Physical and AI Based Diagnostics of Electrical Machines

OData support
Dr. Vajda István
Department of Electric Power Engineering

The topic of my thesis is focused on the various types of electrical machine faults. I examine what are the main reasons of certain faults, what kind of mechanism causes these failures. I present several physical models which are useful to describe these mechanisms.

I deal with the most common artificial intelligence methods. Applications is presented through some samples.

The current signature analysis is a very useful tool to detect different faults. I demonstrate the usage of this technique, especially on harmonic detection. I summarize the effects of these harmonics and I present some detection methods.

In the final part I make a diagnosis on a synchronous generator. I show some simulation results. With FEMM and Agros2D finite element programs I simulate the airgap flux density and after with Matlab I evaluate the harmonic content of this sign.


Please sign in to download the files of this thesis.