Analytical Modeling and Optimization of PM Synchronous Machines Using Novel R262-TLBO Algorithm

Abstract

This article presents analytical modeling and optimization for permanent magnet (PM) synchronous machines using a novel algorithm named realistic 262-teaching-learning-based optimization (R262-TLBO). Through enhancing the convergence speed and searching capability, this study aims to uncover the most favorable compromise solutions that lie within the Pareto front, which can be utilized across a range of combinations of pole and slot numbers. To accomplish this objective, a modified complex permeance (CP) model is developed to calculate the electromagnetic performance that can be applied to any winding configurations. Subsequently, the proposed R262-TLBO algorithm is employed with the CP model, and the outcomes are compared with those obtained from the widely recognized algorithms. The test results demonstrate that the amalgamation of the R262-TLBO algorithm and the modified CP model outperforms other combinations featuring pre-existing algorithms, thereby confirming its immense potential to be utilized in PM design problems. Finally, the proposed analytical results are verified by a finite element analysis (FEA).