Detection and Classification of Faults on Transmission Line Using Wavelet Transform and Neural Network

Mayuresh Rao & R. P. Hasabe

Abstract - This paper presents a discrete wavelet transform and neural network approach to fault detection and classification in transmission line faults. The detection is carried out by the analysis of the details coefficients energy of the phase signals, and as an input to neural network to classify the faults on transmission lines. Neural network perform well when faced with different fault conditions and system parameters. Keywords - Fault detection, classification, transmission line, neural networks.

carried out by the analysis of the wavelets coefficients energy related to phase currents.[1]. The proposed algorithm consists of time-frequency analysis of fault generated transients using Wavelet transform, followed by pattern recognition using artificial neural network. The wavelet transform has the ability to detect the faults, time localization is property very good. The neural networks have the ability to learn, generalize and parallel processing, have made their applications for many systems ideal. The use of neural network as pattern classifiers is among their most common and powerful applications. The MATLAB/SIMULINK is used to generate the fault signals and verify the correctness of the algorithm.[9].Wavelet toolbox and Neural Networks toolbox are used. The proposed scheme is insensitive to variation of different parameters such as fault type, fault resistance. ANN based techniques have been used in power system protection and encouraging results are obtained. Neural networks are used for different applications as classification, pattern recognition etc.In classification, the objective is to assign the input patterns to one of the different classes.[9]. II. DISCRETE WAVELET TRANSFORM Discrete Wavelet Transform is found to be useful in analyzing transient phenomenon such as that associated with faults on the transmission lines. The fault signals are generally non stationary signals, any change may spread all over the frequency axis. Under this condition the Fourier transform techniques are less efficient in tracking the signal. S.T.F.T uses a fixed frequency window to localize sharp transitions for non-stationary signals. The practical power system faults encountering the change in frequencies. The wavelet transform technique is well suited to wide band signals that may
ISSN (Print): 2278-8948, Volume-2, Issue-5, 2013

I.

INTRODUCTION

Along with other electrical components, the transmission line suffers from the unexpected failures due to various faults. Protecting of transmission lines is one of the important tasks to safeguard electric power systems. For safe operation of EHVAC transmission line systems, the protection system need to able to detected, classified, located accurately and cleared is fast as possible to maintain stability in the network. The protective systems are required to prevent the propagation of these faults. The occurrence of any transmission line faults gives rise to transient condition. Fourier transform technique is used for detecting the transmission line faults. Fourier transform gives information about all frequencies that are present in the signal but does not give any information about the time at which these frequencies were present.[1]. Wavelet transform has the advantage of fast response and increased accuracy as compared to conventional techniques. The wavelet transformation is a tool which helps the signal to analyze in time as well as frequency domain effectively. It uses short windows at high frequencies, long windows at low frequencies. Using multiresolution analysis a particular band of frequencies present in the signal can be analysed. The detection is

International Journal of Advanced Electrical and Electronics Engineering, (IJAEEE)

not be periodic and may contain both sinusoidal and non sinusoidal components .The wavelet transform has the ability focus on short time intervals for high frequency components and long time intervals for low frequency components. Multi-Resolution Analysis(MRA) is one of the tools of Discrete Wavelet Transform(D.W.T), which decomposes original, typically non-stationary signal into low frequency signals called approximations and high frequency signals called details, with different levels or scales of resolution. It decomposes signal into different scales and resolutions.

neuron is the result of applying a specific function, known as a transfer function, to the sum of its inputs, plus a threshold value called bias. With these general characteristics it is able to develop different network structures. ANNs have a high degree of robustness and ability to learn. ANNs are prepared to work with incomplete and unforeseen input data.[8]. ANNs have the ability to learn from examples. Once the network is trained, it is able to properly resolve the different situations that are different from those presented in the learning process. The weights of the network are adjusted automatically to get a particular target output for specific input. The neural networks can have several layers. Each neurons in one layer has direct connections with other neurons in the another layer. There can also be hidden layers. By inserting hidden layers, increasing its size and number, the non-linear model of the system is developed. The multilayered feed forward network has the ability of handling complex and nonlinear input-output relationship with hidden layers. In this method, errors are propagated backwards; the idea of back- propagation algorithm is to reduce errors until the ANN learns the training data. The training begins with the random weights and the goal is to adjust them so that the error will be minimal. The multilayered feed forward n/w has been chosen to process the prepared input data obtained from the W.T. ANNs are selected depending upon: 1) ANN architecture 2) Transfer function of each neuron 3) Initial weights and biases. 4) Learning Rule. IV. TRANSMISSION LINE MODEL Load1 G Load2 .
Fig. 2: Transmission Line Single Line Model.

Fig.1: Wavelet filter Bank

It uses filters with different cut off frequencies to analyze a signal (image in our case) at different resolutions. The signal is passed through a series of high-pass filters, g also known as wavelet functions, to analyze the high frequencies and it is passed through a series of low-pass filters, h also known as scaling functions, to analyze the low frequencies. In first decomposition signal is decomposed into D1 and A1, the frequency band of D1 and A1 is /2- /4, 0 /4.where is the sampling frequency.In the second decomposition step the A1 is decomposed to give A2 and D2.The frequency band of D2 component is /4 /8, and A2 is 0- /8.the signal of desired frequency component can be obtained from repetitive decompositions. The mother wavelet determines the filters h and g used to analyse signals. In this paper Daubechies 4 wavelet was chosen because of its success in detecting faults [4], [5]. III. ARTIFICIAL NEURAL NETWORKS Artificial Neural Networks (ANNs) are inspired by biological nervous systems and they were first introduced as early as 1960.[8]. A neural network is a massively parallel distributed processor that has a natural propensity for storing experiential knowledge and making it available for use. Artificial Neural Networks simulate the natural systems behavior by means of the interconnection of processing basic units called neurons. Neurons are highly related with each other by means of links. The neurons can receive external signals or signals coming from other neurons affected by a factor called weight. The output of the

Load3

Load4

In this, above shown model of transmission line a 220kv, 200km transmission line is taken. Generator rating of 500MW is connected at one end and various faults are simulated on that line by varying various parameters. Ratings of power system model are shown in Table No. I.As shown in Fig.3 model is prepared in MATLAB7.10.

ISSN (Print): 2278-8948, Volume-2, Issue-5, 2013

International Journal of Advanced Electrical and Electronics Engineering, (IJAEEE)

Table No. I Transmission Line Parameter. 1. Generator 500MVA,13.8Kv,50Hz,synchrono us generator pu model

power system. The fault resistance, fault location, and fault type are changed to generate different training patterns.
DATA ACQUISITION OF CURRENT SIGNALS

2. Transformer1 13.8KV/220KV, 500MVA. 3. Transfomer2 4. LOAD1 5. LOAD2 6. LOAD3 7. LOAD4 8 220kv/13.8kv, 500MVA. 50MW, 220kv, 50MW, 1Mvar, RL load. 50MW, 220KV,50MW, 1MVar.,RL load 13.8KV,50Hz,40Mw,RL load 13.8kv,40Mw,50Hz.,RL load

D.W.T MULTIRESOLUTION . ANALYSIS

FEATURE EXTRACTION, ENERGY CALCULATION

Transmission Length=200km. line

ANN BASED CLASSIFICATION OF FAULTS Fig. 3 : Process of fault detection and classification.

The transmission line positive and zero sequence parameters areR1=0.10809/km, R0=02188/km, L1=0.00092H/km,L0=0.0032H/km,C1=1.25*108 f/ km, C0=7.85*109 f/km. The distributed parameter model of transmission line is considered for analysis. The current signals of three phases and neutral are taken at sampling frequency of 20kHz.Different faults as single line to ground ,double line to ground ,line to line ,3line to ground are simulated on transmission line at various system conditions V. DESIGN OF FAULT DETECTION AND CLASSIFICATION The design process of proposed fault detection and classification approach is as follows, 1) Creating of the power system model, data acquisition of current and voltage signals. 2) Changing the system parameters and, data acquisition of current and voltage signals and storing and analyzing results. 3) Application of D.W.T on the current signals and calculating detail coefficients energy. 4) Selection of suitable ANN topology for given application. 5) Training of ANN and validation of the trained ANN using test patterns to check its correctness and generalization. Combination of different fault conditions are to be considered and training patterns are required to be generated by simulating different kinds of faults on the

VI. FAULT DETECTION The signals taken from the scope are filtered, sampled at 20 kHz sampling frequency. Then DWT is applied upto level 5, and detail coefficients and approximate coefficients are calculated and detail coefficients energy is calculated. Therefore, for sampling frequency of 20kHz,the detail level 1 coefficients are in the frequency range of of 5kHz10kHz, detail coefficients at level 2 are in the frequency range of 2.5-5kHz., the detail coefficients at level 3 are in the frequency range of 1.25-2.5kHz, the detail coefficients at level 4 are in the frequency range of 0.625-1.25kHz, and detail coefficients at level 5 are in the range of 0.3125-0.625kHz.A moving data window of one cycle(400samples) is taken and decomposition is done and energy of the details coefficients at level 5 is obtained for each data window. Then, we come to know that detail level 5 contains highest amount of energy than the level 4.As the fault signals contain the high amount of harmonic components, the energy of the signal increases at the occurrence of fault as shown in Fig.4 Here, for detecting the fault we have taken a difference of energies between two adjacent windows. The energy of detail coefficients for a window is given by,[10]. Ed (k) =
2 N i=1 Dl (i)

(1)

Where,k=window number,l=level of the DWT,N=length of detail coefficients at level l.

ISSN (Print): 2278-8948, Volume-2, Issue-5, 2013

International Journal of Advanced Electrical and Electronics Engineering, (IJAEEE)

Fig. 4: Energy of the detail level 5 vs. window number.

For accurately detecting the presence of faults, the difference between the two consecutive energies of the moving windows is calculated as; and shown in Fig.5.[7]. F. D k = F. D k 1 + [Ed k Ed k 400 ] (2) In this sampling frequency of 20 kHz give 400 samples for each cycle of 20ms.Here, moving window slides taking only1 new samples at each move and keeping 399 previous samples. So the one cycle corresponds to nearly 400 data windows.
2 2

neurons and 4output neurons are selected. The two hidden layers are selected. Feed forward multilayer back propagation neural network is selected. The average value of energies of currents half cycle after the occurrences of fault are given as input to the neural network, along with the three lines energies, zero sequence current energy is also given as fourth input to the neural network. Three outputs shows the statuses of the three phases, if fault is present it is shown by the presence of 1, otherwise with presence of 0.Similarly fourth output indicates the ground fault. Ground is involved in the fault will be indicated by the presence of 1, otherwise it is presented by 0.This is shown in Table No.II. Generation of different training patterns is done as shown in Table No.III. Table No. II Fault Type AG BG CG AB BC CA ABG BCG CAG ABC A 1 0 0 1 0 1 1 0 1 1 B 0 1 0 1 1 0 1 1 0 1 Table No.III C 0 0 1 0 1 1 0 1 1 1 G 1 1 1 0 0 0 1 1 1 0

Fig. 5 : F.D index for single line to ground fault vs. window number

Type of fault Location of fault (%) Fault resistance

LG, LLG, LL, LLL. 20,30,40,50,60,70,80 5,10,15,20 .

The S.F.D value is compared with threshold value for consecutive 20data windows, and then decision is made whether fault is permanent or short time. By using these F.D index the faults can be accurately detected. For different phases different threshold values are set and the fault detection is achieved. The transient energy is present mainly during fault inception and clearing. The high frequency content energy is smaller than the low frequency content energy of the current signals. VII. NEURAL NETWORK BASED FAULT CLASSIFICATION All different faults are simulated for different conditions and training patterns are generated from the energy values of the detail coefficients. The 4input

Feature information is acquired by the network through a learning process and interconnection synaptic weights are used to store the knowledge. And, then the neural network can generalize, and can give response to any untrained data. For training neural network different fault conditions are simulated, features are extracted and network is trained. At 7 different locations on the transmission line fault is created at 20, 30, and 40,50,60,70,80% of the transmission line length from the sending end,4 different values of fault resistances can be used and total 10 different faults are created, and this gives7*4*10=280cases for training neural network. Once the network is trained with given training parameters, the network can be simulated for any input and target values and performance measures can be found.
ISSN (Print): 2278-8948, Volume-2, Issue-5, 2013

International Journal of Advanced Electrical and Electronics Engineering, (IJAEEE)

The different training algorithms are present to train the neural network; they use the gradient of the performance function to determine how to adjust the weights to minimize a performance function. The gradient is determined using back propagation technique, which involves performing computations backwards through the network. A variation of back propagation algorithm called Levenberg-Marquardt (LM) algorithm was used for neural network training, since it is one of the fastest methods for training moderate sized feed forward neural networks. LM algorithm to weight update is given by, xk+1 = xk [J T J + I]1 J T e (3)
Fig. 8 : Performance Curve

Where J is Jacobean matrix that contains first derivatives of the network error with respect to the weights and biases, e is a vector of network errors. J T J is an approximation of the Hessian Matrix, the gradient is J T e and is the scalar affecting performance function.So,LM algorithm based method for training neural network is much faster than the other methods. Multilayered feed forward network can be prepared as shown in Fig.6

The data used for training data division is done randomly; training function used is LM algorithm. Performance function used is Mean Square Error .The performance goal taken is 106 .Performance curve is as shown in Fig.8.For 4-22-4 and 4-10-4 configurations cannot distinguish between the faults with ground and faults without ground. VIII. TEST RESULTS A validation data set consisting of different fault types was generated using the power system model shown in fig. Table no. IV. Test results
Fault type AG Fault location (%) 30% 50% 50% 50% 30% 70% 70% Fault resistance . 10 15 10 10 10 5 5 A 1.0001 0 1 1*10^-3 1.00 0.9996 1.018 B C G 1.00 1.00 0.998 1.00 0.00 1.000 0.0526

2*10^-3 4*10^-3 1.00 0 0.000 1.00 -3*10^-4 1.0847 0.9989 1.00 1.00 0.999 0.997 0.1587

Fig.6: Multilayer feed forward network for fault classification

Fig.7: 4-22-22-4 tansig,logsig,purelin configuration.

AB

Network with 2 hidden layers worked out to be better than the 1hidden layer network.4-22-22-4 configuration give better results than the 4-22-4, 4-10-4 configurations.[9]. Activation functions used for the hidden layers are tansig and logsig output layer is purelin.The network is as shown in Fig.7.

The validation test patterns were different than they were used for the training of the neural network .For different faults of the validation set, fault type, fault location and fault resistance values are changed to investigate the effects of these factors on the performance of the proposed algorithm. Test results are as shown in Table IV.

ISSN (Print): 2278-8948, Volume-2, Issue-5, 2013

International Journal of Advanced Electrical and Electronics Engineering, (IJAEEE)

IX. CONCLUSION In this paper accurate fault detection and fault classification technique is designed. This technique depends upon the current signals. The features are extracted from the current signals by using wavelet transform. The feature vector is then given as input to the neural network. The capabilities of neural network in pattern classification were utilized. Simulation studies were performed and the performance of the scheme with different system parameters and conditions was investigated. Though the paper deal with fault classification can be extended to other power system protection problems such as finding fault location. X. REFERENCES [1] K. Gayathri and N. Kumarappan, Comparative Study of Fault Identification and Classification on EHV Lines Using Discrete Wavelet Transform and Fourier Transform Based ANN, World Academy of Science, Engineering and Technology pp.822-831.2008 D. V. Coury, D. C. Jorge, Artificial Neural Network Approach To Distance Protection Of Transmission Lines., IEEE Transactions on Power Delivery, Vol. 13, No. 1, January 1998. H. Khorashadi-Zadeh, M. R. Aghaebrahimi, A Novel Approach to Fault Classification and Fault Location for Medium Voltage Cables Based on Artificial Neural Network, World Academy of Science, Engineering and Technology, pp.1100-1103.2008.

[4]

V.S Kale,S.R. Bhide,P.P.Bedekar,Faulted Phase Selection on Double circuit Transmission Line using Wavelet Transform and Neural Network.,Third International Conference On Power Systems,Kharagpur,INDIA,December 2729. A. Abdollahi,S. Seyedtabaii, Comparison of Fourier & Wavelet Transform Methods for Transmission Line Fault Classification, The 4th International Power Engineering and Optimization Conf. (PEOC2010), Shah Alam, Selangor, MALAYSIA: 23-24 June 2010. V.S.Kale, S.R.Bhide, P.P.Bedekar, G.V.K.Mohan, Detection and Classification of Faults on Parallel Transmission Lines using Wavelet Transform and Neural Network, International Journal of Electrical and Computer Engineering,pp.1063-1067,16, 2008. N. Zamanan, M. Gilany, A Sensitive WaveletBased Algorithm for FaultDetection in Power Distribution Networks ACEEE Int. J. onCommunication, Vol. 02, No. 01, Mar 2011.pp.46-50. S. Haykin, Neural Networks, IEEE Press, New York, 1994. H. Demuth and M. Beale, "Neural Network Toolbox - For Use with Matlab 1992. Mamta Patel, R. N. Patel, Fault Detection and Classification on a Transmission Line using Wavelet Multi Resolution Analysis and Neural Network, International Journal of Computer Applications Volume 47 No.22, June 2012.pp.27-33.

[5]

[6]

[7]

[2]

[8] [9] [10]

[3]

ISSN (Print): 2278-8948, Volume-2, Issue-5, 2013