Development of a predictive ECMS based on short-term velocity forecast for a fuel-cell hybrid electric vehicle considering component aging

This study proposes a predictive equivalent consumption minimization strategy (P-ECMS), based on short-term velocity prediction for a heavy-duty fuel cell vehicle while considering fuel cell degradation. The long-short term memory (LSTM) based predictor has been trained on data deriving from realistic driving cycles. The P-ECMS is compared with a typical adaptive-ECMS from the literature, the optimal ECMS, and a rule-based strategy for two different driving cycles in terms of battery SOC sustenance, equivalence factor evolution, hydrogen consumption, and fuel cell degradation. Results show that P-ECMS can reduce hydrogen consumption by up to 3% compared to the reference A-ECMS. It also reduces fuel cell degradation in relation to the optimal ECMS.