The multiple objective optimization of high-speed rotor supported by magnetic bearing in BLDCM

In order to find an algorithm to get a better optimization result of high-speed rotor supported by magnetic bearing in BLDCM, we presented a multiple objective optimization results which included three algorithms' in this paper. They are the local Sequential Quadratic Program (SQP) algorithm, the global Genetic Algorithm (GA) and the combined optimization strategy algorithm which combines GA and SQP. The parametric optimization model of a 100 kW BLDCM supported by magnetic bearings was constituted with software ANSYS and an effective connection between software ANSYS and iSIGHT was used to execute the whole optimization process. To insure the best performance, mass and strength were chosen as the optimization goals, meanwhile, the static strength, dynamic modal, shape and magnetic force of the rotor subassembly were used as the main constrains. Six main dimensions of the subassembly were optimized. The optimization results indicated that the GA can get a higher optimization precision than the other two algorithms and the SQP was not effective in the optimization of magnetic suspended motor rotor subassembly. The GA's optimization result made the mass decrease 7.62 percent with the safe factor is 3.15. The 100 kW BLDCM supported by magnetic bearings was designed and fabricated; the multiple objective optimization results were verified by the prototype.