Saliency Enhancement Design of PMaSynRM through Kriging Surrogate Modeling Optimization

PMaSynRM is emerging as a promising alternative to PMSMs used in electric vehicle propulsion systems. PMaSynRM inherits high reluctance torque with minimal magnet contribution. The traditional method of optimizing the performance of electric motors is complex, as it involves many parameters and is also time-consuming. In the context of improving the performance of PMaSynRM with minimal computational cost, this paper proposes a Kriging surrogate modeling approach integrated with the Non-Dominated Sorting Genetic Algorithm-II. In this approach, initially, the rotor parameters that influence the performance characteristics of PMaSynRM, such as torque, torque ripple and saliency, are identified. These significant rotor parameters are sampled within their bounds using the Halton sequence method, and with these samples the Design of Experiments (DOE) is conducted using the full factorial method. The responses obtained from DOE through FEA simulations are considered as true responses. A surrogate model is also created using these samples to predict the response and its accuracy is assessed by comparing it with the true response. To showcase the effectiveness of the optimization approach, this paper considers a 1.5 kW 36-slot 6-pole PMaSynRM with a fluid flux barrier rotor. The proposed approach results in a saliency of 6.64 and torque ripple of about 7.03 % respectively. Saliency is improved nearly by 2.5 times compared to a recent work having the same torque ripple.