Optimal energy management strategies for hybrid electric vehicles: A recent survey of machine learning approaches

Abstract

Hybrid Electric Vehicles (HEVs) have emerged as a viable option for reducing pollution and attaining fuel savings in addition to reducing emissions. The effectiveness of HEVs heavily relies on the energy management strategies (EMSs) employed, as it directly impacts vehicle fuel consumption. Developing suitable EMSs for HEVs poses a challenge, as the goal is to maximize fuel economy yet optimize vehicle performance. EMSs algorithms are critical in determining power distribution between the engine and motor in HEVs. Traditionally, EMSs for HEVs have been developed based on optimal control theory. However, in recent years, a rising number of people have been interested in utilizing machine-learning techniques to enhance EMSs performance. This article presents a current analysis of various EMSs proposed in the literature. It highlights the shift towards integrating machine learning and artificial intelligence (AI) breakthroughs in EMSs development. The study examines numerous case studies, and research works employing machine learning techniques across different categories to develop energy management strategies for HEVs. By leveraging advancements in machine learning and AI, researchers have explored innovative approaches to optimize HEVs’ performance and fuel economy. Key conclusions from our investigation show that machine learning has made a substantial contribution to solving the complex problems associated with HEV energy management. We emphasize how machine learning algorithms may be adjusted to dynamic operating environments, how well they can identify intricate patterns in hybrid electric vehicle systems, and how well they can manage non-linear behaviors.