Search space reduction strategies for unit commitment problem

摘要

In electricity markets, independent system operators solve the unit commitment (UC) problem to determine the on/off status of generating units with respect to the day-ahead load forecast. However, as the size of power systems grows, computational complexity of the UC problem drastically increases. In this paper, based on the greedy adaptive search procedure (GRASP) technique, we propose a novel heuristic algorithm to reduce the number of variables and constraints in the day-ahead unit commitment problem. Once the size of the problem is reduced, we employ a state-of-the-art mixed integer programing method to solve the problem. The numerical results show the effectiveness of the proposed algorithm.