Hybrid methodology for fault distance estimation in series compensated transmission line. Issue 5 (1st May 2013)

• Record Type: Journal Article
• Authors: Ray, Papia; Panigrahi, Bijaya Ketan; Senroy, Nilanjan
• Is Part Of: IET generation, transmission & distribution. Volume 7:Issue 5(2013)
• Page Start: 431
• Page End: 439
• Publication Date: 2013-05-01
• Subjects: discrete wavelet transforms -- fault location -- neural nets -- power engineering computing -- power transmission lines; fault inception angle -- fault resistance -- ANN -- forward feature selection algorithm -- discrete WPT -- discrete WT -- signal extraction -- artificial neural network -- wavelet packet transform -- hybrid fault location technique -- series compensated transmission line -- fault distance estimation; Electric power production -- Periodicals; Electric power transmission -- Periodicals; Electric power distribution -- Periodicals; 621.3105
• DOI: 10.1049/iet-gtd.2012.0243
• Languages: English
• ISSNs: 1751-8687

Abstract

Abstract : This study proposes an accurate hybrid fault location technique, with wavelet transform (WT) and wavelet packet transform (WPT) combining artificial neural network (ANN) in series compensated transmission line. Proposed method uses single end measurement of current and voltage signal of one cycle duration from the inception of fault. Thereafter features of collected signal are extracted by discrete WT and WPT and optimal feature is selected using forward feature selection algorithm. Selected features are then fed as input to ANN for fault location. Feasibility of the proposed method has been tested for all ten types of fault on two different test systems. Testing of the proposed method is done for varying fault resistance and fault inception angle over different fault location for each type of fault. Performance of the proposed method is evaluated in terms of absolute error and mean error which is found to be quite satisfactory. Simulation result shows that maximum error of less than 0.35% and mean error of less than 0.25% is achieved, which demonstrate high accuracy and robustness of the proposed technique. Results are compared with other different feature selection techniques.