摘要
为了提高锂电池SOC估算的精度,采用改进粒子滤波算法。首先在Thevenin模型的基础上考虑了电流漂移和温度对SOC估算的影响,并对模型参数求解,同时校正了锂电池SOC估算模型,减少了计算误差,使得SOC估算更加精确;通过UKF算法更新粒子,比对权值大小,只有权值大的粒子才能够进入复制组被重新采样,小的则被抛弃,进入复制组的粒子通过线性函数生成新粒子,如果抛弃组粒子数目大于复制组粒子时,循环使用抛弃组粒子;最后给出了算法流程。试验结果表明,改进算法提高了SOC估算精度,本文模型结果与试验标准结果的误差能够控制在较小的范围内,最大误差为1.846%,明显低于采用卡尔曼滤波和粒子滤波算法的SOC估计误差。
In order to improve the accuracy of SOC estimation of lithium batteries,an improved particle filter algorithm is proposed. Firstly,model is considered the current drift and temperature effect on SOC estimation based on the Thevenin,and the parameters was calculated. The lithium battery SOC estimation model was then corrected,the calculation error was reduced,so it made the SOC estimation more accurate; Secondly,the UKF algorithm was updated the particles,and after comparing the weights,only larger weight particles could be copied into the sampled group,smaller ones were abandoned. The sampled group particles were used to generate new particles through a linear function.If the number of discarded particles was greater than that of the replication group,the loop was used the discard group.Finally,the process was given. The experimental results show that the SOC estimation accuracy was improved,the error between the model results and the experimental results could be controlled in a small range. The maximum error is 1.846%,and the SOC estimation error is lower than Kalman filter and particle filter algorithm significantly.
出处
《实验室研究与探索》
CAS
北大核心
2018年第1期134-138,共5页
Research and Exploration In Laboratory
基金
河南省科技厅鉴定项目(豫科鉴委字(2015)第971号)