High-Performance Computing Based Fully Parallel Security-Constrained Unit Commitment With Dispatchable Transmission Network

The co-optimization of security-constrained unit commitment and dispatchable transmission network is a large-scale and computationally complex optimization problem that brings big challenges for the traditional centralized and/or master-slave based decomposition approaches. This paper proposes a fully parallel approach to achieve an efficient and fast solution while considering post-contingency corrective actions. In this paper, the original co-optimization problem is decomposed into six major solution modules for unit commitment, optimal power flow, and transmission switching in base case and contingencies, respectively. Moreover, each major solution module is further decomposed by unit, period, dispatchable line, and/or contingency to make the proposed approach favorable in a fully parallel computing environment. The effectiveness and efficiency of the proposed fully parallel approach are justified using numerical cases on a high-performance computing (HPC) platform.