Date of Award

Spring 2004

Document Type

Thesis - Restricted

Degree Name

Master of Science (MS)

Department

Electrical and Computer Engineering

First Advisor

Arkadan, Abdul-Rahman A.

Second Advisor

Yaz, Edwin E.

Third Advisor

Brown, Ronald H.

Abstract

The performance characteristics of synchronous generator systems in stand alone or no-break power transfer (NBPT) mode are predicted using computational electromagnetics in conjunction with Artificial Neural Network based Data Mining approach. The approach has two main building blocks: The first involves the use of off-line FE-SS method to set up a data base. The second component uses Data Mining technique and is used to predict "quickly" and "accurately" the performance characteristics of highly nonlinear systems. The approach was applied stand alone synchronous generator systems typically used in aerospace applications. In this work, the system included prototype three-phase, 90 kVA, 208 V, 400Hz, synchronous generators feeding a combination of ac and dc loads in stand alone and NBPT mode.

Share

COinS

Restricted Access Item

Having trouble?