Referring to the idea of stochastic dynamic programming (SDP), the required drive power in the new European driving cycle(NEDC) was abstracted as a stochastic process that changed with the vehicle velocity. According to the transition probability matrix of required drive power, the issue of energy management strategy was simplified as the optimization problem of engine torque and the discrete particle swarm optimization (DPSO) was applied to optimize the gear selection. Based on the state balance of charge and the optimum of gear efficiency, the optimal engine torque was obtained through the strategy iteration method to achieve the minimal fuel consumption. The simulation was carried out with Matlab. The test results show that the proposed strategy can reduce fuel consumption by 9.42% without the loss of vehicle velocity in the NEDC cycle in comparison with that of the rulebased energy management strategy.