Equalization of series connected lithium‐ion batteries based on back propagation neural network and fuzzy logic control. (1st March 2020)

• Record Type: Journal Article
• Authors: Wang, Biao; Qin, Feifei; Zhao, Xiaobo; Ni, Xianpo; Xuan, Dongji
• Is Part Of: International journal of energy research. Volume 44:Number 6(2020)
• Page Start: 4812
• Page End: 4826
• Publication Date: 2020-03-01
• Subjects: back propagation neural network -- battery equalization -- fuzzy logic control -- improved Buck‐Boost equalization circuit -- state of charge; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Power resources -- Research -- Periodicals; 621.042
• DOI: 10.1002/er.5274
• Languages: English
• ISSNs: 0363-907X

Abstract

Summary: In this article, a nondissipative equalization scheme is proposed to reduce the inconsistency of series connected lithium‐ion batteries. An improved Buck‐Boost equalization circuit is designed, in which the series connected batteries can form a circular energy loop, equalization speed is improved, and modularization is facilitated. This article use voltage and state of charge (SOC) together as equalization variables according to the characteristics of open‐circuit voltage (OCV)‐SOC curve of lithium‐ion battery. The second‐order RC equivalent circuit model and back propagation neural network are used to estimate the SOC of lithium‐ion battery. Fuzzy logic control (FLC) is used to adjust the equalization current dynamically to reduce equalization time and improve efficiency. Simulation results show that the traditional Buck‐Boost equalization circuit and the improved Buck‐Boost equalization circuit are compared, and the equalization time of the latter is reduced by 34%. Compared with mean‐difference algorithm, the equalization time of FLC is decreased by 49% and the energy efficiency is improved by 4.88% under static, charging and discharging conditions. In addition, the proposed equalization scheme reduces the maximum SOC deviation to 0.39%, effectively reducing the inconsistency of batteries. Abstract : The improved Buck‐Boost equalization circuit makes the series connected batteries form a circular energy loop, equalization speed is improved and modularization is facilitated. The second‐order RC equivalent circuit model of lithium‐ion battery and back propagation neural network is used to estimate the state of charge (SOC) of lithium‐ion batteries. Fuzzy logic control based on voltage and SOC is used to reduce equalization time and improve efficiency, which is compared with the mean‐difference algorithm.