constant-current pulse discharge and ECE15 were carried out. The results indicate that the proposed algorithm effectively corrects the initial error of SOC and guarantees the estimation accuracy of SOC.

Key words: electric vehicle, lithium-ion battery, state of charge, parameter identification, extended Kalman filtering