A Parallel Decentralized Solution for Multi-Regional Unit Commitment with Convex AC Power Flow Constraints

Abstract

Scheduling a multi-regional unit commitment is an efficient way to operate the power system economically. The conventional unit commitment models, which adopts the DC power flow, fail to take the network loss and voltage drop constraints into account. This paper proposes a novel AC power flow model to fill this gap. The proposed model first achieves convexity by decomposing the nonlinear part of the network loss into the voltage amplitude part and the phase angle part, and then is formulated into two sets of linear inequalities. After that, this paper integrates it into a multi-regional unit commitment. Considering the independence and parallelism of operation between regions, the established unit commitment is solved via a parallel decentralized solution, where the simplicial decomposition method (SDM), the nonlinear block Gauss-Seidel (GS) method, and the augmented Lagrangian method (ALM) (abbreviated as SDM-GS-ALM) are integrated. Finally, the efficiency and applicability of the multi-regional unit commitment with convex AC power flow constraints are verified on IEEE Test Systems.