Optimal Dispatch of Integrated Energy System Based on Deep Reinforcement Learning

In allusion to the uncertainty of renewable energy and load in integrated energy system, an optimal dispatch method based on deep reinforcement learning was proposed. Firstly, the methodology of the deep reinforcement learning was expounded, and an optimal dispatch model based on the deep reinforcement learning, in which the state space, action space and reward function were designed, was proposed. Secondly, the model solving process based on asynchronous advantage actor-critic (abbr. A3C) algorithm was designed. Finally, the results of example simulation show that the proposed method can adaptively respond to the uncertainty of source and loads, and its optimization effect is similar to that of traditional mathematical programming method.