一种改进T-S模糊神经网络估计锂电池SOC的方法  

An improved fuzzy neural network method based on T-S model to estimate state of charge of lithium batteries

作　　者：钱建文 杜翀[2] 田欣[2] 朱士彬 QIAN Jan-wen;DU Chong;TIAN Xin;ZHU Shi-bin(School of Microelectronics,University of Chinese Academy of Sciences,Beijing 101408,China;Security Products Laboratory,Shanghai Advanced Rescarch Institute,Chinese Academic of Sciences,Shanghai 201210,China)

机构地区：[1]中国科学院大学微电子学院,北京101408 [2]中国科学院上海高等研究院安全产品实验室,上海201210

出　　处：《电源技术》2020年第9期1270-1273,共4页Chinese Journal of Power Sources

基　　金：中国科学院战略性先导科技专项(C类)(XDC02070800)。

摘　　要：针对传统模糊神经网络估计锂电池荷电状态(SOC)方法精度低、收敛速度慢的问题,采用模糊规则优化算法对神经网络的结构进行优化,加快了网络的收敛速度。通过分析锂电池实际使用工况,将影响电池当前容量的两个参数,即循环次数与循环之间的静置时间,与电池电压、电流、温度统一作为影响SOC估计精度的因子。MATLAB仿真结果表明,改进后的模糊神经网络算法的精度和收敛速度较传统的模糊神经网络算法更优。Aiming at the problems of low accuracy and slow convergence speed of the traditional fuzzy neural network for estimating the state of charge(SOC) of lithium batteries,a fuzzy rule optimization algorithm was adopted to accelerate the convergence speed of the network by optimizing the structure of the neural network.By analyzing the actual operating conditions of the lithium batteries,two parameters that affect the current capacity of the battery,namely the number of cycles and the standing time between cycles,the battery voltage,current and temperature were unified as the factors affecting the accuracy of the SOC estimation.The MATLAB simulation results show that the accuracy and convergence speed of the improved fuzzy neural network algorithm are better than the traditional fuzzy neural network algorithm.

关 键 词：T-S模型 模糊神经网络 锂电池 荷电状态估计 容量衰减 

分 类 号：TM912.9[电气工程—电力电子与电力传动]