Design of a synchronous reluctance motor using finite element method, parametrized model of the motor and evolutionary optimization

U ovom radu projektiran je sinkroni reluktantni motor s kružnim barijerama u rotoru primjenom dvije metaheurističke metode. Na početku je dan kratak osvrt na teoriju reluktantnih motora te detaljan opis korištenih optimizacijskih metoda. Nakon toga povezan je programski paket MATLAB s programskim paketom MagNet preko ActiveX sučelja kako bi se automatizirao postupak definiranja modela stroja na temelju različitih geometrijskih podataka motora. Pomoću algoritama diferencijalne evolucije i algoritma kolonije mrava koje je implementiran unutar računalnog paketa MIDACO realizirana je optimizacija motora. Optimizacijama je maksimiziran moment stroja po volumenu uz uvjete da motor daje nazivnu snagu pri nazivnoj brzini, ima indukcije u zubu i jarmu unutar zadanih granica, korisnost stroja je veća od IE4 korisnosti i umnožak strujnog obloga i gustoće struje je manji od zadanog. Na kraju su uspoređeni rezultati dobiveni objema optimizacijskim metodama.In this thesis, a Synchronous Reluctance Machine (SynRM) with circular barriers in the rotor was designed using two metaheuristic methods. At the beginning, a brief overview of the theory of reluctance motors and a detailed description of the used optimization methods were given. After that, MATLAB software system was linked with Infolytica MagNet software package using ActiveX software framework to automate the process of defining the design of SynRM rotor and stator geometry. The applied methods were differential evolution and ant colony optimization algorithm implemented within the MIDACO computer software. The primary design goal was to design a motor with highest torque density, i.e. torque/volume ratio under the conditions that the motor delivers rated power at rated speed, maximal flux densities in the stator tooth and yoke are within the given limits, efficiency is greater than value prescribed by the IEC standard (design effort is to reach the IE4 (Super-Premium Efficiency) efficiency level), and current density multiplied with linear current density is less than prescribed for a motor of this type and size. Finally, the results obtained by both optimization methods were compared