A New Wavelet Transform Modulus Maximum Calibration Method Based on Distribution Fault Traveling Wave Location

Abstract

The present invention discloses a novel wavelet transform modulus maximum calibration method based on distribution fault traveling wave location, including: performing wavelet transform decomposition and reconstruction on fault traveling wave data to obtain the waveform of the reconstructed first-layer high-frequency detail coefficients; dividing the time horizontal axis of the waveform of the reconstructed first-layer high-frequency detail coefficients into several equal-length time periods, each time period containing the same number of time points, and sequentially marking the maximum amplitude of the waveform of the reconstructed first-layer high-frequency detail coefficients in each time period. Starting from the amplitude at the 0 time point of the original reconstructed waveform, sequentially connecting the maximum amplitudes of each time period to obtain the upper envelope of the original reconstructed waveform. On this basis, extract all modulus maxima and their corresponding time points in the upper envelope. Finally, set a floating threshold value, compare all modulus maxima with the floating threshold value, and obtain the arrival time of each fault wave head according to the time point corresponding to the modulus maximum, thereby completing the calibration.

Description

Technical Field

[0001] The present invention belongs to the technical field of relay protection for distribution networks, and particularly relates to a new wavelet transform modulus maximum calibration method based on distribution fault traveling wave location. Background Art

[0002] At present, among the methods for locating faults in distribution network lines, the most common one is the traveling wave method. It mainly uses the arrival time of the transient traveling wave generated at the fault point along the distribution network line at the traveling wave speed close to the speed of light to the traveling wave detection device, as well as the traveling wave characteristics and the distribution network topology structure to determine the traveling wave propagation path, so as to accurately locate the fault location. The key factor affecting the location accuracy of the fault traveling wave location method lies in the accuracy of the time calibration of each traveling wave arriving at the measurement point; a calibration error of only 0.1 μs can cause a fault location error of dozens of meters. And wavelet transform can quickly and accurately identify the arrival time of the wavefront.

[0003] When the fault traveling wave arrives at the detection point, the transient component of the traveling wave will undergo irregular mutations at this moment. After the wavelet transform decomposition and reconstruction of the traveling wave signal, the waveform of the high-frequency detail coefficient has a large amplitude near the mutation point of the transient component. Therefore, the arrival time of the initial fault traveling wave and each reflected wave can be determined by calibrating the wavelet transform modulus maximum. As the number of wavelet transform decomposition layers increases and the scale increases, the subtle mutation characteristics of the traveling wave signal representing the arrival time of the traveling wave become less and less obvious. In order to accurately calibrate the subtle mutation points of the arrival time of each traveling wave, usually the waveform of the high-frequency detail coefficient obtained by decomposing and reconstructing the transient traveling wave to the 1st layer is used to calibrate its modulus maximum sequence.

[0004] In the past, people could determine the arrival time of the traveling wave by observing the time corresponding to each surge wave modulus maximum point in the high-frequency coefficients of the wavelet transform. However, in reality, not all modulus maximum points corresponding to the arrival time of the traveling wave can be accurately calibrated.

[0005] Since the distribution network mostly has a complex radial line structure of overhead line-cable hybrid connection with many branches, short sections, and many nodes, the length of most sections is between several hundred meters and two kilometers. In this case, the frequent reflection and refraction of the traveling wave at the impedance discontinuity points easily lead to fast attenuation of the traveling wave, resulting in unclear arrival characteristics of the reflected wave at the fault point and signal aliasing. In addition, the influence of factors such as grounding resistance, fault initial phase angle, and fault distance will also cause unclear singularity characteristics, making it difficult to calibrate the arrival time of the fault traveling wave at the detection point. And there are many detection points in the distribution network, and the calibration workload is huge.

[0006] When the traveling wave is decomposed and reconstructed by wavelet to the high-frequency detail coefficients of the first layer, due to the signal aliasing caused by the short section, the arrival times of the traveling waves corresponding to the modulus maxima of each surge are very close. At the same time, there are a large number of time points with amplitudes lower than those of the surges between these times, which may lead to the situation that the modulus maxima corresponding to the arrival time of the wavefront are missed or mislabeled. Summary of the Invention

[0007] To solve the above technical problems, the present invention proposes a new wavelet transform modulus maxima calibration method based on distribution fault traveling wave location, which further processes the high-frequency detail coefficients obtained by wavelet decomposition and reconstruction of the traveling wave to the first layer, so that the arrival times of each fault traveling wave can be accurately calibrated regardless of fault factors such as different fault distances, fault initial phase angles and grounding resistances, or in the case of frequent reflection and refraction in a distribution network with many short branches, especially the arrival time of the second traveling wave required for single-end location.

[0008] To achieve the above object, the present invention provides a new wavelet transform modulus maxima calibration method based on distribution fault traveling wave location, including the following steps:

[0009] Collect fault traveling wave data, perform wavelet transform decomposition and reconstruction on the fault traveling wave data to the high-frequency detail coefficients of the first layer, and obtain the waveform of the reconstructed high-frequency detail coefficients of the first layer;

[0010] Divide the time horizontal axis of the waveform of the reconstructed high-frequency detail coefficients of the first layer into several equal-length time periods, each time period contains several equal-length time points, and sequentially mark the maximum amplitude of each time period in the waveform of the reconstructed high-frequency detail coefficients of the first layer;

[0011] Starting from the amplitude at the 0 time point of the original reconstructed waveform, connect the maximum amplitudes of each time period in sequence to obtain the upper envelope of the original reconstructed waveform;

[0012] Based on the upper envelope of the original reconstructed waveform, extract all the modulus maxima and the corresponding time points in the upper envelope;

[0013] Set a floating threshold value, extract all the modulus maxima higher than the floating threshold value, and obtain the arrival times of each fault wavefront based on all the modulus maxima higher than the floating threshold value and the corresponding time points, thus completing the calibration of the new wavelet transform modulus maxima.

[0014] Optionally, the number of layers of wavelet transform decomposition is not less than 3.

[0015] Optionally, the method for obtaining several of the time points includes:

[0016] T = s×f = m×n

[0017] Wherein, s is the duration of the collected fault traveling wave, f is the sampling frequency, m is the number of time periods, n is the number of time points included in each time period, and T is the total number of time points of the collected fault traveling wave data.

[0018] Optionally, the relationship between the time points, time periods, and the maximum amplitude marked for each time period includes: the time axis of every n time points is a time period, and a total of m maximum amplitudes are marked.

[0019] Optionally, the method for obtaining the upper envelope of the original reconstructed waveform is as follows:

[0020] d b (t) = [max(d(1), d(2), … d(n)), …, max(d(1+(m - 1)×n), d(2+(m - 1)×n), … d(m×n))] 1×m

[0021] Wherein, d(t) is the original reconstructed waveform function with the time point t of the real-time collected fault traveling wave data as the horizontal axis, m is the number of time periods, n is the number of time points included in each time period, and d b (t) is the upper envelope waveform function composed of m maximum amplitudes with the time point t corresponding to the m maximum amplitudes as the horizontal axis.

[0022] Optionally, the conditions satisfied by all modulus maxima in the upper envelope are:

[0023] d b (i) > d b (i - 1), d b (i) > d b (i + 1), (i = 2, 3, 4, …, m - 1)

[0024] Wherein, m is the number of time periods, i is the serial number of several time periods, i represents the 2nd to the (m - 1)th time period, and d b (t) is the upper envelope waveform function.

[0025] Optionally, a floating threshold value is set, and all modulus maxima higher than the floating threshold value are extracted. The method for obtaining the arrival time of each fault wavefront based on all modulus maxima higher than the floating threshold value and the corresponding time points includes: setting a floating threshold value, comparing all modulus maxima with the floating threshold value in sequence and performing filtering processing, filtering out the modulus maxima lower than the floating threshold value, extracting the modulus maxima higher than the floating threshold value and the corresponding moments, and marking the modulus maxima higher than the floating threshold value and the corresponding moments as the arrival times of each fault wavefront.

[0026] Optionally, the floating threshold value is 0.45% of the maximum amplitude of the original reconstructed waveform.

[0027] Technical effects of the present invention: The present invention discloses a new wavelet transform modulus maximum calibration method based on traveling wave location for distribution network faults. Based on the original wavelet transform modulus maximum calibration method, by adjusting the value of n to process the waveform of the high-frequency detail coefficients reconstructed to the first layer, the interference modulus maximum points with relatively low amplitudes mixed in each surge wavelet in the original reconstructed waveform are processed by drawing an envelope line to filter out the interference values, making the modulus maximum points of each surge wavelet more clearly calibrated. In this way, the arrival times of the first few traveling waves that may be aliased, especially the modulus maximum of the wavelet transform corresponding to the arrival time of the second traveling wave required for single-ended location, can be more clearly and accurately calibrated; this method avoids the problems of traveling wave attenuation, signal aliasing, and unclear singularity characteristics caused by the interference of complex network structures such as many branches and short sections in the distribution network and various fault factors in the original calibration method, and further solves the problem of missing or mislabeling the modulus maximum due to the very close arrival times of the traveling waves corresponding to the modulus maximum of each surge and the small amplitude of the modulus maximum, thereby improving the accuracy and efficiency of traveling wave location for distribution network faults. Description of the Drawings

[0028] The drawings forming a part of this application are used to provide a further understanding of this application. The schematic embodiments of this application and their descriptions are used to explain this application and do not constitute an improper limitation to this application. In the drawings:

[0029] Figure 1 It is a schematic flow chart of the new wavelet transform modulus maximum calibration method based on traveling wave location for distribution network faults in an embodiment of the present invention;

[0030] Figure 2 It is a schematic diagram of the line-mode voltage traveling wave data and its wavelet transform decomposition and reconstruction to the first layer of high-frequency detail coefficients in an embodiment of the present invention, where (a) is the line-mode voltage traveling wave, and (b) is the waveform of the wavelet transform decomposition and reconstruction of the line-mode voltage traveling wave to the first layer of high-frequency detail coefficients;

[0031] Figure 3 It is a schematic diagram of the process of the new wavelet transform modulus maximum calibration in an embodiment of the present invention, where (a) is the original reconstructed waveform and its upper envelope waveform diagram, (b) is the bar chart of the upper envelope, and (c) are the modulus maximum values of the original reconstructed waveform and their corresponding wavefront arrival times. Detailed Embodiments

[0032] It should be noted that, without conflict, the embodiments in this application and the features in the embodiments can be combined with each other. The following will refer to the drawings and combine with the embodiments to detail this application.

[0033] It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer-executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0034] As Figures 1-3 shown, in this embodiment, a new wavelet transform modulus maximum calibration method based on distribution network fault traveling wave positioning is provided, including the following steps:

[0035] Collect fault traveling wave data, perform wavelet transform decomposition and reconstruction on the fault traveling wave data to the first-level high-frequency detail coefficients, and obtain the waveform of the first-level high-frequency detail coefficients after reconstruction;

[0036] Divide the time horizontal axis of the waveform of the first-level high-frequency detail coefficients after reconstruction into several equal-length time periods, each time period contains several equal-length time points, and sequentially mark the maximum amplitude of each time period in the waveform of the first-level high-frequency detail coefficients after reconstruction;

[0037] Taking the amplitude at the 0 time point of the original reconstructed waveform as the starting point, connect the maximum amplitudes of each time period in sequence to obtain the upper envelope of the original reconstructed waveform;

[0038] Based on the upper envelope of the original reconstructed waveform, extract all modulus maxima and corresponding time points in the upper envelope;

[0039] Set a floating threshold value, extract all modulus maxima higher than the floating threshold value, and obtain the arrival time of each fault wave head based on all modulus maxima higher than the floating threshold value and corresponding time points, thus completing the calibration of the new wavelet transform modulus maximum.

[0040] The number of layers of wavelet transform decomposition is not less than 3.

[0041] The time axis of every n time points is a time period, and a total of m maximum amplitudes are marked. The value range of m is related to the sampling frequency and the length of the discrete fault traveling wave data collected. The value range of n determines the accuracy of the traveling wave arrival time calibration, and it should not be too large or too small. If n is too large, it may lead to missing marking of modulus maxima, and if too small, it may lead to incorrect marking of modulus maxima. Through a large number of simulation experiments, it is verified that the time period composed of n time points is 0.5 μs. Then the total number of time points of the collected fault traveling wave data satisfies the following relationship

[0042] T = s × f = m × n

[0043] where s is the duration of the fault traveling wave, f is the sampling frequency, m is the number of time periods, n is the number of time points, and T is the total number of time points of the collected fault traveling wave data.

[0044] In the present invention, N is set to 200, that is, a maximum value point is marked every 5 time points. Further, in step 3, let d(t) be the original reconstructed waveform. Its upper envelope waveform function d b (t) consists of the following m values:

[0045] d b (t) = [max(d(1), d(2), … d(n)), …, max(d(1 + (m - 1)×n), d(2 + (m - 1)×n), … d(m×n))] 1×m

[0046] where d(t) is the original reconstructed waveform function with the time point t of the fault traveling wave data collected in real time as the horizontal axis, m is the number of time periods, n is the number of time points included in each time period, and d b (t) is the upper envelope waveform function composed of m maximum amplitudes with the time point t corresponding to the m maximum amplitudes as the horizontal axis.

[0047] If the amplitude at a certain time point in the upper envelope waveform is greater than the maximum amplitudes of its previous time period and the next time period, the amplitude corresponding to this time point is the preliminary modulus maximum value. Each surge wavelet in the original reconstructed waveform may be mixed with interference modulus maximum value points with relatively low amplitudes. After the upper envelope processing, these interference values have been filtered out, and at the same time, the modulus maximum value points of each surge wavelet are more clearly marked, thus avoiding the missing or mislabeling of the modulus maximum value caused by the interference of complex network structures and external fault factors. The conditions for all modulus maximum values satisfying the upper envelope include:

[0048] d b (i) > d b (i - 1), d b (i) > d b (i + 1), (i = 2, 3, 4, …, m - 1)

[0049] where m is the number of time periods, i is the serial number of several time periods, i represents the 2nd to the (m - 1)th time period, and d b (t) is the upper envelope waveform function.

[0050] Set a floating threshold value, compare all modulus maximum values with the floating threshold value. The method for obtaining the arrival time of each fault wavefront based on the time point corresponding to the modulus maximum value includes: based on the floating threshold value, compare all modulus maximum values with the floating threshold value in turn, filter out the modulus maximum values lower than the floating threshold value, extract the modulus maximum values higher than the floating threshold value and the corresponding moments, and mark the modulus maximum values higher than the floating threshold value and the corresponding moments as the arrival times of each fault wavefront.

[0051] According to Figure 1Perform modulus maximum calibration according to the steps of Figure 2 This is a specific embodiment of the present application. Figure 2 (a) shows the line-mode voltage traveling wave data collected during a grounding fault in a distribution network line where the grounding resistance of the simulation platform is 1000 Ω and the transmission distance of the second reflected wave of the initial traveling wave ratio on the line is less than 0.2×2 km. Figure 2 (b) shows the wavelet transform decomposition and reconstruction of the line-mode voltage traveling wave to the 1st layer of high-frequency detail coefficients. In the figure, the db9 wavelet function is used to decompose and reconstruct the line-mode voltage traveling wave to the 1st layer of high-frequency detail coefficients. The acquisition duration of the fault traveling wave is s = 0.1 ms, and the sampling frequency is f = 10 MHz. Then the total number of time points of the collected fault traveling wave data is T = s×f = 1000. The 1000 time points of the original reconstructed waveform are divided into m time periods with each period length of 0.5 μs, that is, every five time points are in one segment, and a total of 200 segments are divided. Mark the maximum amplitude of the waveform in each time period in turn. Taking the amplitude at the 0 time point of the original reconstructed waveform as the starting point (such as Figure 2 (b) the starting point is d(1) = 6.164×10 -5 ), connect the maximum amplitudes marked in the 1st to the 200th time periods in turn to form the upper envelope waveform of the original reconstructed waveform with these 200 maximum amplitudes as the sequence, as shown in Figure 3 (a). The upper envelope is expressed as follows:

[0052] d b (t) = [max(d(1), d(2), …, d(5)), …, max(d(996), d(997), …, d(1000))] 1×200 That is, d b (t) = [d(1), d(8), … d(35), d(37), d(41), …, d(73) d(77), d(81), d(87), d(91), d(97), …, d(999)] 1×200

[0053] Such as Figure 3 (b) is the bar chart of the upper envelope. By comparing Figure 2 (b), it can be clearly seen that in Figure 2 (b), there are interference modulus maximum points with relatively low amplitudes mixed in each surge wavelet. After the envelope drawing process, as shown in Figure 3 (b), its interference values have been filtered out, and at the same time, the modulus maximum points of each surge wavelet are more clearly calibrated, avoiding the missing or mislabeling of modulus maxima caused by the interference of complex network structures and external fault factors. Therefore, in the upper envelope sequence, extract the preliminary modulus maxima that are larger than the maximum amplitudes of its previous and next time periods and their corresponding time points. The extracted preliminary modulus maxima and the conditions they satisfy are as follows:

[0054] d b (37) = 2.505×10 -6 > d b (35), d b (37) > d b (41); d b (77) = 19.808 > d b (73), d b (77) > d b (81);

[0055] d b (91) = 1.001 > d b (87), d b (91) > d b (97); …d b (983) = 0.056 > d b (977), d b (983) > d b (989);

[0056] Finally, take 0.45% of the maximum amplitude d(77) = 19.81 of the original reconstructed waveform, that is, 0.089 as the floating threshold value, filter out the initial modulus maxima smaller than the threshold value, and further remove the interfering modulus maxima corresponding to other non-fault traveling waves arriving at the detection point. The moments corresponding to each modulus maximum higher than the threshold value are the arrival moments of each fault wavefront. As shown in Figure 3 (c), the filtered modulus maxima and the arrival moment sequences of each fault wavefront corresponding to them are shown as follows:

[0057] d m (t) = [d(77), d(91), d(130), …]

[0058] Calculate the difference in the transmission distance of the initial traveling wave and the second reflected wave on the line:

[0059] L = (91 - 77) × 10 -7 × v = 0.204 × 2 km

[0060] where v = 2.92 × 10 5 km / s, which is the fault traveling wave velocity obtained according to frequency-dependent characteristic parameters, etc. Therefore, the fault distance error calculated by using the new wavelet transform modulus maximum calibration method to calibrate the arrival moment of the fault wavefront in this embodiment is only 4 × 2 m.

[0061] By this method, the influences under different fault distances, fault types and fault resistances were tested. When single-phase grounding faults with different fault distances occurred on cables and overhead lines, the new wavelet transform modulus maximum calibration method was used to calibrate the arrival time t1 of the initial wavefront and the arrival time t2 of the second wavefront, so as to locate the fault distance L at one end. The simulation results are shown in Table 1. The location results for different fault types and different fault resistances when the fault distance is 0.4 km are shown in Tables 1 to 3. Table 1 is the single-phase grounding location result for different fault distances, Table 2 is the location result for different fault types at 0.4 km, and Table 3 is the location result for different fault resistances at 0.4 km.

[0062] Table 1

[0063]

[0064] Table 2

[0065]

[0066] Table 3

[0067]

[0068] As can be seen from Tables 1 to 3, under different fault distances, fault types and fault resistances, this method can accurately calibrate the arrival times of the initial fault traveling wave and the second fault traveling wave through the new wavelet transform modulus maximum calibration method. The error of the obtained location results is basically within 50 meters, with high location accuracy and meeting the actual engineering requirements.

[0069] The above is only a preferred specific embodiment of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed in the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A novel wavelet transform modulus maximum calibration method based on distribution fault traveling wave location, characterized in that Including the following steps: Collect fault traveling wave data, perform wavelet transform decomposition and reconstruction on the fault traveling wave data to the 1st layer of high-frequency detail coefficients, and obtain the waveform of the reconstructed 1st layer of high-frequency detail coefficients; Divide the time horizontal axis of the waveform of the reconstructed 1st layer of high-frequency detail coefficients into several equal-length time periods, each time period contains several equal-length time points, and sequentially mark the maximum amplitude of each time period in the waveform of the reconstructed 1st layer of high-frequency detail coefficients; Taking the amplitude at the 0 time point of the original reconstructed waveform as the starting point, sequentially connect the maximum amplitudes of each time period to obtain the upper envelope of the original reconstructed waveform; The method for obtaining the upper envelope of the original reconstructed waveform is: d b d(t) = [max(d(1), d(2), … d(n)), …, max(d(1+(m-1)×n), d(2+(m-1)×n), … d(m×n))] 1×m Among them, d(t) is the original reconstruction waveform function with the time point t of the real-time collected fault traveling wave data as the horizontal axis, m is the number of time periods, n is the number of time points included in each time period, and d b (t) is the upper envelope waveform function composed of m maximum amplitudes with the time points t corresponding to the m maximum amplitudes as the horizontal axis; Based on the upper envelope of the original reconstructed waveform, extract all modulus maxima and corresponding time points in the upper envelope; The conditions satisfied by all modulus maxima in the upper envelope are: d b (i) > d b (i - 1), d b (i) > d b (i + 1), (i = 2, 3, 4, …, m - 1) Where m is the number of time periods, i is the sequence number of several time periods, i represents the 2nd to the (m - 1)th time period, and d b (t) is the upper envelope waveform function; Set a floating threshold value, extract all modulus maxima higher than the floating threshold value, and obtain the arrival time of each fault wave head based on all modulus maxima higher than the floating threshold value and the corresponding time points, completing the calibration of the new wavelet transform modulus maxima; The method for setting a floating threshold value, extracting all modulus maxima higher than the floating threshold value, and obtaining the arrival time of each fault wave head based on all modulus maxima higher than the floating threshold value and the corresponding time points includes: setting the floating threshold value, sequentially comparing all modulus maxima with the floating threshold value and performing filtering processing, filtering out modulus maxima lower than the floating threshold value, extracting modulus maxima higher than the floating threshold value and the corresponding moments, and calibrating the modulus maxima higher than the floating threshold value and the corresponding moments as the arrival time of each fault wave head; the floating threshold value is 0.45% of the maximum amplitude of the original reconstructed waveform.

2. The novel wavelet transform modulus maximum calibration method based on distribution fault traveling wave positioning according to claim 1, wherein The number of layers of the wavelet transform decomposition is not less than 3.

3. The novel wavelet transform modulus maximum calibration method based on distribution fault traveling wave positioning as described in claim 1, wherein The method for obtaining several of the time points includes: T = s×f = m×n Wherein, s is the duration of the collected fault traveling wave, f is the sampling frequency, m is the number of time periods, n is the number of time points included in each time period, and T is the total number of time points of the collected fault traveling wave data.

4. The novel wavelet transform modulus maximum calibration method based on distribution fault traveling wave positioning according to claim 3, characterized in that The relationship among the time points, time periods, and the maximum amplitude marked for each time period is: the time axis of every n time points is one time period, and a total of m maximum amplitudes are marked.

Images