An adaptive multipopulation genetic algorithm for the optimization of active magnetic bearings

Abstract

Abstract Active magnetic bearings (AMBs)are mechatronic systems to support the rotors without any physical contact. Compact AMBs are required for better performance. Present work aims at optimization of AMB for achieving the compact size. Genetic algorithms (GAs) are mostly used for the optimization of such nonlinear design problems. However, GAs with fixed parameters may result in convergence to local optimal solutions and sluggish convergence speed.To remove these deficiencies, the adaptive multipopulation genetic algorithm (AMPGA) is used for the optimization of AMB in the present work. In this algorithm, crossover and mutation probabilities and migration rateare regulated based onpopulation diversity. In addition, real-encoded chromosomes are used. Performance of the designed AMPGA is studied in terms of the convergence speed and optimal solution. It is found that AMPGA converges at a faster rate and gives a better optimal solution than both multipopulation genetic algorithm (MPGA) and simple genetic algorithm (SGA). The optimal value of the AMB size obtained by AMPGA is found to be smaller than that obtained by MPGA and SGA.