Fault diagnosis for transmission lines using chromatic processing



Almajali, Ziyad
Fault diagnosis for transmission lines using chromatic processing. PhD thesis, University of Liverpool.

[img] Text
AlmajaliZiy_Jul2015_2019779.pdf - Unspecified
Available under License Creative Commons Attribution.

Download (12MB)

Abstract

Diagnosing the type of fault and its location in a transmission lines is performed by a variety of techniques and mainly relies on monitoring currents and voltages in the transmission line. Accurate fault diagnosis plays an important role in improving the overall system reliability, has a significant effect on the quality of service provided, improves the protection system efficiency, reduces power outage time and limits the risks and the economic losses. Transmission lines extend over wide areas and are exposed to vulnerable situation, to the harsh and uncontrolled environment random events (e.g.~lightning), this can lead to loss of lines due to various faults. This fact has an attraction for researchers to focus on utilising possible methods to improve protection system and supporting fault diagnosis solutions to overcome many of the transient fault conditions. This thesis explores an alternative method of fault diagnosis and location. The approach uses chromatic methodology to extract information from current and voltage waveforms from a simulated transmission line with different fault conditions. These waveforms are processed chromatically. The process involves two steps, filtering which is performed on a cycle by cycle basis of the three symmetrical components for each waveform, and then using the chromatic transformations to represent the outputs in an information space. Various chromatic models are available but the hue, lightness and saturation (HLS) model is employed in this study and the relation between changes in the waveforms and changes in the chromatic parameters forms the foundation for building the proposed diagnosis algorithms. A fault type classifier algorithm for the asymmetrical faults has been proposed for both, double and single line transmission systems. It employs the chromatic H parameter variation with the fault type for the negative sequence component. The processed waveforms are either the voltage or the current at a single terminal of the transmission line. L chromatic parameter values of the zero sequence component are incorporated in the algorithm to add the ground fault distinguishing element and the L parameter values of the rectified negative sequence component were used to support the classification decision even with high fault resistance. Another algorithm for fault location estimation for all types of faults has been used for the double transmission line system. It employs the L chromatic parameter values of the rectified positive sequence component. The processed waveforms are the current collected from both terminals of the transmission line. Finally, the proposed algorithms have been tested by variation of possible conditions of the faults, such as changing the fault location, the fault resistance, the line configurations and parameters, etc. In addition to robustness testing with different fault scenarios. Experimental results taken from a lumped parameter laboratory system have been also used to verify the outputs of the chromatic processing. The performance of the chromatic approach and other reported methods have been compared. The error of the chromatic method compares favourably with others. As such overall performance can be described as being good, this is encouraging and future work through proposing diagnostic tools for other power system components is needed.

Item Type: Thesis (PhD)
Additional Information: Date: 2015-08-11 (completed)
Subjects: ?? TK ??
Depositing User: Symplectic Admin
Date Deposited: 21 Jan 2016 11:57
Last Modified: 17 Dec 2022 01:54
DOI: 10.17638/02019779
URI: https://livrepository.liverpool.ac.uk/id/eprint/2019779