Intelligent Controller Based on Distributed Deep Reinforcement Learning for PEMFC Air Supply System

Air supply system is an important subsystem in a PEMFC engine system. Research on the control strategy of air great importance and significance engineering. In this paper intelligent controller based distributed deep reinforcement learning which exerts better over flux proton exchange membrane fuel cell (PEMFC) proposed. addition, collective intelligence exploration multi-delay deterministic policy gradient (CIED-MD3) algorithm presented for controller. This improved developed basis (DDPG) adopted enables full environment. classification experience replay mechanism introduced to improve training efficiency. A number techniques are employed effort address Q-value overestimation problem DDPG, including clipped multiple Q-learning, delayed update smooth regularization target policy. Finally, application CIED-MD3 (with its global search ability optimization speed) demonstrated model-free The simulation results show that proposed greater Compared with other methods, exhibits performance robustness. future research.