Multi-objective optimization design of permanent magnet drive based on Kriging model

Permanent magnet drive (PMD) design problems are studied and a multi-objective optimization design approach based on Kriging approximate model is put forward, which takes a variety of structural parameters and performance parameters (such as thickness of permanent magnets, air gap, etc.) as design variables and the minimum eddy current loss and maximum out torque as double optimization goals. Firstly initial sampling data is obtained with the orthogonal experiment design and finite element method (FEM); then the approximate model between the structrual parameters and the eddy current loss and out torque are set up with Kriging method; finally the structure parameters of PMD is optimized by using multi-objective particle swarm algorithm(MOPSO). In the finite element simulation experiment, the magnetic braking force and eddy current density distributions before and after the improvement are compared, the results prove that the parameter optimization method is feasible and has realized the optimization of PMD structrual parameters configuration, and improved working efficiency of the system.