电池管理系统算法综述  

Overview of electric vehicle battery management system algorithms

作　　者：罗承东 吕桃林[1] 解晶莹[1] 付诗意 吴磊 LUO Chengdong;LV Taolin;XIE Jingying;FU Shiyi;WU Lei(Shanghai Institute of Space Power-Sources,Shanghai 201100,China)

机构地区：[1]上海空间电源研究所,上海201100

出　　处：《电源技术》2021年第10期1371-1375,共5页Chinese Journal of Power Sources

基　　金：国家重点研发计划资助(2018YFB0104400)。

摘　　要：锂离子电池由于其高能量密度和使用寿命长等优点成为储能的首选,锂离子电池系统的安全运行、状态估算、剩余寿命预测都由电池管理系统(BMS)管理,故BMS对电池系统的使用和稳定至关重要。主要关注以动力电池为代表的电池系统,综述了数据驱动方法的各种算法在BMS系统的应用。概述了BMS电模型的电化学模型、等效电路模型、数据驱动模型的特点,热建模的电模型、热模型、热-电耦合、热-电化学耦合模型的应用范畴。介绍了卡尔曼滤波、神经网络、向量机在SOC估计的应用,粒子群算法、HI-DD-AdaBoost.RT(不等式漂移检测自适应增强学习/阈值回归算法)、卡尔曼滤波在SOH估计的应用剩余寿命预测方面,介绍了经验预测、滤波预测、时间序列预测法。Lithium ion battery becomes the first choice of energy storage because of its high energy density and long service life.The safety,operation,state estimation and residual life prediction of lithium ion battery system are managed by battery management system(BMS),so BMS is very important for the use and stability of battery system.The battery system represented by power battery was focused on,and the application of data-driven method in BMS system was summarized.The characteristics of electrochemical model,equivalent circuit model and data-driven model of BMS electrical model,and the application scope of electrical model,thermal model,thermal electrical coupling and thermal electrochemical coupling model of thermal modeling were summarized.The application of Kalman filter,neural network,vector machine in SOC estimation was introduced,particle swarm optimization,Hi-DD-Adaboost.RT(inequality drift detection adaptive enhanced learning/threshold regression algorithm)and Kalman filter were applied to SOH estimation in terms of remaining life prediction,empirical prediction,filtering prediction and time series prediction were introduced.

关 键 词：电池管理系统 SOC SOH 数据驱动 卡尔曼滤波 寿命预测 时间序列 

分 类 号：TM912[电气工程—电力电子与电力传动]