Traveling wave head calibration method, system and equipment based on discrete projection transformation and medium

Abstract

The invention discloses a traveling wave head calibration method, system and equipment based on discrete projection transformation, and a medium, relates to the field of operation and maintenance of power transmission lines of power systems, and aims to solve the influence of sudden change data on projection transformation and simplify a projection result judgment process. Comprising the following steps: monitoring traveling wave current data of a fault channel, retrieving a current abrupt change moment by using a time window accumulation method, intercepting a data segment containing an abrupt change point, converting the data segment into a traveling wave fault graph, and normalizing accumulated current data to obtain a current fault quantity; converting the graph into a two-dimensional binary pixel matrix and assigning according to an original signal time sequence; mojette transformation is carried out on the matrix to obtain a projection value set of different projection angles; searching the maximum value of the set, determining the most sensitive projection angle and coordinates of each wave head, and calculating the arrival time of the wave heads and the time difference between the wave heads. Wave head features are accurately captured through discrete projection transformation, calibration accuracy is improved, and reliable support is provided for power system fault positioning.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power system transmission line operation and maintenance, and particularly relates to a traveling wave front calibration method, system, device and medium based on discrete projection transformation. BACKGROUND

[0002] In the power system, the transmission line is the key link of power transmission. However, due to the long line and the influence of cross-regional operation environment, climate factors and topographic conditions, the line fault occurrence rate is high. Accurate fault location technology plays a crucial role in quickly troubleshooting, saving human resources, ensuring power grid stability and improving system economy. Traveling wave location technology uses the time difference between the traveling wave at the fault point and the measurement point to measure the fault distance. This method is not affected by factors such as fault point transition resistance and line structure, has high location accuracy and wide application range.

[0003] At present, with the in-depth research, the traveling wave location technology is continuously improved and perfected. For example, by analyzing the influence of frequency on the amplitude of the traveling wave front, a fault traveling wave location method based on frequency selection is proposed to improve the location accuracy and reliability. In addition, methods such as wavelet transform, S transform and Hilbert Huang transform are used to determine the traveling wave front, further improving the accuracy of location.

[0004] However, the subsequent wave front of the line fault traveling wave is inevitably affected by multiple fold reflections and line attenuation, and the amplitude will be significantly reduced, making it difficult to identify multiple subsequent wave fronts based on the characteristic detection method of local singularity of sampling data. The above algorithm based on data transformation cannot guarantee the accuracy of identifying the second wave front, which seriously affects the accuracy of location, and there is an urgent need to study a single-ended traveling wave subsequent wave front identification method based on robust macroscopic features.

[0005] In the prior art, as in document 1: Machine Vision Based Traveling Wavefront Calibration Method and document 2: Single-ended traveling wave recognition and fault location of transmission line based on wave shape projection features, a wave head monitoring method based on Radon transform is provided, the maximum sensitive angle corresponding to the traveling wave wave head is solved within a certain angle range, and then the projection curve corresponding to the projection transformation is obtained, and the time difference between the wave heads is judged through the intersection of the curve and the time scale. The method needs to transform the fault signal waveform curve correspondingly, and has a strong correlation with the length of the curve track. However, in the actual measured waveform, the track curve is greatly affected by the sudden data, and the document adopts the measure of repeatedly discriminating the projection curve corresponding to the maximum value, and sets the corresponding criterion through the axisymmetric characteristics of the sudden data, and then eliminates the projection curve corresponding to the sudden data track. The operation is relatively cumbersome, which is not conducive to the realization of the process of the automatic monitoring algorithm, and the robustness in the real scene may be insufficient.

[0006] Therefore, it is urgent to propose a wave head calibration method to solve the above problems. SUMMARY

[0007] The technical problem to be solved by the present application is how to solve the influence of sudden data on projection transformation and reduce the judgment process of the projection result of sudden data.

[0008] The present application solves the above technical problems by the following technical means: a traveling wave wave head calibration method based on discrete projection transformation, comprising:

[0009] S1, monitoring the current traveling wave data of the fault channel, using a time window accumulation method to retrieve the sudden time of the current, intercepting the traveling wave data segment containing the mutation point and converting it into a traveling wave fault pattern;

[0010] S2, converting the traveling wave fault pattern into a two-dimensional binary pixel matrix, wherein the value of the binary pixel matrix element is assigned according to the time sequence of the original current traveling wave signal;

[0011] S3, performing Mojette transformation on the binary pixel matrix to obtain a set of projection values under different projection angles;

[0012] S4, searching for the maximum value in the set of projection values, determining the most sensitive angle and projection coordinates corresponding to each wave head, calculating the arrival time of each wave head relative to the starting time of the data, and then obtaining the time difference between the subsequent wave head and the initial wave head.

[0013] The application searches the current mutation time through the time window accumulation method, can accurately locate the key time point of fault occurrence, provides accurate starting data segment for subsequent traveling wave head calibration, ensures that the subsequent analysis is based on effective traveling wave data containing the mutation point, and improves the pertinence and accuracy of fault analysis. The traveling wave fault pattern is converted into a two-dimensional binary pixel matrix, and the traveling wave signal is presented in a matrix form, which lays a foundation for the extraction of wave head characteristics. The binary pixel matrix is subjected to Mojette transformation, the projection value set can be obtained under different projection angles, the characteristics of the traveling wave pattern can be captured from multiple angles, the morphological information of the traveling wave head can be more comprehensively reflected, and multi-dimensional data support is provided for the accurate identification of the subsequent wave head. The maximum value in the projection value set is searched, the most sensitive angle and projection coordinates corresponding to each wave head are determined, the arrival time of each wave head relative to the starting time of the data is calculated, and then the time difference between the subsequent wave head and the initial wave head is obtained, so that the accurate calibration of the traveling wave head is realized, key and accurate time parameters are provided for power system fault location and fault type judgment, and the efficiency and reliability of power system operation and maintenance are improved.

[0014] Further, the step S1 adopts the method of time window accumulation to search the mutation time of current, specifically:

[0015] A fixed time window step is set, the current traveling wave data is accumulated by sliding according to the time window width, and the current data I after time window accumulation is obtained step (i), the current traveling wave data under different time windows is independent and subject to normal distribution N(mu, sigma 2 ), and the current data mutation judgment threshold is set as

[0016] It is assumed that the current traveling wave satisfies the normal state hypothesis H0 when no fault occurs, and the distribution S satisfies the following formula:

[0017]

[0018] When the current is mutated, the fault current traveling wave data satisfies the fault state hypothesis H1, and the detection data value I step (i) satisfies the following formula:

[0019]

[0020] Wherein, mu i is the mean value of the current traveling wave data in the i th time window under the normal state, sigma is the standard deviation of the current traveling wave data under the normal state, is the 2 critical point of the normal distribution N(mu, sigma ).

[0021] Further, the current data I after time window accumulation I step(i) a current fault quantity is obtained by normalization, specifically:

[0022] Distribution of current traveling wave data in normal state I step (i) ~ N(μ,σ 2 , the current fault quantity W step is obtained by normalizing the current data accumulated by the time window, and the calculation formula is as follows:

[0023]

[0024] Wherein, W step (0) is the initial value, μ(i) is the mean value of the data before i step , and σ(i) is the standard deviation of the data before i step .

[0025] Further, the step S2 is specifically:

[0026] Firstly, according to the pixel value of the traveling wave fault pattern, a pixel matrix with pixel longitudinal and horizontal arrangement as coordinates is constructed, and the pixelization expression of the pixel matrix is {f(k,l)|k=1,2,...,K;l=1,2,...,L};

[0027] Then, all elements of the pixel matrix are assigned 0, that is, for all (k,l), f(k,l)=0;

[0028] Finally, the pixel matrix elements are assigned values according to the time sequence of the original current traveling wave signal, that is,

[0029]

[0030] Wherein, f(k,l) is the pixel brightness value of the kth row and lth column position in the image, and K and L represent the horizontal and vertical pixel numbers of the image respectively.

[0031] Further, the Mojette transformation in the step S3 is specifically:

[0032] Suppose the angles of different projection rays are {θ j |tanθ j =q j / p j , gcd(p j ,q j )=1}, each value in the projection set is obtained by summing all elements on the corresponding projection line, and the transformation formula is:

[0033]

[0034] Wherein,

[0035]

[0036] where gcd(p j ,q j ) represents the greatest common divisor of two integers p j and q j , the subscript j represents a sequence of projection lines, each element after transformation is obtained by accumulating the pixel values on the line b+kq-pl=0, b, q and p are three parameters that determine the projection line, and when 1 projection angle {θ j |tanθ j =q j / p j , gcd(p j , q j )=1} is determined, b represents the new coordinates of the matrix element after the projection transformation, and the corresponding element set can be represented as:

[0037]

[0038] where I is the number of element sets;

[0039] The number of elements in each projection set is:

[0040] n j =(K-1)|q j |+(L-1)|p j |+1.

[0041] Further, the most sensitive angle corresponding to each wave front in the step S4 satisfies the following formula:

[0042]

[0043] where θ sena is the most sensitive angle, h a is the amplitude intensity of the mutation, Δθ is the change of the most sensitive angle of the observation waveform caused by the line distortion, ΔH is the change of the amplitude intensity of the line integral caused by the line loss at the most sensitive angle, and H is the amplitude intensity of the line integral.

[0044] Further, the calculation formula of the time difference between the subsequent wave front and the initial wave front in the step S4 is:

[0045]

[0046] where m a is the horizontal pixel coordinate of the starting point of the a-th wave front in the pixel matrix, M is the total number of image horizontal pixels, m0 is the horizontal pixel coordinate corresponding to the starting time 0 of the waveform, and N is the total time length corresponding to the total number of image pixels.

[0047] The application further provides a traveling wave front calibration system based on a discrete projection transformation, comprising:

[0048] A signal extraction module is configured to monitor current traveling wave data of a fault channel, retrieve a mutation time of the current by using a time window accumulation method, intercept a traveling wave data segment containing the mutation point, and convert the traveling wave data segment into a traveling wave fault pattern.

[0049] A matrix construction module is configured to convert the traveling wave fault pattern into a two-dimensional binary pixel matrix, wherein values of elements of the binary pixel matrix are assigned according to a time sequence of the original current traveling wave signal.

[0050] A projection transformation module is configured to perform Mojette transformation on the binary pixel matrix to obtain a projection value set under different projection angles.

[0051] A front calibration module is configured to search for a maximum value in the projection value set, determine a most sensitive angle and a projection coordinate corresponding to each wave front, calculate an arrival time of each wave front relative to a starting time of data, and further calculate a time difference between a subsequent wave front and an initial wave front.

[0052] The application further provides a processing device comprising at least one processor and at least one memory connected to the processor in communication, wherein the memory stores program instructions executable by the processor, and the processor calling the program instructions can execute the above method steps.

[0053] The application further provides a computer readable storage medium storing computer instructions, and the computer instructions enable the computer to execute the above method steps.

[0054] The application has the following advantages:

[0055] The application retrieves a mutation time by using a time window accumulation method, intercepts a traveling wave data segment containing the mutation point, and converts the traveling wave data segment into a traveling wave fault pattern, so as to preliminarily focus on effective data, reduce invalid determinations caused by subsequent mutation data interference, construct a binary pixel matrix, convert the traveling wave signal into a regular matrix form, reduce the influence of mutation data, provide a data basis for projection transformation, obtain projection values by using Mojette transformation, compared with a traditional method (such as Radon transformation) which often needs to perform multiple determinations on a projection curve and remove mutation interference, the application can more stably process data and reduce a complex determination process of projection results of mutation data, search for projection maximum values to determine a most sensitive angle and a projection coordinate corresponding to each wave front, calculate an arrival time of each wave front relative to a starting time of data, and further calculate a time difference between a subsequent wave front and an initial wave front, so as to realize accurate calibration of the traveling wave front, provide key and accurate time parameters for fault location and fault type determination of a power system, and help to improve the efficiency and reliability of operation and maintenance of the power system. BRIEF DESCRIPTION OF DRAWINGS

[0056] Figure 1 Flow chart of a wave front calibration method based on discrete projection transformation for embodiment 1 of the present application;

[0057] Figure 2 Extraction of fault passage wave data graph for embodiment 1 of the present application;

[0058] Figure 3 Interception of fault current waveform graph for embodiment 1 of the present application;

[0059] Figure 4 Mutation data graph after assigning values according to time series of signals for embodiment 1 of the present application;

[0060] Figure 5 Projection principle graph based on Mojette transformation for embodiment 1 of the present application;

[0061] Figure 6 Graph using original image pixels for embodiment 1 of the present application;

[0062] Figure 7 Sensitivity angle graph of different wave fronts for embodiment 1 of the present application. DETAILED DESCRIPTION

[0063] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in a clear and complete manner in conjunction with the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0064] Embodiment 1

[0065] As shown in Figure 1 , a wave front calibration method based on discrete projection transformation includes:

[0066] S1, monitoring current wave data of a fault passage, using a time window accumulation method to retrieve a current mutation time, intercepting a wave data segment containing a mutation point and converting it into a wave fault graph.

[0067] Specifically, for the monitored current wave data of the fault passage, first, the mutation current data in a suitable time period is intercepted. Based on the mutation point, a suitable current wave data segment is extracted. As shown in Figure 2 , the current wave data of the fault passage is extracted.

[0068] Suppose the measured fault current data is i sIt contains current traveling wave data from s time observation points. The abrupt change in current is retrieved using a time-window accumulation method. A fixed time window step is set, and the collected fault current data is accumulated by sliding according to the time window width to obtain the accumulated current data I. step (i) Current traveling wave data under different time windows are independent of each other and follow a normal distribution N(μ1,σ). 2 The statistical characteristics of the accumulated current data for all time windows are a normal distribution with a mean of μ1 and a variance of σ.

[0069] The basis for monitoring abrupt change points is to monitor the average value of all time windows. Assuming that the average value of the time window changes after point n, then the current data I of the time window after point n... step (n+1), I step (n+2), ..., I step The accumulated data from (n+1) onwards follow a normal distribution with mean μ2 and variance σ. The criterion for determining abrupt changes in the fault current traveling wave waveform is μ1 < μ2.

[0070] In practice, |μ1-μ2|>σ / 2 is set as the threshold for detecting abrupt changes. When performing sliding accumulation on the current traveling wave data, the threshold for detecting current data is set to...

[0071] Assuming no fault occurs, the current traveling wave satisfies the normal state assumption H0, and its distribution S satisfies the following formula:

[0072]

[0073] When a sudden change in current occurs, if the fault current traveling wave data satisfies fault state assumption H1, then the detected data value I... step (i) Satisfies the following formula:

[0074]

[0075] Where, μ i Let σ be the mean of the current traveling wave data within the i-th time window under normal conditions, and σ be the standard deviation of the current traveling wave data under normal conditions. The distribution is normally distributed N(μ,σ). 2 )of The critical point of division.

[0076] Distribution of current traveling wave data under normal conditions I step (i)~N(μ,σ 2 ), where μ is the mean of the normal distribution and σ is the variance of the normal distribution, the current fault quantity W is obtained by normalizing the current data after the time window accumulation. step The calculation formula is as follows:

[0077]

[0078] where W step (0) is the initial value, μ(i) is the mean of the data before i step σ(i) is the standard deviation of the data before i step The current traveling wave data in normal state refers to the data in the first half of the traveling wave data without mutation.

[0079] Set the mutation threshold w set1 is the critical point of σ / 2 of N(0, 1), which corresponds to the normalized Compare the detected data, and record the position of the mutation point when it exceeds the threshold. Take the mutation point position as the reference to intercept the appropriate traveling wave fault pattern, generally select 2000 data points, select the waveform graph of 200 μs before the initial mutation and 1800 μs after the initial mutation, and the image pixel is selected as 1201*901.

[0080] S2, convert the traveling wave fault pattern into a two-dimensional binary pixel matrix, wherein the value of the binary pixel matrix element is assigned according to the time sequence of the original current traveling wave signal.

[0081] Specifically, there are a large amount of mutation data in the traveling wave fault pattern, and the waveform lines between the mutation data will form a long pixel track on the image, which will affect the transformation result in the projection transformation. A large number of studies propose the projection of mutation data through the recognition of waveform characteristics, and the present application adopts a discretized data structure to reduce the influence of data mutation.

[0082] First, according to the pixel value of the traveling wave fault pattern, a pixel matrix with pixel vertical and horizontal arrangement as coordinates is constructed, and the pixelization expression of the pixel matrix is {f(k, l) | k = 1, 2,..., K; l = 1, 2,..., L};

[0083] Then, all elements of the pixel matrix are assigned 0, that is, for all (k, l), f(k, l) = 0;

[0084] Finally, the pixel matrix elements are assigned values according to the time sequence of the original current traveling wave signal, that is,

[0085]

[0086] where f(k, l) is the pixel brightness value of the kth row and the lth column in the image, and K and L represent the horizontal and vertical pixel numbers of the image, respectively.

[0087] Figure 3 The intercepted current waveform graph is greatly affected by data mutation. The discretized matrix after the assignment processing will reduce the length of the mutation data in the pixel matrix. For example, Figure 4As shown in Figure 4 In the pixel matrix without assignment processing, the pixel value content corresponding to the mutation data is continuous, so in the process of projection transformation, the superposition of continuous pixel values will increase the value of the mutation data after transformation, which will cause misidentification results in the maximum value search. However, in the pixel matrix after assignment processing, the continuous pixels corresponding to the mutation data are replaced by the original data, and the pixel value in the matrix is greatly reduced, thereby reducing the weight in the maximum value search process.

[0088] S3, performing Mojette transformation on the binary pixel matrix to obtain a projection value set under different projection angles.

[0089] Specifically, Mojette transformation is a set of projection relationships of pixel values of an image block at specific angles, as shown in Figure 5 Assuming that the pixelization expression of the pixel matrix is {f(k, l) | k = 1, 2,..., K; l = 1, 2,..., L}, the angles of different projection rays are {θ j |tanθ j =q j / p j ,gcd(p j ,q j )=1}, and each value in the projection set is obtained by summing all elements on the corresponding projection line. The transformation formula is:

[0090]

[0091] Wherein:

[0092]

[0093] Wherein, gcd(p j ,q j ) represents the greatest common divisor of two integers p j and q j , subscript j represents the sequence of projection straight lines, and each element after transformation is obtained by accumulating the pixel values on the straight line b+kq-pl=0, b, q and p are three parameters determining the projection straight line, and when the projection angle {θ j |tanθ j =q j / p j ,gcd(p j ,q j )=1} is determined, b represents the new coordinates of the matrix element after projection transformation, and the corresponding element set can be represented as:

[0094]

[0095] Wherein, I is the number of element sets;

[0096] The number of elements in each projection set is:

[0097] n j =(K-1)|q j |+(L-1)|p j |+1.

[0098] The Mojette transform is applied to the transform of the traveling wave data, the curve image of the traveling wave data is regarded as a two-dimensional plane matrix, the pixels of the image correspond to the horizontal and vertical coordinates of the matrix, f(k,l) represents the value of the binary digital image at the (k,l) pixel, K and L represent the horizontal and vertical pixel numbers of the image. According to the pixel level of the image, K=1201 and L=901.

[0099] In this embodiment, the Mojette transform is performed on the discretization in the range of [30°, 90°], the polar radial discretization is performed on the diagonal pixel 1503 of the waveform image 1201*901 by 1 pixel unit, and the element set M of the waveform image under different angles is obtained j f(k,l) is obtained. The Mojette transform is performed at the projection angle of 0°, and the horizontal coordinate position is solved.

[0100] By performing the Mojette transform on the matrix at different projection angles, the projection value set corresponding to each angle can be obtained. By searching for the maximum value and its coordinates in the set, the most sensitive projection angle θ seni of each wave head can be determined.

[0101] S4, searching for the maximum value in the projection value set, determining the most sensitive angle and projection coordinates corresponding to each wave head, calculating the arrival time of each wave head relative to the starting time of the data, and further obtaining the time difference between the subsequent wave head and the initial wave head.

[0102] Specifically, in the time-domain waveform image of the fault line, the wave head has a significant mutation straight line feature. And affected by the multiple reflections of the traveling wave in the line, multiple subsequent wave heads are presented, and the most sensitive projection angle θ seni of the subsequent wave head compared with the initial wave head and the amplitude intensity l i generated by the mutation gradually decrease regularly, it is assumed that the most sensitive angle and the mutation intensity corresponding to each wave head satisfy the following formula:

[0103]

[0104] Wherein, θ sena is the most sensitive angle, h aΔθ represents the amplitude intensity of the abrupt change, ΔH represents the change in the line integral abrupt change intensity at the angle most sensitive to the observed waveform due to line distortion, and H represents the amplitude intensity of the line integral.

[0105] Because the most sensitive angle and the intensity of wavefront abrupt changes gradually decrease in the time domain, the Mojette transform results show a successive appearance of peaks as the projection angle decreases. The projection angles corresponding to different peaks reflect the most sensitive angles of different wavefronts. Furthermore, performing a Mojette transform at a 0° projection angle allows for the retrieval of the straight line corresponding to the horizontal abscissa through a maximum value search.

[0106] Search for the maximum value among all elements in the projection value set to obtain the most sensitive angle θ corresponding to different wavefronts. sena The amplitudes of different wavefronts within the sensitivity angle range are shown in Table 1. Figure 6 and Figure 7 The original pixel matrix and the projection transformation results of the matrix after assignment are shown respectively. Figure 6 In the process, because abrupt changes create a long straight line trajectory in the waveform curve, they can easily be misidentified as the curve corresponding to the wavefront during projection transformation. However, after assignment, the weight of the abrupt changes in the transformation is greatly reduced, and the projected straight line corresponding to the fault wavefront is effectively identified.

[0107] Table 1. Amplitude of different wavefronts within the sensitivity angle range

[0108]

[0109]

[0110] The intersection of the straight lines representing different wavefronts with the straight line on the horizontal axis represents the starting coordinate m of each wavefront. a The arrival time t of each wavefront relative to the data start time is obtained by converting the pixel horizontal coordinate to the data point reference position using the following formula. a Then, the time difference between the subsequent wavefront and the initial wavefront can be obtained.

[0111]

[0112] Where, m a is the horizontal pixel coordinate of the starting point of the a-th wavefront in the pixel matrix, M is the total number of pixels in the horizontal direction of the image, m0 is the horizontal pixel coordinate corresponding to the starting time 0 of the waveform, and N is the total duration corresponding to the total number of pixels in the image.

[0113] By searching for the maximum value of the projected lines corresponding to different wavefronts and the projected lines corresponding to their x-coordinates, the pixel positions of the intersection points are solved simultaneously. See Table 2 for the wavefront origin coordinates and arrival coordinates.

[0114] Table 2 Wave front start coordinate and arrival time coordinate

[0115] Wave front Initial wave front 1st subsequent wave front 2nd subsequent wave front most sensitive angle θ sen / deg 68 51 43 Start horizontal pixel / px 310 372 434 Coarse time stamp / μs 199 359 520 Fine time stamp / μs 204 371 539

[0116] Embodiment 2

[0117] Based on embodiment 1, embodiment 2 of the application further provides a traveling wave front calibration system based on discrete projection transformation, comprising:

[0118] The signal extraction module is configured to monitor current traveling wave data of the fault channel, retrieve a mutation time of the current by using a time window accumulation method, intercept a traveling wave data segment containing the mutation point, and convert the traveling wave data segment into a traveling wave fault pattern.

[0119] Specifically, retrieving the mutation time of the current by using the time window accumulation method specifically comprises:

[0120] A fixed time window step is set, and the current traveling wave data is accumulated by sliding according to the time window width to obtain current data I step (i) after time window accumulation. 2 (i) under different time windows are independent of each other and subject to normal distribution N(μ,σ

[0121] It is assumed that when no fault occurs, the current traveling wave satisfies a normal state hypothesis H0, and the distribution S satisfies the following formula:

[0122]

[0123] When the current is mutated, the fault current traveling wave data satisfies a fault state hypothesis H1, and the detection data value I step (i) satisfies the following formula:

[0124]

[0125] Wherein, μ i is the mean value of the current traveling wave data in the i th time window under the normal state, and σ is the standard deviation of the current traveling wave data under the normal state, 2 is the critical point of the normal distribution N(μ,σ

[0126] The current data I step (i) after time window accumulation is normalized to obtain a current fault quantity, specifically:

[0127] The distribution I step (i) of the current traveling wave data under the normal state is subject to N(μ,σ 2 ), and the current fault quantity W is obtained by normalizing the current data after time window accumulation.​step , the calculation formula is as follows:

[0128]

[0129] wherein, W step (0) is the initial value, μ(i) is the mean value of the data before i step , σ(i) is the standard deviation of the data before i step .

[0130] A matrix construction module is configured to convert the traveling wave fault pattern into a two-dimensional binary pixel matrix, wherein the value of the binary pixel matrix element is assigned according to the time sequence of the original current traveling wave signal.

[0131] Specifically, first, a pixel matrix with pixel longitudinal and horizontal arrangement as coordinates is constructed according to the pixel value of the traveling wave fault pattern, and the pixelization expression of the pixel matrix is {f(k, l) | k = 1, 2,..., K; l = 1, 2,..., L}.

[0132] Then, all elements of the pixel matrix are assigned 0, that is, for all (k, l), f(k, l) = 0.

[0133] Finally, the pixel matrix element is assigned a value according to the time sequence of the original current traveling wave signal, that is, f(k, l) = 1.

[0134]

[0135] wherein, f(k, l) is the pixel brightness value of the kth row and the lth column position in the image, and K and L represent the horizontal and vertical pixel numbers of the image, respectively.

[0136] A projection transformation module is configured to perform Mojette transformation on the binary pixel matrix to obtain a projection value set under different projection angles.

[0137] Specifically, the Mojette transformation is as follows:

[0138] Suppose the angles of different projection rays are {θ j | tanθ j = q j / p j , gcd(p j , q j ) = 1}, and each value in the projection set is obtained by summing all elements on the corresponding projection line, and the transformation formula is as follows:

[0139]

[0140] wherein,

[0141]

[0142] where gcd(p j ,q j ) represents the greatest common divisor of two integers p j and q j , the subscript j represents a sequence of projection lines, each element after transformation is obtained by accumulating the pixel values on the line b+kq-pl=0, b, q and p are three parameters that determine the projection line, when 1 projection angle {θ j |tanθ j =q j / p j , gcd(p j ,q j )=1} is determined, b represents the new coordinates of the matrix element after the projection transformation, and the corresponding element set can be represented as:

[0143]

[0144] where I is the number of element sets;

[0145] The number of elements in each projection set is:

[0146] n j =(K-1)|q j |+(L-1)|p j |+1.

[0147] The wave head calibration module is configured to search for the maximum value in the projection value set, determine the most sensitive angle and the mutation intensity corresponding to each wave head, calculate the arrival time of each wave head relative to the starting time of the data, and further obtain the time difference between the subsequent wave head and the initial wave head.

[0148] Specifically, the most sensitive angle and the mutation intensity corresponding to each wave head satisfy the following formula:

[0149]

[0150] where θ sena is the most sensitive angle, h a is the amplitude intensity of the mutation, Δθ is the change of the most sensitive angle caused by the line distortion, ΔH is the change of the amplitude intensity of the line integral at the most sensitive angle caused by the line loss, and H is the amplitude intensity of the line integral.

[0151] The calculation formula of the time difference between the subsequent wave head and the initial wave head is:

[0152]

[0153] where m ais the horizontal pixel coordinate of the start of the a-th wave crest in the pixel matrix, M is the total number of horizontal pixels of the image, m0 is the horizontal pixel coordinate corresponding to the start time 0 of the wave, and N is the total time length corresponding to the total number of pixels of the image.

[0154] Embodiment 3

[0155] Based on embodiment 1, embodiment 3 of the present application further provides a processing device, comprising at least one processor, and at least one memory connected to the processor in communication, wherein: the memory stores program instructions executable by the processor, and the processor calling the program instructions can execute the method steps of embodiment 1.

[0156] Embodiment 4

[0157] Based on embodiment 1, embodiment 4 of the present application further provides a computer readable storage medium, which stores computer instructions, and the computer instructions make the computer execute the method steps of embodiment 1.

[0158] The above embodiments are only used to illustrate the technical solutions of the present application, but not limit it; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part 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 application.

Claims

1. A method for calibrating the wavefront of a traveling wave based on discrete projection transformation, characterized in that, include: S1. Monitor the current traveling wave data of the fault channel, use the time window accumulation method to retrieve the moment of current change, extract the traveling wave data segment containing the change point and convert it into a traveling wave fault pattern. S2. Convert the traveling wave fault pattern into a two-dimensional binary pixel matrix, wherein the values ​​of the binary pixel matrix elements are assigned according to the time series of the original current traveling wave signal. S3. Perform a Mojette transformation on the binary pixel matrix to obtain a set of projection values ​​under different projection angles; S4. Search for the maximum value in the set of projection values, determine the most sensitive angle and projection coordinates corresponding to each wavefront, calculate the arrival time of each wavefront relative to the data start time, and then obtain the time difference between the subsequent wavefront and the initial wavefront.

2. The traveling wave front calibration method based on discrete projection transformation according to claim 1, characterized in that, The specific steps in step S1, where the time window accumulation method is used to retrieve the abrupt change in current, are as follows: Set a fixed time window step, and slide the current traveling wave data according to the time window width to obtain the current data I after time window accumulation. step (i) Current traveling wave data under different time windows are independent of each other and follow a normal distribution N(μ,σ). 2 Let the threshold for determining sudden changes in current data be... Assuming no fault occurs, the current traveling wave satisfies the normal state assumption H0, and its distribution S satisfies the following formula: When a sudden change in current occurs, if the fault current traveling wave data satisfies fault state assumption H1, then the detected data value I... step (i) Satisfies the following formula: Where, μ i Let σ be the mean of the current traveling wave data within the i-th time window under normal conditions, and σ be the standard deviation of the current traveling wave data under normal conditions. The distribution is normally distributed N(μ,σ). 2 )of The critical point of division.

3. The traveling wave front calibration method based on discrete projection transformation according to claim 2, characterized in that, Current data I after time window accumulation step (i) Normalization is performed to obtain the current fault quantity, specifically: Distribution of current traveling wave data under normal conditions I step (i)~N(μ,σ 2 The current fault quantity W is obtained by normalizing the current data accumulated over the time window. step The calculation formula is as follows: Among them, W step (0) is the initial value, μ(i) is I step The mean of the data up to i, σ(i) is I. step The standard deviation of the data before i.

4. The traveling wave front calibration method based on discrete projection transformation according to claim 1, characterized in that, Step S2 specifically involves: First, based on the pixel values ​​of the traveling wave fault pattern, a pixel matrix is ​​constructed with the vertical and horizontal arrangement of pixels as the coordinates. The pixelated expression of the pixel matrix is ​​{f(k,l)|k=1,2,...,K;l=1,2,...,L}; Then, assign 0 to all elements of the pixel matrix, that is, for all (k,l), f(k,l)=0; Finally, values ​​are assigned to the pixel matrix elements based on the time series of the original current traveling wave signal. Where f(k,l) is the pixel brightness value at the k-th row and l-th column position in the image, and K and L represent the number of horizontal and vertical pixels in the image, respectively.

5. The traveling wavefront calibration method based on discrete projection transformation according to claim 1, characterized in that, The Mojette transform in step S3 is specifically as follows: Assume the angles of different projected rays are {θ} j |tanθ j =q j / p j ,gcd(p j ,q j The transformation formula is: )=1}, where each value in the projection set is obtained by summing all elements on the corresponding projection line. in, Among them, gcd(p j ,q j ) represents p j and q j The greatest common divisor of two integers, where the subscript j represents the sequence of projected lines. Each transformed element is obtained by accumulating the pixel values ​​on the line b + kq - pl = 0. b, q, and p are the three parameters that determine the projected line. When one projection angle {θ} j |tanθ j =q j / p j ,gcd(p j ,q j After )=1} is determined, b represents the new coordinates of the matrix element after the projection transformation, and the corresponding set of elements can be represented as: Where I is the number of elements in the set; The number of elements in each projection set is: n j =(K-1)|q j |+(L-1)|p j |+1。 6. The traveling wave front calibration method based on discrete projection transformation according to claim 1, characterized in that, The most sensitive angle corresponding to each wavefront in step S4 satisfies the following formula: Where, θ sena For the most sensitive angle, h a Δθ represents the amplitude intensity of the abrupt change, ΔH represents the change in the line integral abrupt change intensity at the angle most sensitive to the observed waveform due to line distortion, and H represents the amplitude intensity of the line integral.

7. The traveling wavefront calibration method based on discrete projection transformation according to claim 1, characterized in that, The formula for calculating the time difference between the subsequent wavefront and the initial wavefront in step S4 is as follows: Where, m a is the horizontal pixel coordinate of the starting point of the a-th wavefront in the pixel matrix, M is the total number of pixels in the horizontal direction of the image, m0 is the horizontal pixel coordinate corresponding to the starting time 0 of the waveform, and N is the total duration corresponding to the total number of pixels in the image.

8. A traveling wave front calibration system based on discrete projection transformation, characterized in that, include: The signal extraction module is used to monitor the current traveling wave data of the fault channel. It uses the time window accumulation method to retrieve the moment of current change, extracts the traveling wave data segment containing the change point and converts it into a traveling wave fault pattern. The matrix construction module is used to convert the traveling wave fault pattern into a two-dimensional binary pixel matrix, wherein the values ​​of the binary pixel matrix elements are assigned according to the time series of the original current traveling wave signal. The projection transformation module is used to perform Mojette transformation on the binary pixel matrix to obtain a set of projection values ​​under different projection angles. The wavefront calibration module is used to search for the maximum value in the set of projection values, determine the most sensitive angle and projection coordinates corresponding to each wavefront, calculate the arrival time of each wavefront relative to the data start time, and then obtain the time difference between the subsequent wavefront and the initial wavefront.

9. A processing device, characterized in that, The method includes at least one processor and at least one memory communicatively connected to the processor, wherein the memory stores program instructions executable by the processor, and the processor can execute the method as described in any one of claims 1 to 7 by invoking the program instructions.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause the computer to perform the method as described in any one of claims 1 to 7.

Images