An efficient hybrid ladybug beetle and heterogeneous context-aware graph convolutional network for optimizing energy management in fuel cell hybrid electric vehicle.

This article focuses on a novel hybrid approach for optimizing energy management (EM) in Fuel Cell Hybrid Electric Vehicles (FCHEVs) by integrating Ladybug Beetle Optimization (LBO) and Heterogeneous Context-Aware Graph Convolutional Network (HCAGCN). The LBO algorithm dynamically allocates power between the fuel cell (FC) and battery to reduce hydrogen (H₂) consumption, while HCAGCN predicts future energy demands based on driving patterns and environmental factors. Implemented in MATLAB and tested against existing methods, the LBO–HCAGCN approach demonstrated superior performance with the lowest H₂ consumption (161.23 g), reduced operating costs ($564), and highest efficiency (92%). Limitations include sensitivity to rapid environmental changes, data quality issues, scalability challenges, and computational demands, with future work suggested to enhance adaptability, data processing, scalability, and computational efficiency.