Urban rail train starts and brakes frequently in it’s movement. It is important to improve the utilization efficiency of electric energy and reduce the traction energy consumption in the field of metro transit. At peak hours, the overlap time between two trains in the same power supply interval is longer and there is much more renewable energy generated by the train’s braking due to a large increasement in passenger flow and the number of departure. In this paper, a method based on pareto multi-objective genetic algorithm is proposed to optimize energy consumption. By optimizing the stopping time of trains in each station, train schedule is optimized and the regenerative braking energy can be used more efficiently.