-
Renata Rezende Reis
Universidade Federal de Mato Grosso do Sul
-
Marcio L. M. Kimpara
Universidade Federal de Mato Grosso do Sul
-
Raymundo C. Garcia
Universidade Federal de Mato Grosso do Sul
-
Gabriel Gentil
Universidade Federal do Rio de Janeiro, RJ
-
Thyago Estrabis
Universidade Federal do Rio de Janeiro, RJ
-
João Onofre Pereira Pinto
Oak Ridge National Laboratory, Oak Ridge, TN
Keywords:
Switched Reluctance Motor, Genetic Algorithm, Finite Elements, Optimization, Torque Ripple
Abstract
The use of a Genetic Algorithm (GA) as a tool for the solution to engineering optimization problems has been proven to be effective and easy to implement. This paper presents a Genetic Algorithm- based optimization of the firing angles of a 2.2 kW 8/6 Switched Reluctance Machine aiming to reduce a well-known disadvantage of the SRMs: the torque ripple. Initially, the machine was modeled in Matlab/Simulink® by means of lookup tables obtained via finite element simulation. Thus, based on a co- simulation model also implemented in Matlab/Simulink®, the algorithm returned the optimized commutation angles of the phase excitation that minimize the SRM torque ripple for a chosen operation point. The results confirm that the torque performance of the SRM is sensitive to the commutation angles and, therefore, it was possible to define switching commands that improved the torque performance of the SRM drive up to 53% when comparing non-optimized and optimized angles for four different operating points.
