| |
Neural Network and Genetic Algorithm as a New Approach in Design Optimization of Induction Machines
| [View]
[Download]
|
| Author(s) |
A. Tokic; I. Kapetanovic; Z. Haznadar; V. Madzarevic; S. Berberovic; J. Smajic |
| Abstract |
A new approach for optimization of induction machines is presented in this paper. This algorithm is based on the neural networks and genetic algorithm. Neural networks are used as a tool for prediction of induction machine characteristics with respect to induction machine geometrical variables. The optimization process is based on genetic algorithm, as a method from the category of probabilistic algorithms. Optimization algorithm is tested on the optimization of double-cage induction machine starting characteristics. Numerical calculations of induction machine starting characteristics, as an important part of this algorithm, are performed using finite element method for electromagnetic field computation inside of induction machine. Results of electromagnetic field computation are used for synthesis of induction machine equivalent circuit. Full algorithm for calculation of induction machine starting characteristics is developed and named as “voltage balanced iterative process”. Results of starting characteristics calculation are compared with results of experimental measurement. Result of optimization process is induction motor with better starting performances. |
| Download |
| Filename: | EPE-PEMC2002 - T8-027 - Tokic.pdf |
| Filesize: | 464.8 KB |
|
| Type |
Members Only |
| Date |
Last modified 2004-05-25 by System |
| |
 |