Advanced techniques for speed-up the Security Constrained Unit Commitment problem

Security Constrained Unit Commitment (SCUC) is one of the tools used by system operators for day/week-ahead planning operations. It's considered one of the most difficult optimization problems, due to its number/type of variables and constraints. As, due to the operational requirements, it has to be solved in a significantly short time, SCUC still cannot be considered a “well-solved” problem in the literature. In this thesis, a wide literature survey on the algorithms developed for the SCUC problem is provided and a comparison with some standard KPIs is carried out. A novel heuristic algorithm is proposed to better handle the combinatory complexity and to reduce the problem dimension through the identification and elimination of redundant constraints. The computational efficiency of the proposed algorithm is tested by solving the SCUC on a 179-bus system and a 588-bus system and compared with the algorithm usually employed at present by the power industry. The results showed the effectiveness of the proposed algorithm with a significant speed-up of the problem and without loosing in optimality or creating infeasibilities