Research on Coordinated Optimization of Transmission and Distribution Networks Based on Deep Reinforcement Learning

Coordination and optimization of the restoration process of the transmission and distribution grids during the frequent occurrence of extreme events leading to major power outages is essential to increase the speed and reduce the cost of restoration. However, transmission and distribution grids belong to different operators with limited interactive information. To achieve coordinated operation between the two during the load recovery stage, this paper proposes a coordinated recovery optimization method for transmission and distribution networks. The transmission and distribution network restoration operation model are constructed with the minimum restoration cost as the objective function, and the global transmission and distribution network restoration problem is decomposed into two types of sub-problems, namely, transmission network restoration and distribution network restoration, for coordinated optimization by using the analytical method of cascade of objectives (ATC) and reinforcement learning method (DDPG). The effectiveness, convergence, and computational efficiency of the proposed method are verified by taking the transmission and distribution system composed of IEEE30-node transmission grid and IEEE33-node distribution grid as an example.