249: Improving Toyota Prius Traction Motor Performance Using Evolutionary Algorithm Optimization

Voiceover

Presenter(s)

Marcus Wolff

Abstract or Description

The optimization of electric machine design is paramount within both commercial and personal contexts. Bettering the performance of vehicle motors is no exception, as consumers demand increased performance in their cars with each year that passes. This research aims to optimize torque performance while reducing torque pulsations in a 2010 Toyota Prius motor by altering ten different motor variables, while remaining subjected to nine design constraints to ensure a feasible geometry. Machine design is being optimized using the multi-objective evolutionary algorithm NSGA-II. Currently, the research involves running an algorithm that uses Pymoo framework to generate random values for design solutions, communicate these solutions to a MATLAB script and update the machine model in Altair Flux-2D using these values, solve the updated model to get objective function values, use NSGA-II within Pymoo to optimize the objective functions, and generate the next generation of design solutions to start the cycle over again. While the focus of the research is optimizing the Toyota Prius drive motor, further research questions can be considered in ways to reduce computational time as well.