Volume 55

Ageing-Aware Deep Reinforcement Learning for Adaptive Fast Charging of Li-ion Battery Considering Coupled Degradation mechanisms Ben Shang, Yinglong He, Lei Wang, Zeyu Sun, Jianwei Shao, Constantina Lekakou, Jing Zhao, Youping Fan

Abstract

A major challenge in fast charging is to reduce the charging time without accelerating battery aging. The interaction between different aging mechanisms has not been investigated in the fast-charging optimization. To address this issue, we propose an adaptive, aging-aware fast charging method utilizing deep reinforcement learning to balance charging speed with the aging effects of multiple degradation mechanisms. This method incorporates a coupled degradation model to capture the dynamic characteristics of batteries throughout their lifecycle. Using deterministic policy gradient and Markov decision process (MDP) frameworks, the algorithm optimally adjusts charging currents based on safety constraints and internal coupled aging. We demonstrate the effectiveness of our approach through simulations using the PyBAMM model. Results indicate significant adaptability to variations in battery characteristics at different aging stages.

Keywords Optimal charging control, Coupled battery model, deep reinforcement learning