The efficient optimization of geometry and parameters of electric machines is crucial in the design process, particularly when dealing with uncertainties arising from manufacturing or materials. In this talk we present a method for designing a 3-phase, 6-pole Permanent Magnet Synchronous Motor (PMSM) that takes these uncertainties into account. Our optimization method uses the open source package GeoPde, which uses Isogeometric Analysis (IGA) and is well suited to flexibly optimize the motor's shape. Our research has two main goals. Firstly, we aim to improve PMSM designs by combining free form and parameter optimization. Secondly, our robust optimization approach addresses uncertainties by formulating a min-max problem, ensuring consistent motor performance. We solve this bilevel problem by working with the value function of the lower level maximization problems and apply then a generalized version of Danskins Theorem to calculate generalized derivatives. The adjoint method enables us to solve these problems efficiently using gradient-based optimization. A nonsmooth optimization method is used to solve the outer robust problem. To conclude, we present numerical results for the proposed approach.This is joint work with my supervisor Stefan Ulbrich, Michael Wiesheu and Sebastian Schöps

Bio: Theodor Komann is a Ph.D. student in Mathematics at the Technical University of Darmstadt, currently in his second year in Stefan Ulbrich's nonlinear optimization group. His research focuses on robust design optimization of electric machines using Isogeometric Analysis for PDE and domain discretization. He holds a Master's degree in Mathematics from the same university.

See all upcoming talks at https://www.anl.gov/mcs/lans-seminars