A Data-Driven Approach for Design Knowledge Extraction of Synchronous Reluctance Machines Using Multi-Physical Analysis.

Using a data-driven approach, this paper provides electromagnetic, structural, acoustic, and thermal guidelines for designing synchronous reluctance machines. Design spaces of different rotor geometries are created and simulated using finite element packages to evaluate the self- and mutual-correlation of different physical performances, such as average torque, torque ripple, efficiency, power factor, saliency ratio, sound pressure level, mechanical stress, and average winding temperature. Then, a statistical analysis is conducted to extract knowledge and guidelines for relating the design and objective spaces. It is demonstrated that not all the objectives must be incorporated into the design process since some of them are non-conflicting. Hence, a motor designer can numerically evaluate which design variables should be changed and by how much in order to fulfill the design specifications. Multiple designs are selected based on different multi-physical requirements. Useful guidelines for selecting the appropriate motor speed and voltage ratings are proposed while considering structurally reliable designs. [ABSTRACT FROM AUTHOR]