Active Constraint Identification Assisted DC Optimal Power Flow

The optimal power flow (OPF) is important for the reliable operation and management of power systems. Due to the uncertainties introduced by the increasing penetration of renewable energy resources (RES), more frequent OPF calculations are compulsorily required, posing significant computational burdens to the timely derivation of optimal dispatching solutions. In this paper, an active constraint identification (ACI) approach is proposed to identify the active constraints under different generation and demand conditions so that the OPF computational time can be reduced. The ACI is based on deep convolutional neural networks. Simulation studies are performed on the IEEE 14/118/300 bus systems, and the optimal power flow is solved by using Gurobi/Python. Simulation results of the proposed methods are compared with those of the state-of-the-art to demonstrate the calculation speed improvement of the proposed method.