The model of the power lines fault location method using time domain reflectometry

The article describes a simulation model of the method, locating power lines fault, using time domain reflectometry. This diagnostic method is a one-side method that can work on both enabled and disabled transmission lines and allows defining the main types of faults, including the high-impedance faults. The developed model consists of two modules: a generation unit implemented in PSCAD/EMTDC and a processing unit implemented in MATLAB. The model is successfully verified and is intended for a comprehensive study of the time domain reflectometry method for power lines diagnostics including analysis for various line topologies using different probing signals and various signal to noise ratio.


Introduction
Timely and accurate power grids diagnostics allows detecting electric line faults at the early stage, preventing the occurrence of major accidents and minimizing the idle time of the power lines [1].
One of the promising methods of power lines diagnostics is the time domain reflectometry method (TDR) based on the time interval measuring between the moments of probe signal sending to the line and the receiving of the reflected signal from the place of the power line fault [2][3][4][5][6]. The diagnostics method of time domain reflectometry is a one-side method, ie, a diagnostics device (locator) is installed on only one side of the line (generally at the beginning). Comparing with other methods of power lines fault location, the described method works on both enadbled and disabled lines, thus allowing an online monitoring and diagnosis of the state of the line.
The use of complex probing signals improves the sensitivity and resolution of the method. A rectangular video pulse can be used as the simpliest probing signal, while more complex signals such as bell-shaped pulses, Barker sequences, chirp signals provide more accuracy and reliability [5].
This article describes the developed model of the power lines diagnostic and fault location method, implemented in PSCAD/EMTDC in conjunction with MATLAB.

The structure of the developed model of the fault location method
The block diagram of the developed power line fault location model using time domain reflectometry is shown in Figure 1.
The simulation model consists of two main units: a reflectogram generation unit and a reflectogram processing unit. The reflectogram generation module is implemented in the PSCAD/EMTDC environment and provides the generation and sending of probing signals to the simulated power lines, as well as the subsequent recording of the resulting trace (reflectogram) and its saving to the file. The reflectogram processing unit is developed in the MATLAB. This module performs processing of the

The reflectogram processing unit
The reflectogram processing unit is implemented in MATLAB. The traces processing task includes differential trace analysis, cross-correlation processing, searching the reflection of the fault and calculating the distance to it. The processing unit also allows studying the power lines diagnostic method for various signal-to-noise-ratios (SNR) by adding noise components to generated reflectogram signals. Additive white Gaussian noise of different power was used in the model. Noise simulation is carried out by mixing additive white Gaussian noise to the resulting reference and actual reflectograms; the noise value is set by a user in decibels relative to the amplitude of the desired signal. After the addition of the noise, the difference trace is calculated by computing the

Verification of the fault location method model
To test the adequacy of the developed model for powerlines diagnostics, firstly impulse responses for the simplest irregularities value with a probe signal generator electric line. Figure 2 shows the simulation results. It can be seen that when the load resistance is less than the wave resistance of the line, the reflected pulse changes its polarity. is larger than the line wave resistance, resistance and wave resistance the reflected pulse is absent. These simulation results are completely consistent with the theory [9]. For further test of the model adequac power lines topologies and the voltage class.
a) The 6-10 kV straight transmission line We will simulate a linear three 14AM) with the length of 40 km. faults are simulated (short circuit faults between lines and between lines and ground, circuit faults) at a distance of 30 km from the beginning of pulse with an amplitude of 1 V and duration of 5 microseconds. line characteristic impedance and connected directly to Figure 3 shows the reference reflecto The probing pulse is sent into the line at time line comes to the beginning of the line probe signal. There was almost changed its shape because of the filtering properties of the line.
difference between the reference and actual traces. When the signal to noise ratio comprising calculation of the cross-correlation function between the sent to the he received echo signal. Extremums of the crossto the front of the reflected probe pulses [8]. Thus, the first extremum of the difference correlation function corresponds to the location of the line failure. Knowing the time of the double path of the probing signal to the fault location, the desired distance is calculated.

fault location method model
To test the adequacy of the developed model for powerlines diagnostics, firstly, we consider obtained impulse responses for the simplest irregularities on the line such as active resistance of value with a probe signal generator (locator) output resistance consistent with wave resistance of Figure 2 shows the simulation results. It can be seen that when the load resistance is less wave resistance of the line, the reflected pulse changes its polarity. When wave resistance, pulse polarity remains the same. In case of resistance and wave resistance the reflected pulse is absent. These simulation results are completely The simulated reflectograms for single wire AC120/19 with the length of 10 km for various ; b) R L = 450 Ω; c) R L = 800 Ω; g) = R L 5000 Ω.
For further test of the model adequacy, we will consider the reflectograms obtained for different voltage class. 10 kV straight transmission line. We will simulate a linear three-phase 6-10 kV power line (conductor AC120/19, tower PS10P 40 km. For more comprehensive testing of the model, short circuit faults between lines and between lines and ground, ) at a distance of 30 km from the beginning of the line. The probe signal amplitude of 1 V and duration of 5 microseconds. The locator is matched to the characteristic impedance and connected directly to the first phase "A".  we consider obtained h as active resistance of the various output resistance consistent with wave resistance of the Figure 2 shows the simulation results. It can be seen that when the load resistance is less When the load resistance In case of equality of the load resistance and wave resistance the reflected pulse is absent. These simulation results are completely Where c -pulse propagation velocity (approximately equal to the speed of light). Thus, the calculated length of the line coincided with the true length (simulated). Figure 4 illustrates the reflectograms of the power line with faults. When simulating a short circuit to ground (Figure 4.a) the reflected pulse changes its polarity. The greatest amplitude of the pulse, reflected from the place of the line fault, is observed when phase "A" is damaged, i.e., the phase connected to the pulse generator. The distance to the fault, calculated by the reflectograms, is equal to 30 km and coincides with a simulated distance. A similar picture of reflected signals is obtained by modeling line-to-line short circuits (Figure 4.b). When simulating breakage of power line wires (Figure 4.c), the reflected pulse retains its polarity. Simulated reflectograms show that the time domain reflectometry method allows locating most of the transmission line faults. b) The 6-10 kV transmission line with one branch Next, we will consider a three-phase 6-10 kV power line with one branch (conductor AC120/19, tower PS10P-14AM). The topology of the simulated power line with the designation of fault places (simulated fault locations marked by red color) is shown in Figure 5. The length of the main (horizontal) segment of the power line is 40 km; the length of the branch line -20 km. Faults were simulated at the points marked with red color at a distance of 5 km from the starting point of the branch. The probe signal -video pulse amplitude of 1 V and a duration of 5 microseconds. The locator, similar to previous studies, is connected directly to phase "A" of the power line and matched to the line impedance. Figure 6 shows the reference reflectogram of the power line without any damage. The reference reflectogram consists of the sent probing pulse (at time t = 0.1 ms), reflection from the branch position (at time t = 0.3 ms), reflections from the ends of the power line and multiple higher-order reflections that attenuate in the amplitude. The amplitude of the pulse reflected from the end of the branch line is significantly smaller than the amplitude of the pulse reflected from the end of the main line segment. This is caused by the fact that at time t = 0.43 ms, two pulses simultaneously come to the signal receiving point (input of the locator) and are added together: the first pulsereflected from the end of the branch line; the second pulse -reflected from a branch place and twice reflected from the end of the main line segment. Since the second pulse has a negative polarity, the amplitude of the resulting signal after the superposition of two pulses will be smaller. The fourth reflected pulse on the reference reflectogram has a maximum amplitude because of the superposition of two positive polarity pulses reflected from both ends of the power line. Further on the reference reflectogram, there are multiple reflections of the probing signal, a detailed examination of which is not of interest for the fault location problem.
Simulation of various faults on the power line with one branch gives results similar to those of the previous case, and agrees with theoretical calculations and experimental data [9,10].

Simulation of the time domain reflectometry method for branched power line with the presence of noise
As an example, let us consider a three-phase power lines of 6-10 kV (wire AC120/19, power line tower PS10P-14AM) with the branched topology, similar to the real power lines (Figure 7).

Conclusion
The developed model of the method consisting of two main blocks -TDR traces of transmission lines with conditions.
The developed model successfully passed verification by comparing experimental results with the theoretical data. The verification as simulating the linear and branched power lines. and three-phase open and short circuits to the ground, were simulated. Testing the developed model showed its adequacy and compliance with the experimental data and theoretical calculations.
The developed model can be used for a range of studies of comparison of the effectiveness of the method for different power lines topologies and various types of accidents, as well as the comparison of different probing signals by the sensitivity and accuracy of the fault location with different signal