Real-time power optimization strategy for fuel cell ships based on improved genetic simulated annealing algorithm

Abstract

In alignment with the global pursuit of energy conservation and emission reduction, the utilization of hydrogen fuel cells (FCs) shipboard vessels are increasingly recognized as a potent means to mitigate environmental impact. However, despite its promising potential, the elevated costs associated with FCs and hydrogen fuel currently pose a significant challenge to market competitiveness. To maximize the economic viability of FC-Battery hybrid system in maritime applications, this study introduces an innovative real-time energy management strategy (EMS) that refines the Genetic Simulated Annealing Algorithm (G-SA) to mitigate FC degradation and enhance overall system cost-effectiveness by integrating the concept of Equivalent Hydrogen Consumption (EHC). The crux of this strategy lies in the implementation of a Nonlinear Autoregressive Neural Network model (NAR-NET), which facilitates the real-time forecasting of vessel load demands. By embedding the EHC of the hybrid power system and the degradation factor of FCs into the objective function, the strategy employs G-SA to determine the optimal power allocation in real-time. Compared with the Equivalent Consumption Minimization Strategy (ECMS), the proposed strategy is shown to achieve a 13 % to 30 % reduction in equivalent fuel consumption and approximately a 34 % slowdown in the rate of performance degradation.