Abstract: ObjectivesTo extend the operational lifespan of fuel cells and energy storage systems in complex operating conditions for ships, the aim of this paper is to construct a composite energy storage system and design an efficient energy management strategy. Methods Firstly, Fourier analysis and wavelet packet decomposition are conducted on the ship's load power to perform an initial power distribution and optimize the charging management. Then, a power adjustment factor is set, and the composite energy storage power after the initial power distribution is further optimized through heuristic rules and genetic algorithms. Next, a fuzzy neural network model is constructed to predict the power adjustment factor by inputting the composite energy storage power and SOC and other characteristic parameters, forming a three-level real-time energy management strategy. Finally, a ship power system model is built in the Matlab/Simulink platform for simulation experiments to verify the strategy. Results The simulation results show that this strategy effectively reduces the output power fluctuations of fuel cells and lithium batteries and significantly improves the ship's power quality, reducing the maximum fluctuation rate of fuel cells by 36% within 10 seconds and reducing the lithium battery loss by 13.91% compared to the prototype ship. Conclusions The research results provide a feasible technical path and strategy reference for extending the lifespan of ship fuel cells and energy storage systems.