GA and PSO Based Approaches for Fast Optimization of External Rotor Switched Reluctance Motor Design Parameters

With the application of computational intelligence methods to different fields, it has also started to be used in the designs of electric motors. In this paper, a 3-phase 18/12 pole 30 kW in-wheel External Rotor Switched Reluctance Motor (ERSRM) is designed for the Electric Vehicles (EVs). For this design, an analytical model is developed using Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) methods. While determining the design dimensions of the ERSRM, it is aimed to obtain maximum efficiency and the desired torque fast in the developed models. As a result of the developed models using the GA and PSO methods, motor design parameters obtained are compared with the results of Finite Element Analysis (FEA). In the designed models, the inductance values at the unaligned and aligned positions, torque, and efficiency values are calculated. The results obtained from the heuristic based models developed are subjected to various error calculation methods. Consequently, it is observed that the results obtained by the GA and PSO methods are much faster than that of the FEA and the errors are acceptable. Besides, it is seen that the ERSRM design realized by the proposed methods has low time cost and high accuracy.