Learning to Solve Large-Scale Security-Constrained Unit Commitment Problems
This work addresses the efficiency problem for power system operators and electricity markets by providing incremental improvements to existing MIP solver performance through data-driven techniques.
The paper tackled the computational challenge of repeatedly solving Security-Constrained Unit Commitment (SCUC) problems in power systems by using machine learning to predict redundant constraints, initial solutions, and solution subspaces from past instances, resulting in average speedups of 4.3x with optimality guarantees and 10.2x without, while maintaining solution quality.
Security-Constrained Unit Commitment (SCUC) is a fundamental problem in power systems and electricity markets. In practical settings, SCUC is repeatedly solved via Mixed-Integer Linear Programming, sometimes multiple times per day, with only minor changes in input data. In this work, we propose a number of machine learning (ML) techniques to effectively extract information from previously solved instances in order to significantly improve the computational performance of MIP solvers when solving similar instances in the future. Based on statistical data, we predict redundant constraints in the formulation, good initial feasible solutions and affine subspaces where the optimal solution is likely to lie, leading to significant reduction in problem size. Computational results on a diverse set of realistic and large-scale instances show that, using the proposed techniques, SCUC can be solved on average 4.3x faster with optimality guarantees, and 10.2x faster without optimality guarantees, but with no observed reduction in solution quality. Out-of-distribution experiments provides evidence that the method is somewhat robust against dataset shift.