Improved Method for Inverse Shape Optimization Using Constrained Condition Gradients and Genetic Algorithms

Cingoski, Vlatko and Hayakawa, Masahiro and Yamashita, Hideo (2000) Improved Method for Inverse Shape Optimization Using Constrained Condition Gradients and Genetic Algorithms. In: The Proceedings of the 9th International Symposium on Non-Linear Electromagnetic Systems.

[img]
Preview
Text
improved.pdf

Download (1136Kb) | Preview

Abstract

For optimization of various electromagnetic devices an optimization method which efficiently optimizes working parameters, design and shape of the device is necessary. Recently, for this purpose various deterministic and stochastic searching methods have been used, among which the algorithm searching procedure appears to be very promising, especially for multivariable optimization. However, the genetic algorithms have also several demerits, among them their long computation time is usually considered as being the most important. In this paper, an improved optimization procedure based on the genetic algorithm (GA) searching procedure and gradient of the constrained condition is proposed. With this method an optimal solution of a problem is searched by using the gradient of the user-defined constrained condition after quasi-optimal solution is initially obtained by means of the genetic algorithm searching procedure. The usefulness of the proposed method is verified by its application for optimization of a power transformer's tank shield model for which very good results are obtained.

Item Type: Conference or Workshop Item (Paper)
Subjects: Engineering and Technology > Electrical engineering, electronic engineering, information engineering
Divisions: Faculty of Electrical Engineering
Depositing User: Vlatko Cingoski
Date Deposited: 19 Dec 2012 15:12
Last Modified: 01 Oct 2013 11:59
URI: http://eprints.ugd.edu.mk/id/eprint/4052

Actions (login required)

View Item View Item