基于麻雀搜索优化LSTM的实车动力电池SOC估计研究  

Research on Real Vehicle Power Battery SOC Estimation Based on Sparrow Search Optimized LSTM

作　　者：洪吉超 裴佳琦 梁峰伟 李萌 邱余龙[1,2] 张磊 HONG Jichao;PEI Jiaqi;LIANG Fengwei;LI Meng;QIU Yulong;ZHANG Lei(School of Mechanical Engineering,University of Science and Technology Beijing,Beijing 100083,China;Shunde Innovation School,University of Science and Technology Beijing,Foshan Guangdong 528000,China)

机构地区：[1]北京科技大学机械工程学院,北京100083 [2]北京科技大学顺德创新学院,广东佛山528000

出　　处：《西南大学学报（自然科学版）》2024年第12期41-50,共10页Journal of Southwest University(Natural Science Edition)

基　　金：国家自然科学基金项目(52107220);广东省自然科学基金项目(2024A1515012804)。

摘　　要：随着电动汽车技术的迅猛发展,准确评估电池的荷电状态(State of Charge,SOC)对于保障车辆性能和安全至关重要.针对现有SOC估计方法的不足,提出了基于麻雀搜索算法进行参数优化的长短期记忆(Long and Short-Term Memory,LSTM)神经网络模型.首先通过实车数据采集与预处理,构建了包含多种实车运行参数的数据库,并利用信息熵和互信息理论对数据进行特征筛选,以识别与SOC高度相关的特征.然后将筛选出的特征输入到长短期记忆神经网络模型中,并使用麻雀搜索算法对模型参数进行优化.研究结果表明:该模型在多种驾驶工况和不同的充放电环境下均能实现高精度的SOC估计,验证了SOC估计的准确性和鲁棒性,为电池管理系统的发展提供了有力的技术支持.With the rapid development of electric vehicle technology,accurately assessing the state of charge(SOC)of the battery is crucial for ensuring vehicle performance and safety.In response to the shortcomings of existing SOC estimation methods,this study proposes a long and short-term memory(LSTM)neural network model based on the sparrow search algorithm(SSA)for optimization of the parameters.The research first constructed a database that included a variety of vehicle operation parameters through real vehicle data collection and preprocessing.Then,the data was feature-filtered using information entropy and mutual information theory to identify features highly related to SOC.Next,the selected features were input into the LSTM neural network model,and the sparrow search algorithm was used to optimize the model parameters.The results show that the model can achieve high-precision SOC estimation under various driving conditions and different charging and discharging environments.Therefore,the proposed method verifies the accuracy and robustness of SOC estimation,providing strong technical support for the development of battery management system.

关 键 词：电动汽车 荷电状态估计 数据驱动 神经网络 优化算法 

分 类 号：TK01[动力工程及工程热物理]