Date of Award
1-10-2024
Publication Type
Thesis
Degree Name
M.A.Sc.
Department
Electrical and Computer Engineering
Keywords
Data driven methods;faults;Sequential learning;Transmission line
Supervisor
Jonathan Wu
Rights
info:eu-repo/semantics/openAccess
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Abstract
With the continuous increase in the power demand and the incidents occurring in the power transmission systems there is a need of fast and accurate solution to identify the class and location of fault. The goal of this study is to create a new single-ended fault classification method using sequential models derived from the artificial neural network and an impedance-based fault location algorithm based on voltage and current analysis during discrete system states that occur during the operation of three phase distribution feeder systems. For further validation, the suggested technique is illustrated utilizing IEEE- 13 distribution feeders. With the availability of voltage and current measurements, real-time analysis and monitoring of power transmission line is achievable. In this study, a method is provided for estimating the type and location of fault using the root mean square voltage and current measurements measured during the instant of fault. The approach relies on optimization of the LSTM network i.e. sequential learning model and an impedance-based algorithm. To show the effectiveness of the proposed classifier, it has been compared with other data driven methods, in order to analyse a case study of a simulated transmission line is provided with precision value of 98.89 % for the proposed classifier model and the accuracy of 98.52% for fault distance estimation. The study and analysis will provide an overview of current and future efforts with the goal of advancing the work closer to industry implementation.
Recommended Citation
KAMAL, AREEB, "Fault detection and classification in an overhead transmission line using single ended measurements and sequential learning models." (2024). Electronic Theses and Dissertations. 9198.
https://scholar.uwindsor.ca/etd/9198