In order to overcome the frequent changes in the load during the operation of the ship, the ship power system fluctuates drastically, the fuel efficiency of the diesel generator is reduced, the economy is poor, and environmental pollution is adversely affected. The energy storage unit (lithium battery) is introduced, and the hybrid ship power system is composed of diesel synchronous generator and lithium battery. In the hybrid energy management system, a particle swarm optimization (PSO) algorithm based on the genetic algorithm (GA) idea is adopted. A dynamic matrix model predictive control strategy is proposed and simulated in MATLAB/Simulink. The simulation results show that the method can make the output of diesel synchronous generator set and lithium battery pack follow the load change well under the premise of more economical, and meet the demand of load fluctuation. The validity and feasibility of the energy management strategy proposed in this paper are verified.