Neurocomputing 168 (2015) 983–993 (ELSEVIER) Impact Factor =  2.08

New algorithm for detection and fault classification on parallel
transmission line using DWT and BPNN based on Clarke’s
transformation

Abdullah Asuhaimi Mohd Zin a, Makmur Saini a,c,*, Mohd Wazir Mustafa a,

Ahmad Rizal Sultan a,c, Rahimuddin b

a Faculty of Electrical Engineering, Universiti Teknologi Malaysia (UTM), Johor Bahru 81310, Malaysia

b Faculty of Engineering, Universitas Hasanuddin, Makassar 90245, South Sulawesi, Indonesia

c Politeknik Negeri Ujung Pandang, Makassar 90245, South Sulawesi, Indonesia

* makmur.saini@fkegraduate.utm.my 

Article history:

Received 2 July 2014

Received in revised form

23 February 2015

Accepted 10 May 2015

Communicated by Hongli Dong

Available online 19 May 2015

a b s t r a c t

This paper presents a new algorithm for fault detection and classification using discrete wavelet

transform (DWT) and back-propagation neural network (BPNN) based on Clarke’s transformation on

parallel transmission. Alpha and beta (mode) currents generated by Clarke’s transformation were used to convert the signal of discrete wavelet transform (DWT) to get the wavelet transform coefficients (WTC) and the wavelet energy coefficient (WEC). Daubechies4 (Db4) was used as a mother wavelet to decompose the high frequency components of the signal error. The simulation was performed using PSCAD/EMTDC for transmission system modeling. Simulation was performed at different locations along the transmission line with different types of fault and fault resistance, fault location and fault initial angle on a given power system model. Four statistic methods utilized are in the present study to determine the accuracy of detection and classification faults. The results show that the best Clarke transformation occurred on the configuration of 12-24-48-4, respectively. For instance, the errors usingmean square error method, the errors of BPNN, Pattern Recognition Network and Fit Network are0.03721, 0.13115 and 0.03728, respectively. This indicates that the BPNN results are the lowest error.

& 2015 Elsevier B.V. All rights reserved

Introduction

Parallel transmission lines have been widely used in modern power systems to improve power transfer, reliability and security for the transmission of electrical energy. The possibility of different

configurations of parallel lines, combined with mutual coupling effects, makes their protection a challenging problem, therefore a fast and reliable protection is needed for rapid fault detection and accurate estimation of fault location errors. This is vital to support the maintenance and restoration services to improve the continuity and reliability of supply. Therefore, a parallel transmission line requires special consideration in comparison with the single transmission line, due to the effect of mutual coupling on the parallel transmission line. It must also comply with the standards of IEEE.STD.114 2004 [1]. One major advantage of parallel transmission is availability of transmission

network during and after the fault. This paper applies discrete wavelet transform (DWT) and backpropagation neural network (BPNN) using Clarke’s transformation to determine the fault detection and classification on the parallel transmission line. This study presents a different approach called alpha-beta transformation based on Clarke’s transformation;which is also a transformation of a three-phase system into a two-phase system [2,3], where the result of the Clarke’s transformation

is changed into discrete wavelets transform. Recently, some applications of wavelet transforms in power systems are power system protection, power system transients, partial discharge, transformer protection and condition monitoring. Among all, the power system protection continues to be a major application area of wavelet transform in power systems [4], while Artificial Neural Network (ANN) continues as an efficientpattern recognition, classification and generalization tool that motivates many algorithms based on ANN to be used for fault detection and classification [5]. In recent years, the combination of ANN and wavelet has been applied on researches regarding various power system planning and operation problems [6,7], as well as power quality [8], fault classification [9], state estimation and control system [10,11]. This paper presents the development of a new decision algorithm for use in the protective relay for fault detection and classification. In this method, fault conditions are simulated using EMTDC/PSCAD [12]. Current waveforms obtained from the simulation are then extracted using Clarke transformation and wavelet transformation. Decision algorithm, therefore, is built based on back-propagation neural network. In this study, the validity of the proposed algorithm had been tested using various initial error angles, location and broken phase

errors. In addition, the results of the proposed algorithms were compared with and without wavelet transform based Clarke transformation.

Conclusion

This paper proposes a technique of using a combination of discrete wavelet transform (DWT) and back-propagation neural networks (BPPN) with and without Clarke’s transformation, in order to identify fault classification and detection on parallel circuit transmission lines. This technique applies Daubechies4 (Db4) as a mother wavelet. Various case studies have been studied, including variation distance, the initial angle and fault resistance. This study also includes comparison of the results of training BPPN and DWT with and without Clarke’s transformation, where the results show that using Clarke’s transformation will produce smaller MSE and MAE, compared to without Clarke’s transformation. Among the three structuresArchitects result was the best, which was 12-24-48-12. Four statistical methods are utilized in the present study to determine the accuracy of detection and classification faults, suggesting that the Back Propagation Neural Network results in the lowest error thus it is the best compared with Pattern Recognition Network and Fit Network.

Acknowledgments

The authors would like to express their gratitude to Universiti Teknologi Malaysia, the State Polytechnic of Ujung Pandang, PT. PLN (Persero) of South Sulawesi and the Government of South

Sulawesi, Indonesia for providing the financial and technical support for this research.