Sensitivity Analysis and Optimal Design of a Stator Coreless Axial Flux Permanent Magnet Synchronous Generator

Abstract

In this paper, the modified initial design procedure and economic optimization design of a stator coreless axial flux permanent magnet synchronous generator (AFPMSG) are presented to improve the design accuracy, efficiency, and economy. Static magnetic field finite-element analysis (FEA) is applied to the magnetic equivalent circuit (MEC) method to increase the accuracy of electromagnetic parameters and reduce the iteration times. The accuracy and efficiency of the initial design is improved by the combination of MEC method and static magnetic field FEA in the design procedure. For the economic optimization, the permanent magnetic (PM) material volume model, which affects the cost of the AFPMSG the most, is derived, and the influence degree of the main structure parameters, to the performance, is distinguished and sorted by sensitivity analysis. The hybrid genetic algorithm that combines the simulated annealing and father-offspring selection method is studied and adopted to search for the best optimization solution from the different influence degree and nonlinear interaction parameters. A 1 kW AFPMSG is designed and optimized via the proposed design procedure and optimization design. Finally, 3D finite-element models of the generator are simulated and compared to confirm the validity of the proposed improved design and the generator performance.