Owing to the increasing concerns about the release of pollution by traditional ships, the use of the renewable energy in ships' power systems is attracting much attention. However, an improperly designed renewable generation system and energy storage system (ESS) will increase costs and greenhouse gas emissions. This paper proposes a mathematical model of a photovoltaic (PV) power generation system for a ship, taking into account the effects of ship rolling. A PV system on the shipboard, unlike one on land, has to confront dramatic power fluctuations that are caused by the motions of the ship and bad weather, so hybrid ESSs play a significant role in a ship's power system. In this work, the discrete Fourier transform (DFT) is employed to decompose the required balancing power into various time-varying periodic components, which are utilized to calculate the maximum required power of the hybrid energy storage systems. A cost analysis is performed using particle swarm optimization (PSO) algorithm to optimize the size and capacity of various types of energy storage systems. Simulation results reveal the efficiency of the optimal allocation of the ESSs. (C) 2017 Elsevier Ltd. All rights reserved.