Optimization of Induction Machine Design for Electric Vehicle Powertrain

This paper proposes a methodology to include induction machine (IM) in an optimization scheme for EV powertrains using Particle Swarm Optimization (PSO) as the optimization algorithm. The IM model is developed based on finite element analysis (FEA) and all the losses are estimated. The main objective is to generate a large database of base designs in a computationally efficient but accurate way. To perform the optimization, scaling methods are used. Detailed transmission and inverter models are also considered based on previous work to estimate the overall powertrain efficiency. The optimization case study shows how the optimizer evaluates a large number of alternatives and picks the optimal IM design and associated transmission and inverter. The proposed methodology can serve as a basis for future research on powertrain optimization for dual-motor driven vehicles.