Abstract: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 (MIP), sometimes multiple times per day, with only minor changes in input data. In this work, we propose a number of machine learning 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 a 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 4.3 times faster with optimality guarantees and 10.2 times faster without optimality guarantees, with no observed reduction in solution quality. Out-of-distribution experiments provide evidence that the method is somewhat robust against data-set shift. Summary of Contribution. The paper describes a novel computational method, based on a combination of mixed-integer linear programming (MILP) and machine learning (ML), to solve a challenging and fundamental optimization problem in the energy sector. The method advances the state-of-the-art, not only for this particular problem, but also, more generally, in solving discrete optimization problems via ML. We expect that the techniques presented can be readily used by practitioners in the energy sector and adapted, by researchers in other fields, to other challenging operations research problems that are solved routinely.

Deep Reinforcement Learning-Assisted Convex Programming for AC Unit Commitment and Its Variants

A Solution To Unit Commitment Problem By Aco And Pso Hybrid Algorithm

Unit Commitment Based on Improved Discrete Particle Swarm Optimization

A parallel algorithm for unit commitment including AC power flow constraints

Research on Unit Commitment Optimization Method Based on Deep Reinforcement Learning

Learning to Solve Large-Scale Security-Constrained Unit Commitment Problems

Synergistic Integration of Machine Learning and Mathematical Optimization for Unit Commitment

Managing power balance and reserve feasibility in the AC unit commitment problem

A Parallel Decentralized Solution for Multi-Regional Unit Commitment with Convex AC Power Flow Constraints

A fast linearized AC power flow-constrained robust unit commitment approach with customized redundant constraint identification method

Improving the Computational Efficiency of the Unit Commitment Problem in Hydrothermal Systems by Using Multi-Agent Deep Reinforcement Learning

Feasibility-guaranteed machine learning unit commitment: Fuzzy Optimization approaches

Decentralized Unit Commitment in Integrated Heat and Electricity Systems Using SDM-GS-ALM

Large-Scale Maintenance and Unit Commitment: A Decentralized Subgradient Approach

Global Solution Strategies for the Network-Constrained Unit Commitment Problem with AC Transmission Constraints

Quantum Reinforcement Learning-Based Two-Stage Unit Commitment Framework for Enhanced Power Systems Robustness

Safe deep reinforcement learning for real-time AC optimal power flow: A near-optimal solution

A Fast Solution Method for Large-scale Unit Commitment Based on Lagrangian Relaxation and Dynamic Programming

Modeling and Solution of the Large-Scale Security-Constrained Unit Commitment

Online Operational Decision-making for Integrated Electric-Gas Systems with Safe Reinforcement Learning

Reinforcement Learning and Mixed-Integer Programming for Power Plant Scheduling in Low Carbon Systems: Comparison and Hybridisation