Search

This paper presents the application of two classes of Evolutionary Algorithms (EA) to determine optimum design of Single-Phase Switched Reluctance Machine (SPSRM). The EA used is Genetic Algorithms (GA) and Differencial Evolution (DE). Due to sensitivity of the output torque to the stator and rotor pole arcs, these are selected as design variables for a multi-objective optimization with the objective of maximizing average torque and torque density, and minimizing copper loss. The proposed optimization is tested on a 4/4 1,25 kW SPSRM, and the results of both algorithms are compared. The performance of the optimized motor are compared to the initial motor through the finite element analysis. The results show improvement in both efficiency and output torque. [ABSTRACT FROM AUTHOR]

The article presents a review of the design procedures for a switched reluctance machine. A finite element simulation is used to obtain the operating data of the designed prototype. These data are flux linkage, inductance, torque and magnetic saturation. With them it is possible to profile and inductance torque and verification of design for the prototype. It also presents a criterion for analysis operational limit for the machine, based on data collected in the simulation. [ABSTRACT FROM PUBLISHER]