Multi-Objective Optimization of a Line-Start Synchronous Machine Using a Self-Organizing Algorithm

This article introduces the design and optimization of line-start synchronous machines in fractional horsepower industry applications. The proposed optimized design will be considered as a replacement for the induction machine driven from an ac drive. The optimized design will be achieved using a 2-D finite-element transient analysis and multi-objective optimization algorithm. For the optimization, a multi-objective self-organizing migrating algorithm was chosen. The optimization results are compared with the more popular nondominated sorting genetic algorithm (NSGA-II) algorithm, to prove algorithm sufficiency with three objectives and five design parameters in the area of the electromagnetic design application.

I. Introduction

Globally, the goal of industrial applications is always to deliver the most reliable components while achieving the best possible characteristics. Electrical machines are being pushed toward their highest possible efficiency for the lowest cost; therefore, tradeoffs must be introduced. A long time has been spent researching and optimizing induction machines (IMs), and in some cases, IM has reached its limits [1]. Thus, the search for highly efficient systems means that older machines must be suitably replaced. One possibility is the line-start synchronous machine (LSSM) because of its superior efficiency and ability to keep direct grid connection capability. LSSMs use a squirrel cage for the startup operation and then synchronize with the stator rotating field to develop synchronous electromagnetic torque at the nominal speed.