A constraint logic programming algorithm for solving thermal unit commitment problem

This paper presents a constraint logic programming (CLP) algorithm to solve the thermal unit commitment (UC) problem. The CLP algorithm provides an efficient and flexible solution method for the UC problem. Constraints satisfaction techniques are used to prune the search space a priori. The solutions that can not appear as a feasible one are removed from further evaluation. These techniques propagate the constraints as much as possible, and assume the values of some variables of the unit on/off schedule until the infeasible solutions are excluded. After pruning the infeasible solutions, branch and bound approach is then employed to find the minimum operation cost from the feasible domains. The algorithm has been implemented to solve the practical Taiwan power 38-unit thermal system UC problem. Results are compared to those of the Lagrangian relaxation (LR) method to demonstrate the feasibility and effectiveness of the proposed approach to solving the thermal UC problem.