Numerically Efficient Calculation of Flux Maps and Efficiency Maps of Electric Motors

Abstract

This thesis presents efficient interpolation techniques to calculate flux linkage, torque, and loss maps in synchronous reluctance machines (SyRMs) and permanent magnetassisted synchronous reluctance machines (PMa-SyRMs). Advanced methods are required due to the computational expense of extensive finite element method (FEM) simulations. Two interpolation methods are implemented. The first is recursive piecewise polynomial interpolation, which accurately handles sharp magnetic changes near the edges by dividing the area into smaller sections as needed. The second method is adaptive sparse grid interpolation, which effectively manages interpolation in multiple dimensions by refining the sampling grid only in regions where the system behaviour changes significantly. These methods are integrated into a MATLAB-FEMM simulation framework and validated against the SyR-e software. Results indicate that both techniques provide accurate performance maps while the methods proposed in this thesis are significantly reducing computational requirements. This framework supports rapid characterization, control algorithm development, and optimization of electrical machines.

Date of Award	2025
Original language	English
Supervisor	Rastko Zivanovic (Supervisor)