Parallel Co-Optimization of Unit Commitment and Transmission Switching

Transmission switching is an efficient way to improve the network controllability and increase the economic benefits of power systems. The co-optimization of the generating unit commitment and transmission switching is a large-scale and computationally complex optimization problem, which is hard to solve by traditional centralized and/or master-slave based decomposition approaches. This paper presents a parallel approach to obtain an efficient and fast solution for this co-optimization problem considering post-contingency corrective actions. Augmented Lagrangian method and auxiliary problem principle (APP) are adopted to decompose the original problem into three solution modules: the unit commitment (UC) module, the optimal power flow (OPF) module and the transmission switching (TS) module. These three modules can be solved simultaneously, which makes the proposed method favorable for parallelization and consequently higher computational efficiency. Numerical cases are tested on a distributed computing cluster with 16 computers to justify efficiency of the proposed approach.