This is the authors' abstract. We don't add key points for this paper.
Battery degradation is another important constraint of the hybrid electric vehicle systems, where the standard approach to energy management focuses on short-term fuel efficiency and does not consider the long-term health of the battery. In this paper, a proposal has been made on an Aging-Aware Attention-Driven Transformer Reinforcement Learning (AAT-DRL) framework of intelligent hybrid battery energy management. Transformer-based temporal feature encoding, along with actor-critic reinforcement learning architecture, is combined in the model to encode patterns that are sensitive to degradation such as temperature, current stress and depth-of-discharge. The multi-objective reward is developed with an aim to optimize the energy efficiency, thermal safety and battery life. The framework is tested at standard driving cycles (UDDS, NEDC and WLTP). In the results, it is shown that the degradation index has reduced by 38%, the SOC has become more stable and thermal fluctuations have been reduced, with a peak efficiency of 95.3%, which is a better performance than the state-of-the-art DRL methods. The solution offers a healthy, scalable and battery management system for the next generation based on the proposed approach.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Raghavendran et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: