Dual - terminal traveling wave ranging wavefront calibration method, device and medium applicable to distribution network

Abstract

A method, device and medium for calibrating the wave head of double - ended traveling - wave ranging applicable to distribution networks according to the present invention select appropriate high - frequency components of wavelets in the characteristic frequency band, and calibrate appropriate modulus extreme points as the preliminary estimation of the wave - head starting point; construct a smooth threshold based on the preliminary estimation and the hard - trigger point on the low - frequency component corresponding to the characteristic frequency band; search for a more accurate wave - head starting point through this adaptive smooth threshold; and finally calculate the fault point location according to the double - ended traveling - wave ranging formula. The present invention proposes a new principle and method for calibrating the wave - head starting point by calibrating the waveform inflection point of the wave head and fully utilizing the high - and low - frequency components of appropriate scales in the wavelet domain to complete coordinated punctuation. Practice has proved that this new calibration principle and method are not easily affected by the traveling - wave dispersion and the attenuation, distortion and noise interference of traveling waves caused by wave impedance, and at the same time can overcome the problem of low calibration accuracy caused by the inconsistency of the characteristic spectra of double - ended traveling - wave heads.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution networks, and particularly relates to a double-ended traveling wave ranging wave head calibration method, device and medium applicable to distribution networks. Background Art

[0002] Traveling wave ranging technology and devices were first applied to transmission lines. They have high measurement accuracy, rapid response, and are not affected by factors such as asymmetrical line structures and transducer transformation errors. In the currently maturely applied wavelet transform modulus maximum principle in high-voltage power grids, the modulus maximum points of wavelet components correspond to the points where the component waveforms change fastest, and such points are located at a certain point on the rising or falling edge of the wave head. Since the rising or falling edge of the traveling wave head in the transmission network has a high steepness, directly calibrating the modulus maximum point as the starting point of the wave head can also obtain a good ranging effect;

[0003] Compared with high-voltage lines, the working conditions faced by traveling wave ranging applications in distribution networks are more complex. In medium- and low-voltage distribution systems, the neutral points mostly adopt non-effective grounding methods, and at the same time, mostly adopt radial network structures, mainly characterized by many branch feeders and a mixture of overhead lines / cable lines. When a grounding or short-circuit fault occurs, the neutral grounding method, grounding resistance, line parameters, topological structure, distributed parameters, and the number of shunt branches directly affect the shape, amplitude, and steepness of the fault traveling wave head;

[0004] At the same time, in distribution network lines, factors such as a large amount of random noise in the working environment of on-site traveling wave ranging devices, the propagation path and length of traveling waves, the magnitude of the transition resistance, and the phase angle of the fault voltage directly affect the shape characteristics and steepness characteristics of the wave head, causing the wave head to be distorted and become flatter. For this, it is difficult for the wavelet modulus maximum principle to effectively and accurately calibrate the starting point of the wave head. Therefore, further wave head starting point calibration principles and methods are required in distribution network traveling wave ranging. Summary of the Invention

[0005] A double-ended traveling wave ranging wave head calibration method, device and storage medium applicable to distribution networks proposed by the present invention can at least solve one of the technical problems in the background art.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] A double-ended traveling wave ranging wave head calibration method applicable to distribution networks includes the following steps,

[0008] Step 1: Extraction of fault traveling waves: Collect a three-phase fault current traveling wave sampling sequence of 1000 points, namely Ia, Ib, and Ic, from 200 points before to 800 points after fault triggering at both the M and N ends of the distribution network line through a current transformer. The sampling rate is selected as 1Mhz, and the sampling time is one millisecond. A high-pass filter device is included in the hardware sampling to remove power frequency and related harmonic signal components below 10khz;

[0009] Step 2: Construction of modulus fault components: Perform a phase-mode transformation on the three-phase fault current traveling wave sequence after sampling and filtering: (I0, I1, I2) = A(Ia, Ib, Ic)’, where A is the phase-mode transformation matrix, I0 is the zero-mode component, and I1, I2 are the line-mode components; Select the line-mode component I1 as the fault modulus vector to be processed, and the fault traveling wave moduli at both the M and N ends are denoted as: IM, IN;

[0010] Step 3: Wavelet decomposition: Decompose the fault traveling wave moduli IM and IN at both the M and N ends into multi-scale high and low frequency components through wavelet transform, and form the modulus matrices mod_M and mod_N with the multi-scale high and low frequency components. Each row of the matrix corresponds to one scale;

[0011] Step 4: Modulus extreme value screening: Based on the modulus matrices at both the M and N ends, set the characteristic frequency band as 10khz~200khz, and screen the modulus extreme points that meet the requirements as the preliminary estimates of the starting points of the fault traveling wave heads reaching both ends, denoted as MaxModPoint_M and MaxModPoint_N respectively;

[0012] Step 5: Further estimation of the wave head starting point: On the basis of the preliminary estimation of the wave head starting point, combine the hardware wave head trigger point hardStarPoint (known during traveling wave sampling) to construct a smoothing threshold, and obtain more accurate estimates of the wave head starting points headStartPoint_M and headStartPoint_N through threshold screening;

[0013] Step 6: Fault distance calculation: Calculate the time difference between the arrival of the fault traveling wave heads at both ends based on the point difference between the accurate wave head starting point estimates headStartPoint_M and headStartPoint_N at both the M and N ends and the waveform recording trigger times at both ends, and then obtain the distances from the fault point to the M end and N end through the ranging formula as:

[0014]

[0015]

[0016] Among them, L is the line distance between the M end and the N end, and v is the fault traveling wave velocity.

[0017] Furthermore, in the selection of the transformation matrix in step 2, on the premise that the fault phase has been determined, the commonly used phase-mode transformation includes the Karenbauer transformation matrix A = (1, 1, 1; 1, -1, 0; 1, 0, -1). In the case where the fault phase is uncertain, the improved phase-mode transformation matrix A = (1, 1, 1; 1, -5, 4; 1, 4, -5) is adopted.

[0018] Furthermore, in the wavelet decomposition process of step 3, for the fault modulus S(n) (representing the fault modulus to be processed: IM or IN), the following discrete dyadic spline wavelet transform is adopted:

[0019]

[0020]

[0021] Here, n represents the serial number, is the approximation coefficient vector of the j-th scale, is the wavelet coefficient vector of the j-th scale, is the convolution symbol, and are the filter bank vectors derived from the spline wavelet (for example, the vectors h and g can be taken as [-3 / 2, 3 / 2] and [1 / 8, 3 / 8, 3 / 8, 1 / 8]), and are the high-pass and low-pass filter bank vectors at the j - 1 scale respectively, formed by and gradually inserting zeros; is the compensation factor used to overcome the influence brought by the scale; the specific implementation of the wavelet decomposition is completed based on the stationary spline wavelet transform of the à Trous framework, and the specific signal space decomposition is represented by the following schematic formula:

[0022]

[0023]

[0024]

[0025]

[0026]

[0027] In the above schematic formula, the fault modulus S(n) can be equivalently regarded as the coefficient representation in the full-space basis. In the subsequent equivalent relationships, the full space is gradually decomposed to form wavelet subspaces of different scales. and It can be understood as the coefficient representation of the basis of the corresponding subspace, where represents the direct sum of independent subspaces, represents the equivalent relationship of the sum of different subspaces, ( , ) represents the paired high and low frequency coefficient combinations of the subspace at a certain scale.

[0028] Furthermore, the definition of the modulus of the j-scale wavelet decomposition component is as follows:

[0029]

[0030] where, represents a modulus extreme value in the modulus of the j-scale wavelet decomposition component, is the maximum value of the absolute value of the wavelet component at the j-th scale. The specific implementation of the modulus of the wavelet component coefficient is to obtain local extreme points and extremes through the first and second order differences of the wavelet component, and the amplitude of non-extreme points can be directly set to zero.

[0031] Furthermore, for the modulus extreme value screening process in step 4, the screening scheme starts from the scale corresponding to the characteristic frequency and traces back in reverse order to the cut-off scale j = 2.

[0032] Furthermore, the specific screening steps in step 4 are as follows:

[0033] (1) Calculate the scale corresponding to the target frequency band: Based on the sampling frequency fs and the median Target_fs of the target frequency band, the scale corresponding to the target frequency band is obtained by the formula r = floor{1 + log( fs / 2 / Target_fs ) / log(2.0)}, where floor{} represents the operation of rounding up the variable in the brackets, and the processing object is the modulus vector of the high-frequency component coefficients of each scale in the modulus matrix ;

[0034] (2) Start searching from the high-frequency component of the R-th scale, R = r+1, with Rat max_R as the threshold, where the proportionality coefficient Rat = 0.35 and max_R is the maximum value of the absolute value of the amplitude of the high-frequency component of this scale. Sequentially search to obtain the point number corresponding to the first maximum or minimum value whose absolute value of the high-frequency component element of this scale is greater than the threshold, and this point number is used as the modulus extreme point of this scale, denoted as MaxModPoint_R;

[0035] (3) At the previous scale R-1, within the range of plus or minus 2 of the point MaxModPoint_R, sequentially search for the element whose absolute value crosses the threshold Rat For the first modulus extreme value with the same sign as MaxModPoint_R and the corresponding point of max_R - 1, record this point as the modulus extreme point of this scale, MaxModPoint_(R - 1). If the search is empty, then record MaxModPoint_(R - 1)=MaxModPoint_R;

[0036] (4) Then enter the next higher scale, repeat the search mechanisms in (2) and (3), stop the search at the second scale, and record the overall modulus extreme point as MaxModPoint = MaxModPoint_2;

[0037] (5) Record the modulus extreme points at both ends as MaxModPoint_M and MaxModPoint_N respectively.

[0038] Further, for step 5 to obtain a more accurate estimate of the wavefront starting point, there are the following implementation steps:

[0039] For obtaining a more accurate estimate of the wavefront starting point at the M - end, based on the selected modulus extreme point MaxModPoint = MaxModPoint_M, determine the pre - wave point beforePoint = min(MaxModPoint / 2, hardStarPoint - 50) with reference to the hardware starting point. The processing object is the low - frequency component coefficient vector of the modulus matrix corresponding to the characteristic frequency band; (2) Determine the smoothing threshold smoothThd based on the waveform before the wavefront point, set the smoothing starting point startPoint = 50, and set the smoothing bandwidth smoothWidth = 20;

[0040] (2.1) For each point from the starting point startPoint to beforePoint, calculate the difference between the amplitude of this point and the average amplitude of the points within smoothWidth before this point, record the corresponding absolute value of the difference, and form a sequence;

[0041] (2.2) Statistically find the maximum value maxValue of this difference sequence, perform appropriate relaxation, and the relaxation coefficient can be selected as 1.2, generating the threshold smoothThd = 1.2 maxValue;

[0042] (3) Based on the threshold smoothThd, within the range of points from (beforePoint, MaxModPoint), compare the amplitudes point - by - point in reverse order, and obtain the first point that exceeds the threshold and record it as a more accurate estimate of the wavefront starting point headStartPoint_M;

[0043] (4) Repeat the above process for the other end to obtain a more accurate estimation of the wavehead starting point headStartPoint_N of the fault traveling wave at this end.

[0044] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which when executed by a processor causes the processor to execute the steps of the above method.

[0045] In still another aspect, the present invention also discloses a computer device including a memory and a processor, where the memory stores a computer program, and when the computer program is executed by the processor, it causes the processor to execute the steps of the above method.

[0046] As can be seen from the above technical solutions, based on the research of wavelet transform modulus extreme value detection theory and test practice, the present invention analyzes the essential process of wavehead detection, and proposes a new type of double-end traveling wave wavehead calibration and ranging method. Specifically, the present invention provides a double-end fault traveling wave ranging method applicable to distribution networks. First, by selecting appropriate high-frequency wavelet components in the characteristic frequency band (10 kHz - 200 kHz), calibrate appropriate modulus extreme points as the preliminary estimation of the wavehead starting point; secondly, based on the preliminary estimation and hard trigger points on the low-frequency component corresponding to the characteristic frequency band, construct a smoothing threshold; furthermore, search for a more accurate wavehead starting point through this adaptive smoothing threshold; finally, calculate the fault point location according to the double-end traveling wave ranging formula. Different from the conventional calibration process in the traditional wavelet transform method where the modulus maximum point is directly calibrated as the wavehead starting point, the present invention proposes a calibration principle with the wavehead waveform inflection point as the wavehead starting point and makes full use of appropriate high and low frequency components in the wavelet domain to coordinately and gradually complete the punctuation and ranging process.

[0047] Compared with the prior art, the present invention has the following advantages:

[0048] The present invention proposes a calibration principle with the wavehead waveform inflection point that is not easily interfered as the wavehead starting point, and a step-by-step progressive screening and calibration method based on modulus maximum screening and adaptive threshold clamping points, which can obtain a punctuation closer to the true wavehead starting point under the condition of a small signal-to-noise ratio of the traveling wave signal, achieving better ranging effects and accuracy. At the same time, the algorithm overcomes the problems that the conventional algorithm cannot recognize or has large errors in the case of wavehead distortion and gentle steepness, and has good ranging effects and strong applicability in the traveling wave measurement of distribution networks.

[0049] At the same time, in the implementation scheme of the present invention, based on the distribution characteristics of the distribution network, the overall structure is organized based on the principle of distributed acquisition and master station calculation, automatically acquires three-phase current traveling waves and uploads waveform data, and obtains accurate ranging effects through master station analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1Schematic diagram of the calculation method for traveling wave ranging disclosed by the present invention;

[0051] Figure 2 Topological structure diagram for implementing traveling wave ranging experiments and tests in a distribution network;

[0052] Figure 3 Wavelet high-frequency component and corresponding modulus diagram of the fault traveling wave modulus signal at the M end;

[0053] Figure 4 Wavelet low-frequency component and corresponding modulus diagram of the fault traveling wave modulus signal at the M end;

[0054] Figure 5 Wavelet high-frequency component and corresponding modulus diagram of the fault traveling wave modulus signal at the N end;

[0055] Figure 6 Wavelet low-frequency component and corresponding modulus diagram of the fault traveling wave modulus signal at the N end;

[0056] Figure 7 Schematic diagram for calibrating the upper modulus maximum value and smoothing threshold of the second-order low-frequency component of the double-end fault traveling wave modulus signal;

[0057] Figure 8 Schematic diagram of the final calibration points of the double-end wavefronts on the original fault traveling wave modulus waveform. Detailed implementation manners

[0058] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention.

[0059] Based on the characteristics of the gentle wavefront in the distribution network analyzed above, in the traditional wavelet transform method, there are inevitable theoretical errors in calibrating the modulus maximum point (the point with the largest amplitude change) as the starting point of the wavefront in the distribution network. Therefore, the present invention proposes a calibration principle using the waveform inflection point before the modulus maximum point as the starting point of the wavefront, and further forms a double-end traveling wave ranging method suitable for the distribution network.

[0060] The embodiments of the present invention specifically adopt the following technical solutions:

[0061] An improved double-end traveling wavefront calibration and ranging method for a distribution network, the method comprising the following steps:

[0062] Step 1: Extraction of fault traveling waves: Collect the three-phase fault current traveling wave sampling sequences of 1000 points before 200 points and after 800 points of fault triggering at both the M end and N end of the distribution network line through a current transformer: Ia, Ib, Ic. The sampling rate is selected as 1Mhz, the sampling time is one millisecond, and a high-pass filter device is included in the hardware sampling to remove the power frequency and related harmonic signal components below 10khz;

[0063] Step 2: Perform phase-mode transformation on the three-phase fault current traveling wave sequences sampled and filtered in S1: (I0, I1, I2) = A(Ia, Ib, Ic)’, where A is the phase-mode transformation matrix, I0 is the zero-mode component, and I1, I2 are the line-mode components; Select the line-mode component I1 as the fault modulus vector to be processed, and the fault traveling wave moduli at both the M end and N end are denoted as: IM, IN;

[0064] Step 3: Wavelet decomposition: Decompose the fault traveling wave moduli IM and IN at both the M end and N end into multi-scale high and low frequency components through wavelet transform, and form the component into a modulus matrix mod_M, mod_N. Each row of the matrix corresponds to a scale;

[0065] Step 4: Modulus extreme value screening: Based on the modulus matrices at both the M end and N end, set the characteristic frequency band as 10khz~200khz, and screen the modulus extreme points that meet the requirements as the preliminary estimates of the starting points of the fault traveling wave heads reaching both ends, denoted as MaxModPoint_M and MaxModPoint_N respectively;

[0066] Step 5: Further estimation of the wave head starting point: On the basis of the preliminary estimation of the wave head starting point, combine the hardware wave head trigger point hardStarPoint (known during traveling wave sampling) to construct a smoothing threshold, and obtain more accurate wave head starting point estimates headStartPoint_M and headStartPoint_N based on threshold screening;

[0067] Step 6: Fault distance calculation: According to the point difference between the accurate wave head starting point estimates headStartPoint_M and headStartPoint_N at both the M end and N end and the waveform recording trigger moments at both ends to calculate the time difference between the arrival of the fault traveling wave head at both ends , and then obtain the distances from the fault point to the M end and N end through the ranging formula:

[0068]

[0069]

[0070] where L is the line distance between the M end and N end, and v is the traveling wave velocity.

[0071] Specifically, in the selection of the transformation matrix in Step 2, on the premise that the faulty phase has been determined, a commonly used phase-mode transformation such as the Karenbauer transformation matrix A = (1, 1, 1; 1, -1, 0; 1, 0, -1) is selected. In the case where the faulty phase is uncertain, the improved phase-mode transformation matrix A = (1, 1, 1; 1, -5, 4; 1, 4, -5) is adopted.

[0072] Specifically, in the wavelet decomposition process of Step 3, for the fault modulus S(n) (representing the fault modulus to be processed: IM or IN), the following discrete dyadic spline wavelet transform is adopted:

[0073]

[0074]

[0075] Here, n represents the serial number, is the approximation coefficient vector of the j-th scale, is the wavelet coefficient vector of the j-th scale, is the convolution symbol, and come from the filter bank vectors of the spline wavelet (for example, the vectors h and g can be taken as [-3 / 2, 3 / 2] and [1 / 8, 3 / 8, 3 / 8, 1 / 8]), and are the high-pass and low-pass filter bank vectors at the (j - 1)-th scale respectively, formed by and gradually inserting zeros; is the compensation factor (compensating for the error caused by the discrete reduction of the scale to make the coefficients more accurate); the specific implementation is completed based on the stationary spline wavelet transform of the à Trous framework, and the specific signal space decomposition is represented by the following schematic formula:

[0076]

[0077]

[0078]

[0079]

[0080]

[0081] In the above schematic formula, the fault modulus S(n) can be equivalently regarded as the coefficient representation in the full-space basis. In the subsequent equivalent relationship, the full space is gradually decomposed to form wavelet subspaces of different scales. and can be understood as the coefficient representations of the corresponding subspace bases, where represents the direct sum of independent subspaces, represents the equivalence relation of the sum of different subspaces, ( , ) represents the paired high and low frequency coefficient combinations of subspaces at a certain scale.

[0082] The definition of the modulus of the further wavelet decomposition components at scale j is as follows:

[0083]

[0084] where, represents a modulus extremum in the modulus of the wavelet decomposition components at scale j, is the maximum value of the absolute value of the wavelet components at the j-th scale. The specific implementation of the modulus of the wavelet component coefficients is to obtain local extreme points and extrema through the first and second order differences of the wavelet components, and the amplitude of non-extreme points can be directly set to zero.

[0085] Specifically, for the modulus extremum screening process in step 4, since the wavelet modulus extrema at low scales are easily affected by high-frequency noise, and the wavelet modulus extrema at high scales correspond to lower frequency bands and cannot reflect the high-frequency characteristics of transient traveling waves, the screening scheme starts from the scale corresponding to the characteristic frequency and backtracks in reverse order to the cut-off scale j = 2. Of course, this scheme can also be set to other scales according to the actual situation (for example, j = 1, 3, 4).

[0086] The specific screening steps are as follows:

[0087] (1) Calculate the scale corresponding to the target frequency band: Based on the sampling frequency fs and the median Target_fs of the target frequency band, the scale corresponding to the target frequency band is obtained by the formula r = floor{1 + log( fs / 2 / Target_fs ) / log(2.0)}, where floor{} represents the operation of rounding up the variable in the brackets, and the processing object is the modulus vector of the high-frequency component coefficients of each scale in the modulus matrix ;

[0088] (2) Starting from the high-frequency components at the R-th scale, search sequentially element by element, where R = r + 1, with Rat max_R as the threshold, the proportionality coefficient Rat = 0.35, and max_R is the maximum value of the absolute value of the amplitude of the high-frequency components at this scale. Search sequentially to obtain the point number corresponding to the first maximum or minimum value whose absolute value of the high-frequency component element is greater than the threshold. This point number is used as the modulus extremum point at this scale, denoted as MaxModPoint_R;

[0089] (3) At the previous scale R - 1, within the range of plus or minus 2 of the point MaxModPoint_R, search sequentially for elements whose absolute value crosses the threshold Rat The first modulus extreme value with the same sign as MaxModPoint_R and corresponding point of max_R - 1, record this point as the modulus extreme point of this scale: MaxModPoint_(R - 1). If the search is empty, record MaxModPoint_(R - 1)=MaxModPoint_R;

[0090] (4)Then enter the previous scale, repeat the search mechanisms in (2) and (3), stop the search at the second scale, and record the overall modulus extreme point as MaxModPoint = MaxModPoint_2.

[0091] (5)Record the modulus extreme points at both ends as MaxModPoint_M and MaxModPoint_N respectively.

[0092] Specifically, for step 5 to obtain a more accurate estimation of the wavefront starting point, there are the following implementation steps:

[0093] (1)For terminal M, to obtain a more accurate estimation of the wavefront starting point, based on the selected modulus extreme point MaxModPoint = MaxModPoint_M, determine the point before the wavefront beforePoint = min(MaxModPoint / 2, hardStarPoint - 50) with reference to the hardware starting point. The processing object is the coefficient vector of the low - frequency component of the modulus matrix corresponding to the characteristic frequency band;

[0094] (2)Determine the smoothing threshold smoothThd based on the waveform before the wavefront point, set the smoothing starting point startPoint = 50, and set the smoothing bandwidth smoothWidth = 20;

[0095] (2.1)For each point from the starting point startPoint to beforePoint, calculate the difference between the amplitude of this point and the average amplitude of the points within smoothWidth before this point, record the corresponding absolute value of the difference, and form a difference sequence;

[0096] (2.2)Statistical maximum value of this difference sequence is maxValue, and appropriately relax (the relaxation coefficient is selected as 1.2) to generate the threshold smoothThd = 1.2 maxValue;

[0097] (3)Based on the threshold smoothThd, within the point range of (beforePoint, modmaxPoint), compare the amplitudes of points in reverse order one by one, and obtain the first point that exceeds the threshold and record it as a more accurate estimation of the wavefront starting point (headStartPoint_M).

[0098] (4) Repeat the above process for the other end to obtain a more accurate estimation of the wavefront starting point (headStartPoint_N) of the fault traveling wave at this end.

[0099] The technical solution of the present invention will be further described in detail below in conjunction with the accompanying drawings of the specification and specific embodiments.

[0100] As Figure 1 shown is the overall process framework of this embodiment. The test topology is composed of Figure 2 shown. The main test line is the line with a length of 5 km from DTU-82 to DTU_83. A double-ended traveling wave signal for determining the fault distance is generated by a traveling wave signal generator, appropriate white noise is added, and then the three-phase fault traveling wave waveform data in the double-ended comtrade standard format is obtained through the hardware filtering circuit and the current transformer acquisition channel, and is transmitted to the master station for processing; the processing methods used include the following steps:

[0101] Step (1): The sampling rate of the hardware circuit is 1 MHz, the interval between two sampling points is 1 microsecond, the sampling time is 1 millisecond, the number of sampling points is 1000, and the three-phase phase currents at both ends are collected: Ia, Ib, Ic.

[0102] Step (2): The double-ended three-phase fault currents obtained in step (1) are subjected to phase-mode transformation (Keren Bell transformation) to obtain the fault traveling wave line-mode current components at both ends of MN (MN corresponds to DTU-82 and DTU_83 respectively) as the processing objects, and the data is as Figure 8 shown.

[0103] Step 3: Perform wavelet transform on the double-ended fault traveling wave line-mode current components, and perform modulus extremum screening on the high-frequency and low-frequency component coefficient vectors of the second scale respectively. The results are 500 and 497. Further, through smoothing threshold punctuation, the calibration results of the double-ended traveling wave wavefront starting points are obtained as 498 and 494; among them, the double-ended wavelet decomposition components and modulus waveforms are as Figure 3 , 4 and 5, 6 shown; the modulus extremum screening and threshold calibration results on the low-frequency component of the second scale are as Figure 7 shown; the illustration of the final wavefront punctuation on the original signal is as Figure 8 .

[0104] Step 4: Set the fault traveling wave velocity as V = 2.93e8 m / s according to the line material, and calculate the fault distances as L_M = 4032 m and L_N = 968 m, with an error of 32 m.

[0105] As described in the above steps, in the modulus extreme value screening and the smoothing threshold clamping point, there are certain differences between the appropriate modulus extreme value points and the threshold clamping points. However, the threshold clamping point is closer to the true wavefront starting point, and the ranging results further demonstrate this conclusion. At the same time, this case shows that the modulus extreme value point only corresponds to the point where the wavefront changes fastest, and the wavefront inflection point is closer to the true wavefront starting point.

[0106] Compared with the prior art, the present invention has the following advantages:

[0107] The overall structure of the present invention is organized based on the principles of distributed acquisition and master station calculation, which is suitable for the distribution characteristics of the distribution network. It automatically acquires three-phase current traveling waves and uploads waveform data, and obtains accurate ranging results through master station analysis;

[0108] The present invention proposes a step-by-step progressive estimation and calibration principle and screening method for the wavefront starting point of fault traveling waves based on modulus maxima and adaptive thresholds, which can obtain punctuation points closer to the true wavefront starting point when the signal-to-noise ratio of the traveling wave signal is small, achieving better ranging results and accuracy. At the same time, the algorithm overcomes the problems of inability to identify or large errors in the case of wavefront distortion and gentle steepness, and has strong applicability in the traveling wave ranging of the distribution network.

[0109] On the other hand, the present invention also discloses a computer-readable storage medium storing a computer program, which when executed by a processor causes the processor to execute the steps of the above method.

[0110] On yet another aspect, the present invention also discloses a computer device including a memory and a processor, where the memory stores a computer program, and when the computer program is executed by the processor, it causes the processor to execute the steps of the above method.

[0111] In another embodiment provided by the present application, there is also provided a computer program product containing instructions, which when running on a computer causes the computer to execute any of the double-ended traveling wave ranging wavefront calibration methods applicable to the distribution network in the above embodiments.

[0112] It can be understood that the systems, devices, and storage media provided by the embodiments of the present invention correspond to the methods provided by the embodiments of the present invention, and the explanations, examples, and beneficial effects of the relevant content can refer to the corresponding parts in the above methods.

[0113] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer, or a data storage device such as a server or data center that includes one or more integrated available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid state disk (SSD)).

[0114] It should be noted that in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise", or any other variant thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device that includes a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or device. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, method, article, or device that includes the element.

[0115] Each embodiment in this specification is described in a related manner. The same or similar parts among the embodiments can be referred to each other, and the differences between each embodiment and other embodiments are emphasized. In particular, for the system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the description of the method embodiments.

[0116] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A double-terminal traveling wave ranging wave head calibration method suitable for distribution network, characterized in that: The following steps are included: S1: The three-phase fault current traveling wave sampling sequence is collected through the mutual inductor, i.e., the first 200 points and the last 800 points of the fault traveling wave at both ends of the line M and N, i.e., a total of 1000 points: Ia, Ib, Ic, with a sampling rate of 1Mhz; S2: Perform phase mode transformation on the sampling sequence in S1: (I0, I1, I2) = A(Ia, Ib, Ic)', where A is the phase mode transformation matrix, I0 is the zero mode component, and I1 and I2 are line mode components respectively; Select I1 as the fault modulus, and the fault moduli at both ends of M and N are recorded as: IM, IN; S3: Decompose IM and IN into multi-scale high and low frequency components through wavelet transform and form the modulus matrices at both ends of M and N respectively; S4: Based on the modulus matrix at both ends of M and N, the characteristic frequency band is set to 10khz~200khz, and the respective modulus extreme points are selected as the preliminary estimation of the starting points of the wave heads at both ends; S5: Based on the preliminary estimation of the wave head starting points at both ends, a smoothing threshold is constructed and a more accurate wave head starting point estimation is screened; S6: The time difference of the fault wave head reaching both ends is calculated by the difference between the more accurate wave head starting point estimation at both ends of M and N and the waveform recording triggering time at both ends of M and N. , the distances from the fault point to the M and N terminals are obtained by the distance measurement formula: L is the line distance between M and N, v is the speed of the fault traveling wave; For S5, constructing a smoothing threshold and screening a more accurate wave head starting point estimate includes the following steps: (1) Obtain a more accurate estimation of the wave head starting point for the M end. Based on the selected modulus extreme point MaxModPoint = MaxModPoint_M, determine the wave head front point beforePoint = min(MaxModPoint / 2, hardStarPoint-50) with reference to the hardware starting point. The processing object is the coefficient vector of the low-frequency component of the modulus matrix corresponding to the scale of the characteristic frequency band; (2) Determine the smoothing threshold smoothThd based on the waveform before the wavefront point, set the smoothing start point startPoint=50, and set the smoothing bandwidth smoothWidth=20; (2.1) For each point from the starting point startPoint to beforePoint, calculate the difference between the amplitude of the point and the average amplitude of the points in smoothWidth before this point, record the corresponding absolute value of the difference, and form a difference sequence; (2.2) The maximum value of this difference sequence is counted as maxValue, and relaxation is performed. The relaxation coefficient is selected as 1.2, and the threshold smoothThd = 1.2 is generated. maxValue; (3) Based on the threshold smoothThd, within the point range (beforePoint, MaxModPoint), compare the amplitudes point by point in reverse order, and obtain the first point that exceeds the threshold as a more accurate estimate of the wave head starting point headStartPoint_M; (4) Repeat the above process for the other end to obtain a more accurate estimate of the wave head starting point headStartPoint_N of the fault traveling wave at that end.

2. The double-terminal traveling wave ranging wave head calibration method applicable to the distribution network according to claim 1 is characterized in that: For the selection of transformation matrix in phase mode transformation in S2, under the premise that the fault phase has been determined, the commonly used phase mode transformation includes Karen Bell transformation matrix A=(1,1,1; 1,-1,0; 1,0,-1). In the case of uncertain fault phase, the improved phase mode transformation matrix A=(1,1,1; 1,-5,4; 1, 4,-5) is used.

3. The double-terminal traveling wave ranging wave head calibration method applicable to the distribution network according to claim 1 is characterized in that: In the wavelet transform process of S3, for the fault modulus S(n), S(n) represents the fault modulus to be processed: IM or IN, the following discrete binary spline wavelet transform is used: Here n represents the sequence number, is the approximate coefficient vector of the jth scale, is the wavelet coefficient vector of the jth scale, is the convolution symbol, and The filter bank vector derived from the spline wavelet, and They are respectively the high-pass and low-pass filter group vectors at the j-1 scale, and are and By gradually inserting zeros at alternate points, is the compensation factor; The specific implementation of wavelet decomposition is based on the stationary spline wavelet transform of the à Trous framework. The specific signal space decomposition is expressed by the following schematic formula: The fault modulus S(n) in the above schematic formula can be equivalently regarded as the coefficient representation in the whole space basis. In the subsequent equivalent relationship, the whole space is gradually decomposed to form wavelet subspaces of different scales. and is the coefficient representation of the corresponding subspace basis, where represents the direct sum of independent subspaces, Represents the equivalence relation between different subspaces and, ( , ) represents the paired combination of high and low frequency coefficients of the subspace at a certain scale.

4. The double-terminal traveling wave ranging wave head calibration method applicable to the distribution network according to claim 3 is characterized in that: The modulus of the j-scale wavelet decomposition component is defined as follows: in, Represents a modulus extreme value in the modulus of the j-scale wavelet decomposition component, is the maximum absolute value of the wavelet component at the jth scale. The specific implementation of the wavelet component coefficient modulus is to obtain local extreme points and extreme values ​​through the first and second order differences of the wavelet components, and the amplitude of the non-extreme points can be directly set to zero.

5. The double-terminal traveling wave ranging wave head calibration method applicable to the distribution network according to claim 3 is characterized in that: For the mode extreme point screening process in S4, the screening scheme starts from the scale corresponding to the characteristic frequency band and traces back in reverse order to the cutoff scale j=2.

6. The double-terminal traveling wave ranging wave head calibration method applicable to the distribution network according to claim 3 is characterized in that: The steps for selecting the extreme value points in S4 are as follows: (1) Calculate and obtain the scale corresponding to the target frequency band: Based on the sampling frequency fs and the median value of the target frequency band Target_fs, the scale corresponding to the target frequency band is obtained by the formula r = floor{1 + log(fs / 2 / Target_fs) / log(2.0)}, where floor{} means rounding up the variables in the brackets, and the processing object is the modulus of the high-frequency component coefficients of each scale in the modulus matrix. ; (2) Starting from the Rth scale high frequency component, search element by element sequentially, where R = r + 1, with Rat max_R is the threshold, the proportional coefficient Rat=0.35, max_R is the maximum absolute value of the high-frequency component amplitude of this scale, and the point number corresponding to the first maximum or minimum value of the high-frequency component element of this scale whose absolute value is greater than the threshold is sequentially searched. This point number is used as the modulus extreme point of this scale, recorded as MaxModPoint_R; (3) In the upward scale R-1, based on the range of plus or minus 2 of the point MaxModPoint_R, sequentially search for elements whose absolute value exceeds the threshold Rat max_R-1, the first modulus extreme value and corresponding point with the same sign as MaxModPoint_R, record this point as the modulus extreme value point of this scale: MaxModPoint_(R-1), if the search is empty, record MaxModPoint_(R-1) =MaxModPoint_R; (4) Enter the next higher scale and repeat the search mechanisms of (2) and (3) until the second scale stops and the next higher scale is reached. The overall modulus extreme point is recorded as MaxModPoint = MaxModPoint_2, where MaxModPoint_2 is the modulus extreme point at the second scale search. (5) The modulus extreme points that meet the requirements at both ends of M and N are recorded as MaxModPoint_M and MaxModPoint_N respectively.

7. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 6.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the computer program is executed by the processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 6.

Images