Abstract: This paper proposed a multi-objective model predictive control for on-board energy storage system in rail transit. The energy storage system included lithium-ion batteries and ultracapacitors. The main objective of the controller is to reduce the system losses and keep the energy storage elements stayed in suitable limit. To do this, the system losses, ultracapacitor voltage, battery current fluctuation and internal resistance variation were considered in the multi-objective cost function. Additionally, a two-level optimization method based on fuzzy control and particle swarm optimization was proposed to design the weight coefficients in the cost function. Moreover, a simplified implementation method based on pre-calculated look-up table according to driving distance was proposed to improve the controller practicability. Finally, simulation results show that the proposed control strategy is feasible and effective, and has high energy efficiency