Date of Award

Summer 1994

Document Type

Thesis - Restricted

Degree Name

Master of Science (MS)

Department

Electrical and Computer Engineering

First Advisor

Arkadan, Abdul-Rahman A.

Second Advisor

Belfore, Lee A.

Third Advisor

Josse, Fabien J.

Abstract

The inverse problems in electromagnetic system design, optimization, and identification received lately much more attention and interest. The inverse electromagnetic field problem can be stated as the problem of predicting the values of the descriptive parameters of a device or system, which produces the required or given performance. The prediction, in general, can be tackled as a system mapping problem. Artificial Neural Network is a powerful tool in pattern recognition and system mapping, and Finite Element Analysis is widely used for solving electromagnetic field problems. In this thesis, The Artificial Neural Networks is investigated and applied in conjunction with Finite Element Analysis to the solution of the inverse electromagnetic field problem. The purpose of this approach is to exploit the recognition and mapping capabilities of neural networks and the analytical strength of Finite Element method.

Share

COinS

Restricted Access Item

Having trouble?