Convergence acceleration techniques for the stochastic unit commitment problem

This paper proposes a two-stage stochastic mix-integer programming formulation that addresses a unit commitment (UC) problem in a probabilistic manner. The objective of this problem is to minimize the expected operating cost for the decision time horizon under load uncertainty. The problem is solved using a standard L-shaped algorithm. We then propose some acceleration techniques to improve convergence behavior. These techniques are tested on a 9-bus system to determine computing efficiency. Results show that the proposed acceleration techniques reduce number of iterations as well as the algorithm convergence time.