Genetic algorithm approach to more consistent and cost effective unit commitment

Genetic algorithms (GAs) have been demonstrated to be an effective technique to solve the unit commitment (UC) problem due to its capability in dealing with mixed-integer problems and ease in handling constraints. However, it is clearly indicated that GAs exhibit two major drawbacks in solving the problem which excludes them from real-time applications: one is their slow processing speed, and the other is their solution inconsistency from different GA runs. This paper proposes a mechanism to solve the UC problem with GAs so that the solution variation is reduced from run to run and the solution speed can be improved. The effectiveness of the proposed method is illustrated on a 6-generator system