Thermal Unit Commitment Using Hybrid Binary Particle Swarm Optimization and Genetic Algorithm

This paper presents a hybrid algorithm which integrates Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) for solving thermal unit commitment. The UC problem consists of two sub-problems: Unit Scheduled problem which is solved by PSO for minimization of transition cost and Economic Dispatch that can be solved by GA by the means of minimizing the production cost. The proposed algorithm is demonstrated for a system including ten thermal units. Running PSO and GA simultaneously justifies the production cost reduction in 24 hour period. Choosing varying PSO acceleration coefficients and inertia weight make the system convergence faster and skip the local optimums.