Detection and classification of high impedance faults in power distribution networks using ART neural networks

Adaptive Resonance Theory (ART) neural networks have several interesting properties that make them useful in the area of pattern recognition. Many different types of ART-networks have been developed to improve clustering capabilities. In this paper, five types of ART neural networks (ART1, ART2, ART2-A, Fuzzy ART and Fuzzy ARTMAP) are applied to detect and classify high impedance faults (HIF) in distribution networks. The features are extracted by applying TT-transform to one cycle of fault current signal. These features include energy, standard deviation and median absolute deviation. Then, they are applied to ART neural networks to detect and classify high impedance fault with broken conductor on gravel, asphalt and concrete, unbroken conductor on tree and also no fault condition. Finally, the results of these ART neural networks are compared with each other.

[1]  Lalu Mansinha,et al.  Localization of the complex spectrum: the S transform , 1996, IEEE Trans. Signal Process..

[2]  Shyh-Jier Huang,et al.  High-impedance fault detection utilizing a Morlet wavelet transform approach , 1999 .

[3]  Stephen Grossberg,et al.  Fuzzy ART: Fast stable learning and categorization of analog patterns by an adaptive resonance system , 1991, Neural Networks.

[4]  David C. Yu,et al.  An adaptive high and low impedance fault detection method , 1994 .

[5]  Kishore Lankalapalli,et al.  Feature recognition using ART2: a self-organizing neural network , 1997, J. Intell. Manuf..

[6]  B. D. Russell,et al.  A digital signal processing algorithm for detecting arcing faults on power distribution feeders , 1989 .

[7]  C. Robert Pinnegar,et al.  A method of time-time analysis: The TT-transform , 2003, Digit. Signal Process..

[8]  Stephen Grossberg,et al.  Art 2: Self-Organization Of Stable Category Recognition Codes For Analog Input Patterns , 1988, Other Conferences.

[9]  Stephen Grossberg,et al.  ART 2-A: an adaptive resonance algorithm for rapid category learning and recognition , 1991, IJCNN-91-Seattle International Joint Conference on Neural Networks.

[10]  Stephen Grossberg,et al.  ART 2-A: An adaptive resonance algorithm for rapid category learning and recognition , 1991, Neural Networks.

[11]  S. Grossberg,et al.  ART 2: self-organization of stable category recognition codes for analog input patterns. , 1987, Applied optics.

[12]  D. Sutanto,et al.  High-impedance fault detection using discrete wavelet transform and frequency range and RMS conversion , 2005, IEEE Transactions on Power Delivery.

[13]  Stephen Grossberg,et al.  Fuzzy ARTMAP: A neural network architecture for incremental supervised learning of analog multidimensional maps , 1992, IEEE Trans. Neural Networks.

[14]  B. Don Russell,et al.  Distribution High Impedance Fault Detection Utilizing High Frequency Current Components , 1982, IEEE Power Engineering Review.

[15]  Roseli A. Francelin Romero,et al.  Investigating the potential of art neural network models for indexing and information retrieval , 2007, Int. J. Intell. Syst..

[16]  Adly A. Girgis,et al.  Analysis of high-impedance fault generated signals using a Kalman filtering approach , 1990 .

[17]  Stephen Grossberg,et al.  A massively parallel architecture for a self-organizing neural pattern recognition machine , 1988, Comput. Vis. Graph. Image Process..