Chalmers’ researchers develop AI charging strategy to extend EV battery life 23% with no increase in charge time

Researchers at Chalmers University of Technology have pioneered an AI-driven fast charging strategy that significantly prolongs the lifespan of lithium-ion batteries used in electric vehicles. Utilising reinforcement learning, the system dynamically adjusts the charging current in real time based on the battery’s current state of charge and overall health. This approach reportedly extends battery life by approximately 23% compared to conventional fast charging methods, without materially increasing the charging duration. The innovation addresses a critical issue in current fast charging practices, which apply uniform charging protocols regardless of a battery’s age or condition. Such uniformity can accelerate degradation, particularly through lithium plating—a process where metallic lithium deposits on the electrode surface rather than integrating properly into the anode. This phenomenon diminishes battery capacity and can lead to dangerous short circuits, risks that intensify as batteries age. By contrast, the Chalmers model tailors charging parameters to the battery’s evolving characteristics, mitigating these effects. According to Meng Yuan, co-author of the study and now assistant professor at Victoria University of Wellington, the method achieves comparable charging speeds to existing protocols but with substantially less long-term damage. Importantly, the strategy could be implemented via software updates to current battery management systems, avoiding the need for new hardware installations. The researchers also highlight the potential for transfer learning to adapt the AI model to different battery chemistries more efficiently, enhancing its applicability across various EV models. While the results are currently based on simulations using models of widely used EV batteries, the team plans to validate the approach through physical battery testing. The study, co-authored by Professor Changfu Zou of Chalmers’ Department of Electrical Engineering, was published in the IEEE Transactions on Transportation Electrification. If successful in real-world trials, this AI charging strategy could represent a significant step forward in extending EV battery longevity and improving sustainability in electric mobility.