基于SSA-BPNN的锂离子电池SOH估算

Estimation of Lithium-ion Battery SOH Based on SSA-BPNN

锂离子电池已被广泛应用于储能系统与电动汽车中,精确地估算锂离子电池健康状态SOH(state-of-health)是保证系统安全可靠运行的必要条件。从容量的角度分析SOH,在恒流-恒压CC-CV(constant current-constant voltage)充电电压和温度曲线中提取了7个健康特征HI(health indicator)作为输入,基于数据驱动法提出了麻雀搜索算法-反向传播神经网络SSA-BPNN(sparrow search algorithm-back propagation neural network)的锂离子电池SOH估算方法,并应用数据增强进一步提高模型的鲁棒性,最终在NASA锂离子电池随机使用数据集上进行验证。通过与未采取数据增强的传统BP神经网络相比,获得SOH估算精度有明显提升,测试集SOH估算的最大绝对误差和均方根误差分别小于3%和1.32%,实验结果表明该方法兼顾误差小,收敛快,全局搜索能力且能够适应电池老化差异特性。

Since lithium-ion batteries have been widely applied in energy storage systems and electric vehicles,the accurate estimation of their state-of-health(SOH)is a necessary condition for ensuring the reliable and safe operation of the system.SOH is analyzed from the perspective of capacity,with seven health indicators which are extracted from the constant current-constant voltage charging voltage and temperature curves as input.Based on the data-driven method,a sparrow search algorithm-back propagation neural network(SSA-BPNN)SOH estimation method for lithium-ion batteries is proposed,and data enhancement is applied to further improve the model’s robustness.Finally,this method is verified on the NASA Randomized Battery Usage Dataset.Compared with the traditional BP neural network without data enhancement,the SOH estimation accuracy of the proposed method is significantly improved.The maximum absolute error and root mean square error of SOH estimation on the test set are less than 3%and 1.32%,respectively.Experimental results show that this method has advantages of small error,fast convergence,global search capability and adaptation to different characteristics of battery aging.