Essay:

Essay details:

  • Subject area(s): Engineering
  • Price: Free download
  • Published on: 7th September 2019
  • File format: Text
  • Number of pages: 2

Text preview of this essay:

This page is a preview - download the full version of this essay above.

Analysis of  Wavelet Transform and ANFIS Based Fault Detection and Classification

Puja Bharti1, M.A.Ansari2, Y.K.Chauhan3 and Ashish Kumar4

1,2,3,4 Department of Electrical Engneering,

Gautam Buddha University, Greater Noida,U.P., India

{pujabharti025, mahmadiitr,ashishelectrical025}@gmail.com, [email protected]

Abstract. Abnormal conditions are occurred in the power system. The intelligent techniques can be applied to know about the fault nature. The identification of fault in the power system is very important for the safeguard of the equipment's from the abnormal conditions which arises due to fault. When the faults are identified then the classification of the fault and after that clear the fault for studying the stability in the system is paramount. For the identification and classification, the use of intelligent techniques in which identification through wavelet transform and classification through ANFIS (Adaptive Neuro Fuzzy Inference System) is proposed in this paper. Wavelet transform identifies the fault in terms of energy and classification of fault by the ANFIS as well as fuzzy inference system which display the type of fault. The fault identifications are based on the fault current which is more dominant for line faults.

Keywords: Power system, Fault Identification, Classification, Wavelet Transform, ANFIS.

1   Introduction

The power system is divided into three main components namely Generation, Transmission and distribution. Generation is very basic component like to generate the power to the system. In India, the generation of voltage level varies between 11Kv to 33Kv and distribution Voltage level varies between 11Kv to 415V [2]. However, in the transmission, as the voltage level varies from high voltages to extra high voltage, it creates the case of faults. So, the safety & stability of the transmission is necessary in the power system. Life without electricity is not possible and consumption of electricity increases gradually. So, firstly identify the fault and after that classify the fault. When the faults are clear then system comes under the stable condition [1]. Normally, the faults which occur in the system are symmetric and unsymmetrical. Symmetrical faults are also called balanced or three phase faults. These types of faults arise when three conductors touch each other with ground or without ground likely (a-b-c) or (a-b-c-g). It occurs 2-5% in the power system [2]. When symmetric faults occur, then it remains in balance condition but it damages the system too much. Unsymmetrical faults are also called unbalanced fault. Unbalanced faults are three types of faults namely- Single line to ground fault (a-g, b-g, c-g), Double line to ground fault (a-b-g, b-c-g, c-a-g), Line to Line fault (a-b, b-c, c-a). Single line to ground (SLG) fault occurs when one phase conductor touches with the ground then this type of fault are arises. These types of faults are very common in the power system and it occurs 65-70% fault in the power system. Double line to ground (DLG) fault occurs when two phase conductor touches with the ground then this type of fault are arises in the power system. It occurs 15-20% fault in the power system. Line to Line (LL) fault arises when two phase conductor touches each other and it occurs in the power system 5-10%. [2].

2   Detection of Faults Using Wavelet Transform

The wavelet transform helps to find the energy of various signals (currents) associated in the power system due to that the identification of faults is carried out. The wavelet transform takes the signal from the workspace and the workspace data comes from the simulation results and the current waveform is considered for identification and classification of the line fault. Wavelet transform decomposes the signal for more accurate studies [3]. At suitable level and wavelet are used to decompose the signal in the power system. In this paper, Daubechies (db.) wavelets are used and their level are 8. The signal decomposed in two components likely, approximate and detail coefficient as shown in fig.1. In which approximate coefficient has the high scale and low frequency as well as detail coefficient has the low scale and high frequency [4]. Approximate (A) and detail (D) help to filter the signal and after that to find the energy for the power system that helps us to identification of fault in the power system.

Fig. 1. (a) 'S' is the signal which is decomposed and it divided into two components A and D mean approximate and detail coefficient and c is the constant value. (b) A and D are filters.

3   Classification of Fault Using ANFIS

ANFIS is used to classification of fault in the power system. ANFIS is a hybrid system which is the mixture of Fuzzy and ANN [5]. In the ANFIS, the data is trained for the results which contain the types of fault. ANFIS takes the data which is decomposed and filter by the wavelet transform [6]. The data is put into the ANFIS and check the result and Fuzzy logic controller display the results. Fuzzy takes the signal in term of FISs (Fuzzy inference system) and then declares the type of fault [7].

4   Flow Chart for Detection and Classification of Faults in the Power System

Fig. 2. Flow Chart for Detection and Classification of faults.

5   Simulink Model of Proposed Fault Detection in Power System

In this Fig.3, It has one generator and one transformer. There are three buses in which Bus-1 is slack bus and is called Swing or Reference bus. Three phase faults are connected in bus -2 in which creation of fault by these block. All the fault declared current base only. Scope is used to see the graph. All the calculation for Bus-2 only.

Fig. 3. Simulation of proposed model for Fault Identification and Classification

 Fig. 4. Current waveform of SLG fault (abnormal condition of Bus-2). [Ia max=170A, Ib max=Ic max=101.5A]

 Fig. 5. Current waveform of DLG fault (abnormal condition of Bus-2). [Ia max=Ib max=170A, Ic max=101.5A]

Fig. 6. Current waveform of LL fault (abnormal condition of Bus-2). [Ia max=160A, Ib max=150A, Ic max=100A]

 Fig. 7. Current waveform of three phase fault (abnormal condition) of Bus-2. [Ia max=Ib max=Ic max=175A]

In Fig. 4 to 7, there is current waveform in different situation. Variation of current which show abnormal in some graph that has the sign of fault and these type of fault are also detect by put the data into the wavelet transform by doing this energy of current will shows that there is fault. It mean detection of fault check by wavelet transform as well as scope graph Fig. 4 to 7 shows the fault current.

6   ANFIS Results and Discussion

In the ANFIS, Trained the data for classification of fault. All the data taken from matrix form and data is taken from the workspace which comes from wavelet transform who give the energy of current waveform of bus-2.

Fig. 8. (a) Checking data in ANFIS and graph shows in FIS output. (b) Surface view of Three phase fault where the phase (a-b-c or a-b-c-g).

In the Fig. 8 (a) Checking data, checking data shows the output 1 '+' like this and '1' denotes the three phase fault and their phase (a-b-c or a-b-c-g). Similarly, all the fault are the output from 1 to 10. From this basis, Classification of fault are occurs. There are the surface view of the Three Phase fault where the phase (a-b-c or a-b-c-g). It start from 0 and to 1. Surface view start from 0 and end with the value 1 for three phase fault.

Fig. 9. Fuzzy inference system to display the type of fault.

Table 1.  Type of fault, phase conductor and display output for the classification of fault.

Type of Fault Phase of conductor Display Output

Three Phase Fault A-B-C or A-B-C-G 1

DLG A-B-G 2

DLG B-C-G 3

DLG C-A-G 4

LL A-B 5

LL B-C 6

LL C-A 7

SLG A-G 8

SLG B-G 9

SLG C-G 10

In the Fig. 9, the fuzzy logic controller display the type of fault as shown in table 1. Fuzzy logic controller take the value from the workspace in FISs form whereas the constant value also taken from workspace in the matrix form. In the Table 1, the display shows the output in display numerical value from 1 to 10 and from this value, classification of faults are occurs. Various type of line fault are classified by the Fuzzy Inference System. When three phase fault are occurs in the system then it shows the value '1' and from this out display, classified the fault. Similarly, All these type of fault like the SLG fault are classified when output come '8/9/10' then classified the fault and which phase also classified by value '8/9/10'. If 8 are come then 'A-G' Phase fault are occurs and it is SLG fault. From this, Classified all fault and after that clear the fault as soon as possible from damage the equipment from the hazard.                        7   Conclusion

Current is the most important factor for the line. For identification of faults, we used the wavelet transform and for classification the ANFIS. Wavelet Transform helps to find the energy of current and Energy of current is varied in fault condition. In fault condition, Energy of faulty phase is more as compared to normal condition due to this condition identify the fault. ANFIS helps to classify the fault by the Display output in table 1. In the FISs system, output display controlled by the fuzzy logic controller.

References

1. Badri Ram and D N Vishwakarma: Power System Protection and Switchgear. Tata McGraw Hill. 2nd Edition, 2007.

2. C. L. Wadhwa: Electrical Power System. New Age International Publishers. 6th Edition, 2005.

3. Mayuresh Rao and R.P. Hasabe: Detection and Classification of Faults on Transmission Line Using Wavelet and Neural Network. International Journal of Advanced Electrical and Electronics Engineering, (IJAEEE). 2278-8948. Volume-2. Issue-5. 2013.

4. S. Saha, M. Aldeen and C.P.Tan: Fault detection in transmission networks of power systems. Scince Direct Electrical Power and Energy Systems 33. pp. 887'900. 2011.

5. Kevin M. Passino and Stephen Yurkovich: Fuzzy Control. 1st Edition. Addison Wesley Longman. Inc. California. 1998.

6. Huseyin Eristi: Fault diagnosis system for series compensated transmission line based on wavelet transform and adaptive neuro-fuzzy inference system. Elsevier. Measurement. Vol 46. pp. 393'401. 2013.

7. J. Zhang, Z.Y. He, S. Lin, Y.B. Zhang, Q.Q. Qian: An ANFIS-based fault classification approach in power distribution system. Elsevier. Electrical Power and Energy Systems. Vol. 49. pp. 243'252. 2013.

...(download the rest of the essay above)

About this essay:

This essay was submitted to us by a student in order to help you with your studies.

If you use part of this page in your own work, you need to provide a citation, as follows:

Essay Sauce, . Available from:< https://www.essaysauce.com/essays/engineering/essay-2016-04-07-000At-.php > [Accessed 20.10.19].