For the pursuit-evasion game problem with target constraints, this paper designs an intelligent maneuvering strategy based on the Soft Actor-Critic algorithm. First, a kinematic dynamics model of the pursuit-evasion game is established and the target constraint conditions are given. Second, according to the characteristics of the tasks of evading the pursuer and reaching the target constraint area, an SAC neural network is designed, and the intelligent maneuvering strategy of the evader is obtained through reinforcement learning. Finally, simulations are carried out and the simulation results show that the intelligent maneuvering strategy designed based on the SAC method enables the evader to effectively evade the pursuer while satisfying the target constraint conditions, proving the effectiveness of the intelligent maneuvering strategy.