TD3-based Adaptive Economic Dispatch Optimization Strategy for Multi-energy Microgrid

Abstract

The multi-energy microgrid (MEMG) improves the overall economy of the system by coupling scheduling among multiple energy sources. However, in the case of renewable energy power generation and load demand fluctuations, traditional methods are difficult to apply to energy dynamic management and control under the changing situation of multi-energy microgrid systems, which poses a huge challenge to the multi-energy coupling optimal operation of MEMG. In this paper, a multienergy allocation model based on deep reinforcement learning (DRL) is established to optimize the multi-energy coupling scheduling, which can automatically adapt to changes in the environment. In order to make the optimal scheduling strategy effectively reduce the cost, a multi-energy scheduling strategy based on the twin delayed deep deterministic policy gradient (TD3) algorithm is proposed. The experimental results display that our proposed strategy can reduce the cost by 21.45% and 14.71% compared with particle swarm optimization algorithm in summer and winter.