Optimal Design of Saturated Switched Reluctance Machine for Low Speed Electric Vehicles by Subset Quasi-Orthogonal Algorithm

This paper investigates the optimal design method of switched reluctance machines (SRMs) for mitigating torque ripple and improving average torque under the low speed operation. In order to produce high torque for hill climbing or overload start-up of micro electric vehicles (EVs) and reduce the converter capacity, an SRM with four times torque overload capability and two times speed range with constant power is needed. To improve the SRM performance, first, a new method called multiphase excitation (MPE) method for calculating resultant static torque is adopted, which takes the cross-coupling and magnetic saturation into consideration to reduce the calculation error bought by the traditional method called single-phase excitation (SPE) method; second, optimal methods, including parameters sensitivity analysis, multi-objective optimization function, and winding connection types, have been taken to release the magnetic saturation, to improve average torque, and to depress torque ripple. Compared with the initial scheme, the torque ripple is minimized from 55.33% to 14.84% and the average torque is improved by 8.36%. The experimental results validate the effectiveness of the proposed methods.