A Data-aided Security Constraint Prescreening Technique and Application to Real-world System

Security constraints are necessary in unit commitment problems to guarantee reliable commitments. However, only a small subset of security constraints can bind in the unit commitment problem. This paper presents a data-aided prescreening approach that could identify and remove non-dominating security constraints. This approach takes advantage of both deterministic and statistical methods, which leverages the historical data and linear programming (LP) solution. Lazy constraints are used in optimization solver to maintain solution integrity. Case studies presented here use Midcontinent Independent System Operator (MISO) data, and simulation results show that the proposed prescreening approach significantly improves computational performance.