A Hybrid FEA and MOGA-Based Optimization Framework for Industrial Switched Reluctance Motors

Abstract

Switched Reluctance Motors (SRMs) are increasingly used in industrial drives and electric vehicles due to their robust construction, fault tolerance, and wide speed range. However, the non-linear magnetic characteristics of SRMs often lead to torque ripple, vibration, and acoustic noise, posing challenges to achieving optimal performance. This paper presents a Hybrid Finite Element Analysis (FEA) and Multi-Objective Genetic Algorithm (MOGA)-based optimization framework for the industrial design of SRMs. The proposed method integrates accurate electromagnetic simulation using FEA with the Non-Dominated Sorting Genetic Algorithm II (NSGA-II) to optimize key geometric parameters, stator pole arc ratio (Es), rotor pole arc ratio (Er), and stator yoke height (Ys), under multiple objectives and design constraints. The objectives are to maximize average torque and efficiency while minimizing torque ripple and total power loss. Constraints on torque, current density, and temperature are incorporated to ensure thermal and electromagnetic feasibility. Compared with conventional single-objective and unconstrained multi-objective SRM designs, the proposed framework offers a clearly interpretable and reproducible optimization route. It achieves a 71.5 % increase in average torque and reduces torque ripple to 3.18% while satisfying electromagnetic and thermal constraints. The framework’s novelty lies in integrating a robust FEA-based evaluation loop with constrained multi-objective decision-making, enabling industrial designers to identify balanced, high-performance SRM geometries efficiently. Future work includes experimental prototyping and machine learning integration to further enhance optimization efficiency.