A Coordinated Optimization Method for New Power System Source Network Load Storage Based on Deep Learning

In response to the problem that most existing methods are difficult to apply to the regulation of new power systems, A coordinated optimization method for new power system source network load storage based on deep learning is proposed. Firstly, construct an optimized architecture for source network load storage in a new power system, achieving unified regulation of distributed power sources, controllable loads, and various types of energy storage. Then, taking into account the characteristics of the source network load storage system, a target model was constructed to minimize operating costs and waste wind and light. Finally, the Proximal Policy optimization algorithm is used to optimize the deep reinforcement learning network and apply it to the solution of the objective optimization model to obtain the optimal coordinated control strategy for the system. Based on the IEEE 33 node system, experimental verification was conducted on the proposed method, and the results showed that the optimized system had only 234700 yuan of operating cost.