基于贝叶斯最小二乘支持向量机的电池SOC预测  

Battery SOC prediction based on LS-SVM within Bayesian evidence framework

作　　者：井娥林[1] 孙正凤[1] 温宏愿[1] 

机构地区：[1]南京理工大学泰州科技学院,江苏泰州225300

出　　处：《电源技术》2015年第12期2616-2619,2642,共5页Chinese Journal of Power Sources

基　　金：江苏省高校研究生创新计划项目(CXZZ13_0683)

摘　　要：针对混合动力汽车电池电能容量判别问题,将最小二乘支持向量机方法应用于混合动力汽车电池荷电状态的预测。考虑到最小二乘支持向量机的参数选择会对预测结果产生较大的影响,提出了基于贝叶斯证据框架优化的最小二乘支持向量机预测方法。通过贝叶斯证据框架自动调整正则化参数和核参数,更好地实现了最小化误差和模型复杂性之间的折中。将电池的工作电压、工作电流和表面温度参数用来预测蓄电池的荷电状态实时值,在美国城市动态驱动工况(UDDS)条件下进行实验验证,结果表明:所设计预测模型具有较高的精度,能够实时准确地预测出SOC值,实用性强,有效性高。The least square support vector machines（LS-SVM） was proposed based approach to predict the battery＇s state of charge（SOC） of the hybrid electric vehicles（HEV）. In consideration of that parameter selection of support vector machines exerts a major influence on SOC predict, this paper presented a SOC prediction algorithm on the basis of Bayesian evidence framework optimized least square support vector machines. Within the evidence framework, the regularization and kernel parameters could be adjusted automatically, to achieve a better tradeoff between the minimum error and model＇s complexity. The working voltages, currents and surface temperature of battery were used to predict the real-time value of SOC. Experiment was verified under different working conditions of urban dynamometer driving schedule（UDDS）, and the results indicated that the prediction model possessed higher predicted accuracy, achieved real-time and accurate SOC prediction.

关 键 词：SOC预测 贝叶斯证据框架 最小二乘支持向量机 混合动力汽车 

分 类 号：TM912.9[电气工程—电力电子与电力传动]