Unit commitment computation by fuzzy adaptive particle swarm optimisation

A fuzzy adaptive particle swarm optimisation (FAPSO) for unit commitment (UC) problem has been proposed. FAPSO reliably and accurately tracks a continuously changing solution. By analyzing the social model of standard PSO for the UC problem of variable resource size and changing load demand, the fuzzy adaptive criterion is applied for the PSO inertia weight based on the diversity of fitness. In this method, the inertia weight is dynamically adjusted using fuzzy IF/THEN rules to increase the balance between global and local searching abilities. Velocity is digitised (0/1) by a logistic function for the binary UC schedule. To improve knowledge, the global best location is also moved instead of a fixed one in each generation. To avoid the system to be frozen, stagnated/idle particles are reset from time to time. Finally, benchmark data and methods are used to show effectiveness of the proposed method