电动车蓄电池荷电状态估计的支持向量机方法

Support Vector Machine Based SOC Estimation for Electric Vehicles

结合电动车蓄电池容量判断问题,将支持向量机方法用于蓄电池荷电状态估计。针对蓄电池本身的非线性特性,使用核函数为非线性核的最小二乘支持向量机算法完成估计器的设计,得到了基于多项式核函数和径向基核函数的估计器。通过实验分析了两种核函数对估计器性能的影响。从实际应用出发,分析了如何合理简化估计器模型的复杂性。结果表明基于多项式核函数的估计器精度较低,但应用过程简单;基于径向基核函数的估计器精度较高,但其应用时需要存储部分训练数据,增加了应用的复杂性。具体应用可以根据实际环境,合理选择核函数,提高估计器的综合性能。

A support vector machine approach is used to estimate the battery＇s state of charge of the electric vehicles. According to a battery is a nonlinear system, nonlinear support vector machines with polynomial kernel and radial basis function kernel are developed for the estimation of the state of charge with least square support vector m.achine algorithm. For the goal of practice, the paper gives the simplified method to decrease the estimator＇s complexity. The results have showed that the estimator with polynomial kernel is less accurate but more agreeable to application, the estimator with radial basis function kernel is more accurate but less agreeable to use. One can select optimal kernel to improve the performance of estimation in practice.