Parametric Optimization of Electric Machines

Siegfried Silber, Team leader at the Department of Electric Drives Linz Center of Mechatronics GmbH (LCM)

The design of electrical machines requires the integration of several interdependent physical domains, including electromagnetic, thermal, structural, and acoustic analyses. Optimizing performance in one domain often adversely affects another; for example, achieving superior electromagnetic properties may negatively affect the mechanical strength of the rotor. This paper presents a comprehensive approach to parametric optimization that simultaneously considers these conflicting requirements. The use of genetic optimization algorithms is explored to effectively navigate the complex multidimensional design space. In addition, accelerated optimization techniques using surrogate models are introduced to reduce the computational burden while maintaining accuracy. The proposed methodology aims to improve the overall performance and reliability of electrical machines by balancing the trade-offs between competing physical constraints.