Energy Management Strategies for Plug-in Hybrid Electric Vehicles Using Snow Geese Optimization and RBAGCN: A Review

Abstract

The rapid development of plug-in hybrid electric vehicles (PHEVs) has necessitated advanced energy management strategies (EMS) to optimize fuel efficiency, reduce emissions, and enhance battery lifespan. Traditional EMS approaches, including rule-based and dynamic programming methods, suffer from limitations such as poor adaptability to real-time conditions and high computational complexity. Recent advances integrate artificial intelligence (AI), metaheuristic optimization, and graph-based deep learning models to address these challenges. In particular, hybrid approaches combining nature-inspired optimization algorithms such as Snow Geese Optimization (SGO) with Graph Convolutional Networks (GCNs) have emerged as promising solutions for handling complex nonlinear relationships and spatio-temporal dependencies in vehicular systems. This systematic review explores recent developments in EMS for PHEVs, focusing on intelligent and hybrid techniques such as reinforcement learning, predictive control, and graph-based neural architectures. Studies from 2020–2023 are analysed to identify trends, methodologies, and performance improvements. Results indicate that data-driven and hybrid optimization approaches significantly outperform traditional techniques in fuel economy, adaptability, and real-time implementation. Furthermore, relational bi-level aggregation GCNs provide enhanced modelling of multi-source vehicular data. The review highlights current challenges, including computational overhead and real-world deployment issues, and outlines future research directions toward scalable, intelligent EMS frameworks for next-generation PHEVs.