Abstract: Aiming at the control problems of water pump and radiator in fuel cell thermal management system, a joint control strategy based on deep deterministic policy gradient（DDPG） was proposed. This strategy replaces the independent controller of the water pump and radiator in the traditional control framework, and uses an intelligent agent with multiple inputs and multiple outputs that can simultaneously control the cooling water flow rate of the water pump and the air flow rate of the radiator. Firstly, the state space and action space of the intelligent agent are determined. Then, the reward function is set by the control goal. Finally, examples are given to verify the effectiveness of the algorithm. The results show that the proposed joint control strategy can effectively control the flow rate of cooling water and air at the same time, thereby improving the operating efficiency of PEMFC.