Development of DSP-Based Transmission Line Fault Detection

Abstract

This paper proposes a new algorithm to detect and classify faults in the electric power transmission line. Transmission line protection is an important issue in power system engineering because 85-87% of power system faults are occurring in transmission lines. This paper presents a technique to detect and classify the different shunt faults on transmission lines for quick and reliable operation of protection schemes. Discrimination among different types of faults on the transmission lines is achieved by the application of evolutionary programming tools. Further, fault signal data are imported into MATLAB for post-processing, and time-frequency analysis using Signal Processing Toolbox in MATLAB.

Introduction

I. INTRODUCTION

Transmission line fault detection and classification are two important features of a protective relay. Detection and classification of the fault must be performed accurately and as fast as possible for the protective relay to de-energize the faulted line to protect the power system from harmful effects such as a cascading outage, thermal overload, and voltage, angular instability, etc. The conventional algorithm for fault detection and classification is based on steady-state components only. That suffers from low protection speed, and also ignores many factors such as fault types, fault resistance, and fault transients which are important from a signal processing point of view for feature extraction of the faulty signal. The commonly used approach for feature extraction is to convert the time-domain signal into the frequency domain. Fast Fourier transform (FFT) is a very well-known signal processing technique to convert the time-domain signal into the frequency domain and is utilized to develop several algorithms for fault detection and classification in electric power systems. However, FFT provides accurate assessment in the case of stationary signals only as it requires the signal to be periodic. Real-world PQ disturbances such as faults, sags, swell, oscillatory transients, etc. are non-stationary signals, and FFT results in an inaccurate assessment of the PQ disturbances due to spectral leakage. Also, FFT is unable to provide any time information on the faulty signal and transient frequency components as computation is performed in the frequency domain only. Regarding the limitations of the conventional FFT method, the application of wavelet transform (WT) is motivated by fault detection and classification. Also, localization as it shows suitable time-frequency localization ability for non-stationary signals. Reference employs discrete wavelet transform (DWT) to design fault classification tool for determining the boundary of operating region for series compensated transmission line, discusses DWT-based technique in detail, and demonstrates that DWT is an excellent online tool for relaying applications.

II. PROPOSED MODEL

TFD-BASED FAULT DETECTION METHOD

In this work, a typical model of a 60 Hz, 200 km, 230 kV high voltage (HV) transmission line with two power sources, sending end and receiving end. The complete power system model is simulated in PSCAD/EMTDC to generate different types of faults. Fault resistance of 0.01 ohm is used in the simulation, and the fault is applied 50 km from sending end at t = 0.2 s and cleared at t = 0.25 s. The positive, negative, and zero sequence components of the transmission line impedance. After PSCAD/EMTDC simulation, faulty signals are imported into MATLAB for post-processing, and the TFD-based time-frequency localization technique is used for feature

TABLE I: Sequence Components of 60 Hz, 200 km, 230 kV Transmission Line Impedance

traction and single out faults from other PQ disturbances. L. Cohen first generalized all (TFD) as

Mathematical Analysis-

Cv(t, ν; f) = 1 4π 2 Z Z Z v(u + τ 2 )v ∗ (u − τ 2 ) × f(ξ, τ )e −jξt−jτ ν+jξudξdτ du   (1)

                                                                                                          : where Cv(t, ν; f)

 is the time-frequency distribution (TFD) of the analytic (complex) representation of a signal v(t), and v ∗ (t) is the complex conjugate of v(t)? f(ξ, τ ) is a two-dimensional parameterization function known as the kernel. The variables ξ and τ represent a frequency domain shift and a time-domain shift, respectively. TFD is considered an advanced DSP technique as it overcomes the limitations of conventional DSP methods such as FFT, and has some advanced features. For example, it provides simultaneous time and variable frequency information and has some advanced applications, for example, radar, sonar, seismic, and medical imaging systems in real-time. Note that in signal processing energy of a signal v(t) is defined as |v(t)| 2 . As seen in (1), a product of v(u + τ 2 ) and its complex conjugate v ∗ (u− τ 2 ) provides instantaneous energy information of the signal v(t) which is obtained via a time-marginal property of TFD. This energy information of the signal in the time-frequency domain is utilized to single out faults from other PQ disturbances. However, TFD is a bi-linear transform, and interference or cross terms (unwanted frequency components) are introduced as a result of the by-product. It may cause inaccurate detection of a fault. Among all TFD, reduced interference distribution (RIDB) has shown the most suitable properties to minimize the interference terms.

Principle Average of IFDR Over Half a Cycle.

However, it is noticed that the value of the principal average is not enough for fault classification as the principal average of faulty phase under different types of faults has a similar value. For example, principal average PA and PB for ABG and AB faults are approximately equal which may result in inaccurate classification of faults. In order to obtain a reliable fault classification method under various conditions, fault indicators FIA, FIB, and FIC for phases a, b, and c are introduced here to indicate the principal average value for different types of faults as

ε = FIA + FIB + FIC.

Fault Classification Based on Fault Indicators.

III. LITERATURE SURVEY

The method is developed based on advanced digital signal processing (DSP) and the time-frequency distribution (TFD) technique. Since it shows suitable properties to extract the time-varying signature of non-stationary signals and high-frequency transients introduced by typical power quality (PQ) disturbances in electric power systems. The proposed method first separates the fault disturbance component from the steady-state signal and represents it in the time-frequency domain. Thereby for feature extraction to single out faults from other common electric PQ disturbances such as voltage sags and oscillatory transients. Once a fault is detected, TFD-based new index Instantaneous Fault Disturbance Ratio (IFDR), which provides energy information of fault disturbance compared to steady-state signal, is utilized to classify different types of faults. The analysis results show that the proposed method can classify faults successfully by setting up thresholds obtained with an IFDR index for different types of faults. In this work, different types of fault signals are generated using PSCAD/EMTDC simulation software. Further, fault signal data are imported into MATLAB for post-processing, and time-frequency analysis using Signal Processing Toolbox in MATLAB.  

Conclusion

This simulation project proposes a new method for fault detection and classification in high voltage transmission lines based on advanced digital signal processing techniques and time-frequency analysis. The efficacy of the proposed method is justified by applying it to simulated 230 kV transmission line faults in PSCAD/EMTD simulation software allowed by post-processing the faulty signals in MATLAB Signal Processing Toolbox employing time-frequency distribution (TFD). It is shown that TFD can be used effectively for feature extraction of non-stationary signals in the time-frequency domain, and to single out the fault from other types of PQ disturbances. MATLAB/Simulink-based method and C language method for the implementation of the proposed Relay Model are also discussed along with the DSP techniques which can be applied in IDMT characteristic implementation to improve the performance of the system during false tripping.

References

[1] K. Chen, J. Hu and J. He, “Detection and classification of transmission line faults based on unsupervised feature learning and convolutional sparse autoencoder,” IEEE Trans. Smart Grid, vol. 9, no. 3, pp. 1748- 1758, May 2018 [2] P. Sanjeevikumar, B. Paily, M. Basu and M. Conlon, “Classification of fault analysis of HVDC systems using artificial neural network,” 2014 49th International Universities Power Engineering Conference (UPEC), Cluj-Napoca, 2014, pp. 1-5 [3] S. V. Hareesh, P. Raja and M. P. Selvan, “Design and implementation of a robust fault detection mechanism for EHV lines,” 2016 IEEE 1st International Conference on Power Electronics, Intelligent Control and Energy Systems (ICPEICES), Delhi, 2016, pp. 1-5. [4] S. D ? zakmi ? c, T. Namas and A. Husagi ´ c-Selman, “Combined Fourier ´ transform and Mexican hat wavelet for fault detection in distribution networks,” 2017 9th IEEE-GCC Conference and Exhibition (GCCCE), Manama, 2017, pp. 1-6. [5] A. Jamehbozorg and S. M. Shahrtash, “A decision-tree-based method for fault classification in single-circuit transmission lines,” IEEE Trans. Power Del., vol. 25, no. 4, pp. 2190–2196, Oct. 2010. [6] M. M. Islam, M. R. Hossain, R. A. Dougal, and C. W. Brice, “Analysis of real-world power quality disturbances employing time-frequency distribution,” 2016 Clemson University Power Systems Conference (PSC), Clemson, SC, 2016, pp. 1-5. [7] W.-M. Lin, C.-D. Yang, J.-H. Lin, and M.-T. Tsay, “A fault classification method by RBF neural network with OLS learning procedure,” IEEE Trans. Power Del., vol. 16, pp. 473–476, Oct. 2001.

Copyright

Copyright © 2022 Monali A. Sawale, Dr. Kiran Dongre . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.