An optimal power management strategy for enhancing fuel economy in fuel cell-battery hybrid electric vehicles

Abstract

Fuel Cell-Battery Hybrid Electric Vehicles (FCBHEVs) are emerging as a promising solution for sustainable transportation, offering high efficiency and zero tailpipe emissions. Optimizing power distribution between energy sources is essential for improving fuel economy. This paper introduces a hybrid approach that combines optimization techniques with machine learning (ML). It uses LSTM networks for real-time estimation of key battery states, such as State of Charge (SoC), enabling informed decision-making. Additionally, an advanced optimisation layer utilises a weighted multi-objective cost function to minimise system costs and weights while maintaining power balance and operational constraints. To further guide energy sharing between the battery and fuel cell, a Hybrid Storage Participation Index (HSPI) is introduced, quantifying the relative contribution of each energy source over a drive cycle. The HSPI approach aims to improve fuel economy and reduce fuel consumption per 100 km, and also dynamically allocates power demand between the fuel cell and the battery according to real-time operating conditions. The results across various drive cycles demonstrate significant improvements in fuel economy, with reductions of up to 70–73 % compared to conventional rule-based strategies. Furthermore, the proposed strategy enhances vehicle fuel efficiency—measured in litres per 100 km—ensuring better energy utilization and extended system longevity.