Optimal Scheduling of Wind-Photovoltaic-Energy Storage System Based on Deep Deterministic Policy Gradient Algorithm

As the ratio of renewable energy sources increases in the grid, the randomness and intermittency caused by weather and other factors pose a great challenge to the optimal dispatch of the grid. In this paper, a wind-photovoltaic-energy storage system (WPESS) optimal scheduling model based on deep deterministic policy gradient (DDPG) algorithm is proposed. And the optimal policy is obtained by continuous learning through the interaction between the agent and the system. Firstly, the optimal scheduling model of WPESS is introduced. Secondly, the optimal scheduling process is transformed into a Markov decision process. And a real-time reward function is set to replace the objective function and equipment constraints. Finally, the validity of the model and algorithm proposed is verified based on the analysis of a case, and the effects of different parameter settings on the convergence of the algorithm are compared.