改进PSO优化核SVR的锂离子电池剩余寿命估计

Remaining Useful Life Estimation for Lithium-ion Battery based PSO Optimization with Kernel SVR

基于传统的剩余寿命估计方法准确率较低的问题,提出一种改进PSO优化核SVR的锂离子电池剩余寿命估计(IPSO-KSVR)方法.首先构建以充放电循环次数为输入,以电池容量为输出的核SVR模型,然后用改进的PSO算法优化惩罚因子C和核方差g,最后在筛选的NASA数据集上进行训练与测试.实验结果表明,与传统的KSVR方法相比,该方法的电池容量估计MSE、RMSE、MAE和MAPE指标的百分比误差分别减少66%、42%、58%和63%,RUL预测百分比误差减少64%,优于其它PSO优化的KSVR方法.该方法具有重要的理论与应用价值.

Remain useful life estimation of lithium-ion battery is the basis to ensure the safe and reliable operation of electric equipment or system.The traditional remain useful life estimation method has low accuracy.To this end,this paper proposes an ipso-ksvr method for optimizing kernel SVR by improved PSO.This method firstly constructs a kernel SVR model with the number of charge and discharge cycles as the input battery capacity as the output,then the improved PSO algorithm is applied to optimize the penalty factor C and the kernel variance g,and finally it carries out training and testing on the selected NASA data set.The experimental results show that the method proposed in this paper reduces the battery capacity estimation percentage error indexes of MSE,RMSE,MAE and MAPE by 66%,42%,58%and 63%respectively,and the percentage error of RUL prediction is reduced by 64%compared with the traditional KSVR method.It is also better than other KSVR methods optimized by PSO,and it has important theoretical and practical application value.