电池组
计算机科学
电池(电)
负载平衡(电力)
拓扑(电路)
荷电状态
占空比
MATLAB语言
电压
电气工程
网格
工程类
数学
功率(物理)
物理
操作系统
量子力学
几何学
作者
Neha Khan,Chia Ai Ooi,Shreasth,Abdulrahman Alturki,Mohd Khairunaz Mat Desa,Mohammad Amir,Ashraf Bani Ahmad,Mohamad Khairi Ishak
标识
DOI:10.1038/s41598-024-68226-9
摘要
In a Battery Management System (BMS), cell balancing plays an essential role in mitigating inconsistencies of state of charge (SoCs) in lithium-ion (Li-ion) cells in a battery stack. If the cells are not properly balanced, the weakest Li-ion cell will always be the one limiting the usable capacity of battery pack. Different cell balancing strategies have been proposed to balance the non-uniform SoC of cells in serially connected string. However, balancing efficiency and slow SoC convergence remain key issues in cell balancing methods. Aiming to alleviate these challenges, in this paper, a hybrid duty cycle balancing (H-DCB) technique is proposed, which combines the duty cycle balancing (DCB) and cell-to-pack (CTP) balancing methods. The integration of an H-bridge circuit is introduced to bypass the selected cells and enhance the controlling as well as monitoring of individual cell. Subsequently, a DC-DC converter is utilized to perform CTP balancing in the H-DCB topology, efficiently transferring energy from the selected cell to/from the battery pack, resulting in a reduction in balancing time. To verify the effectiveness of the proposed method, the battery pack of 96 series-connected cells evenly distributed in ten modules is designed in MATLAB/Simulink software for both charging and discharging operation, and the results show that the proposed H-DCB method has a faster equalization speed 6.0 h as compared to the conventional DCB method 9.2 h during charging phase. Additionally, a pack of four Li-ion cells connected in series is used in the experiment setup for the validation of the proposed H-DCB method during discharging operation. The results of the hardware experiment indicate that the SoC convergence is achieved at ~ 400 s.
科研通智能强力驱动
Strongly Powered by AbleSci AI