Progressive intelligence estimation of SOC based on multiple models

[摘要] In order to accurately estimate stage of charge(SOC) of the electric car lithium-ion battery, the paper used a variety of equivalent circuit model and built space state equation, through real-time online with forgetting factor recursive least squares identification battery model parameters, dynamic real-time update battery model state equation, the experimental condition to make use of the Matlab simulation, based on the battery circuit model of joint (FFRLS - EKF) algorithm, and joined the battery stage of health(SOH), The average error of the obtained SOC estimate is less than 1.8i% and the maximum error is less than 3%.Finally, the accuracy of FFRL-SEKF-SOC is verified, and the error accumulation problem is solved.

[发布机构] Guilin University of Electronic Science and Technology, Guilin City, China^1

[学科分类] 环境科学（综合）

[关键词] Circuit modeling;Equivalent circuit model;Error accumulation;Experimental conditions;Forgetting factors;Matlab simulations;Real-time updates;Recursive least-squares identifications