Electromagnetism-like mechanism particle swarm optimization and application in thermal process model identification

In order to improve the performance of the particle swarm optimization (PSO) algorithm, the electromagnetism-like mechanism (EM) algorithm is introduced and the electromagnetism-like mechanism particle swarm optimization (EMPSO) algorithm is put forward. Each particle´s best value is replaced by the weighted value of all particles´ best value and local search mechanism is used in the algorithm. The results of testing typical function demonstrate that the convergence precision and convergence rate of EMPSO algorithm are better than those of PSO algorithm. EMPSO algorithm is used to identify the typical transfer function of thermal processes and the transfer function between primary air volume and bed temperature in a circulating fluidized bed (CFB) boiler, achieving some satisfactory identification results.