In a proton exchange membrane fuel cell (PEMFC) system, the flow of the air and hydrogen is the main factor affecting the output characteristics of the PEMFC, and there is a coordination problem in the flow control of both. To ensure the real-time gas supply in the fuel cell and improve the output power and economic benefits of the system, a deep reinforcement learning (DRL) controller based on fusion optimization with deterministic policy gradient and a control optimization strategy based on net power optimization are proposed in this paper. The experimental results show that the control algorithm proposed in this paper can effectively improve the dynamic performance and steady-state performance of the system, which is embodied in the average 12.5% increase in the dynamic performance compared with the fuzzy PID control and average 99.54% increase in the steady-state performance compared with the traditional DRL control.