基于改进RBF算法的蓄电池荷电状态估计研究  

Research on Battery SOC Estimation Based on Improved RBF Algorithm

作　　者：汪刘峰 慈兆会 李翔 叶伟 李剑卿 Wang Liufeng;Ci Zhaohui;Li Xiang;Ye Wei;Li Jianqing(Anhui Nanrui Jiyuan Power Grid Technology Co.,Ltd.,Hefei 230031,China)

机构地区：[1]安徽南瑞继远电网技术有限公司,合肥230031

出　　处：《单片机与嵌入式系统应用》2023年第11期63-67,79,共6页Microcontrollers & Embedded Systems

基　　金：南瑞集团有限公司—智慧型站用交直流电源关键设备研制(SGNR0000KJJS2106316)。

摘　　要：针对径向基神经网络在SOC预测过程中随机产生基函数中心和宽度的不稳定性导致预测精度不佳的问题,提出一种改进RBF的蓄电池SOC估计方法,以麻雀搜索算法优化RBF网络参数以提高网络预测精度。在DST工况仿真验证SSA参数优化的有效性,在US06和FUDS工况下,分别利用改进径向基神经网络、RBF、极限学习机与BP神经网络对SOC进行预测。对比分析结果表明,SSA RBF在SOC估计精度方面表现更优,将估计误差降低到2%以内,能够完成高精度的SOC估计,具有一定的理论研究意义与应用价值。To address the issue of poor prediction accuracy caused by the instability of randomly generated center and width of basis func-tions in Radial Basis Function(RBF)during SOC prediction,an improved RBF-based SOC estimation method is proposed.The method optimizes RBF network parameters using the Sparrow Search Algorithm(SSA)to enhance the prediction accuracy of the network.The effectiveness of SSA parameter optimization is validated through simulation experiments under the DST operating condition.Further-more,under the US06 and FUDS operating conditions,the SOC prediction is conducted using the improved Radial Basis Function(SSA RBF),RBF,Extreme Learning Machine(ELM),and Backpropagation Neural Network(BPNN).Comparative analysis of the results demonstrates that SSA RBF outperforms in terms of SOC estimation accuracy,reducing estimation error to within 2%and achieving high-precision SOC estimation.This approach holds significant theoretical research significance and practical value.

关 键 词：蓄电池 SOC估计 RBF神经网络 SSA算法 

分 类 号：TM912[电气工程—电力电子与电力传动]