A Preprocessing Method for Security-Constrained Unit Commitment with AC Power Flows

This paper presents a preprocessing method to solve the security-constrained unit commitment with AC power flows (SCUC-ACPF), which is a large scale mixed-integer non-convex, nonlinear optimization problem. We introduce clustering algorithms and local search to reduce problem complexity, and obtain local optimal solutions within limited solving time and computational capacity. The clustering algorithms group buses into manageable clusters centered around core buses, creating smaller sub-grids for efficient optimization. The local search algorithms iteratively explore and improve solution quality. Our method ensures balanced clusters that respect physical and operational constraints, significantly enhancing optimization efficiency and contributing to more effective grid management and operation.