AI enabled fast charging of lithium-ion batteries of electric vehicles during their life cycle: review, challenges and perspectives

Gradually replacing conventional fuel vehicles with electric vehicles (EVs) is a crucial step towards achieving energy saving and emission reduction in the transportation sector. The large-scale adoption of EVs depends on the rapid energy replenishment of lithium-ion batteries (LIBs). Fast charging (FC) is crucial for the rapid energy replenishment of LIBs. The performance of FC is influenced by multiple factors, including battery design, critical state estimation, and the design of FC control strategies. However, there is a lack of comprehensive reviews that elucidate the limiting factors of battery FC, the critical state estimation methods, and the design of charging control strategies. This paper addresses this gap by first elucidating the limiting mechanisms of battery FC and analyzing the factors affecting FC from both internal and external perspectives. Secondly, this study conducted a comprehensive investigation of diverse battery models tailored for state estimation and FC control, critically assessing their respective advantages and limitations. Thirdly, it provides an in-depth analysis of the key states that are crucial during the battery FC process and systematically examines various state estimation methods for key states. Furthermore, this study critically examines the challenges in rule-based, model-based, and machine learning (ML)-based FC control strategies, elucidating their respective limitations. Finally, based on the investigation of the current state and challenges of FC technology, the prospects for enhancing battery FC performance using artificial intelligence (AI) methods, such as deep learning (DL), deep reinforcement learning (DRL), and Bayesian optimization (BO), have been outlined. This comprehensive review aims to provide valuable insights and guidance for future advancements in battery FC performance. This review presents a thorough investigation of factors affecting fast charging, battery modeling, key state estimation and fast charging control strategies and provides a forward-looking perspective on AI enabled fast charging technology of LIBs.