Robust design of a direct-drive PM hub motor by fuzzy-inference taguchi method

Abstract

In this paper, using the fuzzy-inference Taguchi method with multiple performance characteristics index (MPCI), a novel systematic approach is employed to obtain the robust design of a brushless permanent-magnet (PM) hub motor with both lower torque ripple factor and higher efficiency. According to the geometric parameters of the PM motor, we select the controllable factors and their levels, such as the permanent magnet thickness in rotor, air-gap width, opening slot width and slot tooth width at stator. The proposed method first establishes the orthogonal array (OA) that is required by the Taguchi method. Then, the multiple targets are coordinated by the fuzzy inference mechanism to obtain the better combination of the geometric parameters for achieving multiple quality targets. Finite element method (FEM) is employed as the tool for analyzing the torque ripple and performance of the PM motor, and results have indicated that the proposed method can obtain the suitable motor's geometric parameters for reducing the cogging torque and enhancing the operating efficiency.