Unit commitment strategy of thermal generators by using advanced fuzzy controlled binary particle swarm optimization algorithm

This paper presents a fuzzy controlled and multi-population based binary clustered particle swarm optimization (BCPSO) algorithm to solve short term thermal generation scheduling problem. In order to incorporate the uncertainties regarding forecasted load demand, the spinning reserve requirement and the total production cost, the formulations are modified by employing fuzzy logics. Each of these uncertain entities are associated with fuzzy membership functions which determine the degree of acceptance. The aggregated membership function, which combines the individual membership functions of fuzzified variables, is incorporated with the fitness value to provide the acceptability measurement of a particular candidate schedule. Typically, generation scheduling is a highly non-linear, multi-peak combinatorial optimization problem. In this method, the potential candidate schedules (or individuals) are distributed among several clusters based on their acceptance values. Each individual of a particular cluster then flies through to its cluster-space towards the cluster best while improving its personal best position. Gradually, as the population grows, the cluster space is also increased to ensure the global convergence. Therefore, this algorithm explores a larger search space and thus reduces the probability of local trapping. A dynamic probabilistic mutation operator is applied on the individual solutions based on their associated fitness values. Simulation result is provided to show the effectiveness of BCPSO while considering two different power system configurations. (C) 2012 Elsevier Ltd. All rights reserved.