| PhD Seminar

Name of the Speaker: Mr. Anupuruvadraju Naga Sampath Kumar (EE20D025)
Guide: Prof. Srirama Srinivas
Venue: ESB-244 (Seminar Hall)
Date/Time: 29th January 2025 (Wednesday), 4:00 PM
Title: Design Optimization of PMSynRM for Electric Vehicles Applications Using NSGA-II and Surrogate Modeling

Abstract :

Electric propulsion systems are rapidly evolving to meet the sustainable and high-performance transportation demands. Synchronous Reluctance Motors (SynRMs) and the Permanent Magnet Assisted Synchronous Reluctance Motors (PMSynRMs) are emerging as promising alternatives to the Permanent Magnet Synchronous Motors due to their reliance mainly on reluctance torque and reduced dependency on rare-earth magnets. It is observed that the existing designs of SynRM and PMaSynRMs end up with compromised performance characteristics, lower power density, and increased torque ripple, etc.

This talk presents a novel design approach for ferrite-magnet-based PMSynRMs that utilizes fluid flux barrier geometry in an outer rotor configuration to achieve enhanced saliency and motor performance. The methodology introduces a phase advance angle to optimize torque and employs finite element analysis (FEA) in Ansys Maxwell for high-fidelity simulations. Magneto-static parametric analysis is conducted to determine the optimal phase advance angle, while transient simulations iteratively refine the torque profiles by optimizing carrier and barrier widths using the Particle Swarm Optimization (PSO) algorithm. Further, a linear skew rotor configuration is also introduced to further mitigate cogging torque. The proposed design approach demonstrates significant advancements, achieving an improvement in power density and reduction in torque ripple compared to some existing works. Additionally, the motor withstands up to 3 p.u. of current without irreversible demagnetization, ensuring robustness and reliability which is an important characteristic desirable in EV motors. To further optimize the design process, Kriging surrogate modeling is integrated with the Non-Dominated Sorting Genetic Algorithm-II (NSGA-II) for the inner rotor motor. This proposal mainly reduces the computational complexity while maintaining precision in exploring rotor design parameters. The envisaged optimized design approach achieves a high saliency ratio and also a torque ripple reduction of about 6.63%, underscoring the potential of advanced optimization techniques in electric motor development. The proposed design optimization approaches paves way for the next generation of energy-efficient and robust electric motors for EV applications.