Kriging-Assisted Multi-Objective Optimization Framework for Electric Motors Using Predetermined Driving Strategy

In this paper, a multi-objective optimization framework for electric motors and its validation is presented. This framework is suitable for the optimization of design variables of electric motors based on a predetermined driving strategy using MATLAB R2019b and Ansys Maxwell 2019 R3 software. The framework is capable of managing a wide range of objective functions due to its modular structure. The optimization can also be easily parallelized and enhanced with surrogate models to reduce the runtime. The framework is validated by manufacturing and measuring the optimized electric motor. The method's applicability for solving electric motor design problems is demonstrated via the validation process. A test application is also presented, in which the operating points of a predetermined driving strategy provide the input for the optimization. The kriging surrogate model is used in the framework to reduce the runtime. The results of the optimization and the framework's benefits and drawbacks are discussed through the provided examples, in addition to displaying the properly applicable design processes. The optimization framework provides a ready-to-use tool for optimizing electric motors based on the driving strategy for single- or multi-objective purposes. The applicability of the framework is demonstrated by optimizing the electric motor of a world recorder energy-efficient race vehicle. In this application, the optimization framework achieved a 2% improvement in energy consumption and a 9% increase in speed at a rated DC voltage, allowing the motor to operate at desired working points even with low battery voltage.