Self-adaptive genetic algorithms based characterization of structured model parameters

This study presents a self-adaptive genetic algorithm based parameters characterization of a structured model of an induction machine. By its evolutionary character, the standard genetic algorithm is considered as a product of artificial intelligence techniques. It treats efficiently complex problems despite its relative slowness in seeking a consistent global solution. In order to reduce computing time and hence overcome this shortcoming, a new self-adaptive scheme has been introduced. It is incorporated within the genetic algorithm characterization procedure. This self-adaptive GA procedure is used to best-fit curves of an induction machine by optimally characterizing its model parameters, which build up the machine Park model employed in drive systems. This approach is performed to demonstrate its utility and effectiveness. Finally, the resulted machine performances of the proposed characterization process are compared with the experimental data reference curves.