Aiming at the problems of precise parking， punctuality， comfort and energy consumption in the process of Automatic Train Operation （ATO）， an optimization method of ATO speed curve based on GAPSO （Genetic Algorithm and Particle Swarm Optimization） algorithm was proposed. Firstly， a multi-objective optimization model of train ATO operation was established， the train passing through the neutral zone with power cutoff and coasting was included in the control strategy， and the operation control strategy was analyzed. Secondly， Particle Swarm Optimization （PSO） algorithm was improved， the nonlinear dynamic inertia weight and the improved acceleration coefficient were adopted， and the genetic operator was integrated into it to form a brand-new GAPSO algorithm， and the superiority of GAPSO algorithm in global search and local search ability as well as convergence speed was verified. Finally， GAPSO algorithm was used to optimize the operating mode changing points， and a set of operating mode changing point speeds satisfying multi-objective optimization was obtained， thereby obtaining the optimal target speed curve. Simulation experimental results show that under the premise that the overall running time meets the requirements of punctuality， the optimization method can make the energy consumption reduced by 13.29%， the comfort increased by 26.62%， and the parking error reduced by 21.62%. Therefore， the optimized train target speed curve can meet the multi-objective requirements， and this method provides a feasible solution for train ATO multi-objective optimization.