Date of Award

2009

Publication Type

Master Thesis

Degree Name

M.A.Sc.

Department

Electrical and Computer Engineering

First Advisor

Narayan Kar

Keywords

Applied sciences

Rights

info:eu-repo/semantics/openAccess

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Abstract

Synchronous generators form the principal source of electric energy in power systems. Dynamic analysis for transient condition of a synchronous machine is done under different fault conditions. Synchronous machine models are simulated numerically based on mathematical models where saturation on main flux was ignored in one model and taken into account in another. The developed models were compared and scrutinized for transient conditions under different kind of faults – loss of field (LOF), disturbance in torque (DIT) & short circuit (SC). The simulation was done for LOF and DIT for different levels of fault and time durations, whereas, for SC simulation was done for different time durations. The model is also scrutinized for stability stipulations.

Based on the synchronous machine model, a neural network model of synchronous machine is developed using neural network based characterization. The model is trained to approximate different transient conditions; such as – loss of field, disturbance in torque and short circuit conditions. In the case of multiple or mixture of different kinds of faults, neural network based clustering is used to distinguish and identify specific fault conditions by looking at the behaviour of the load angle. By observing the weight distribution pattern of the Self Organizing Map (SOM) space, specific kinds of faults is recognized. Neural network patter identification is used to identify and specify unknown fault patterns. Once the faults are identified neural network pattern identification is used to recognize and indicate the level or time duration of the fault.

Share

COinS