A Multi-objective Optimization Framework for the Design of a High Power-Density Switched Reluctance Motor

Switched Reluctance Motors (SRMs) are reliable, robust, and magnet-free, but they have relatively lower power density, higher torque ripple and vibration. With the design and control optimization, the drawbacks of an SRM can be addressed. However, the relationship between geometry parameters and optimum control parameters can make the design of an SRM challenging. This paper proposes a multi-objective optimization framework for the design of a high-power density SRM. The proposed framework overcomes the challenges due to the motor geometry and current control interdependencies. It has two main stages. The first stage is the static optimization to find the geometries with high static average torque. The static optimization utilizes diverse non-linear surrogate models and global optimization algorithms to find a comprehensive set of geometries with high average static torque. In the second stage, for each of those geometries, the dynamic optimization loop is employed to improve the average torque, torque ripple, and radial forces. Dynamic optimization is a loop involving gradient-based deterministic optimization and stochastic controls optimization. The variations of geometry parameters in each iteration of the dynamic optimization is constrained to ensure the same control parameters can be maintained. The framework has been employed in the design of a high-power density SRM for an aerospace application. The same framework can be adapted for different SRM applications by modifying the constraints and objectives.