A Fault Fusion Ranging Method and Device for Transmission Lines

Abstract

The present application relates to the technical field of fault detection and location, and particularly relates to a method and device for fusing and ranging transmission line faults. The method includes: collecting a current sequence composed of current traveling wave signals at the moment of a fault occurring on a transmission line; clustering local elements in the current sequence to obtain each fault subsequence; calculating the trend fluctuation index of each fault subsequence; and further obtaining the peak drop difference and the peak interval difference; calculating the period chaos coefficient of the current sequence; and further obtaining the proportion of the main components and the difference index of the singular value sequence of the current sequence; obtaining the signal purity index of the current sequence based on the correlation between the data of each row in the Hankel matrix, the period chaos coefficient, the proportion of the main components, and the difference index; and ranging the transmission line fault based on the signal purity index. The present application improves the accuracy of fault ranging.

Description

Technical Field

[0001] This application relates to the technical field of fault detection and location, and particularly to a method and device for fusing and ranging transmission line faults. Background Art

[0002] Due to the wide span of transmission lines and the complex and diverse terrains they cross, the transmission environment is relatively complex. Transmission lines are affected by various factors, increasing the probability of faults. To ensure the stability of the power system and provide good power quality, it is very necessary to promptly perform fault ranging and troubleshooting on the fault points of transmission lines after a fault occurs to avoid greater impacts. Fault fusion refers to a technology that locates faults by analyzing the fault information of the line after a transmission line fault occurs.

[0003] The traveling wave ranging method is based on the traveling waves generated when a transmission line fault occurs to locate the fault point. It is applicable to various types of transmission lines, and the traveling waves are generated almost immediately after the fault occurs. Therefore, the fault can be quickly located, which is beneficial to quickly restoring power supply. After a transmission line fault occurs, a transient fault voltage will be generated at the fault location, and the current traveling wave signal will have singularities. Since there is a large amount of noise in the traveling wave signal, only by accurately identifying the singularities of the fault signal can the position of the fault point be accurately obtained. Existing technologies usually use wavelet denoising to denoise the traveling wave signal and extract the information of the wave head based on the traveling wave ranging method to achieve fault ranging. However, wavelet denoising usually requires setting a threshold to distinguish the noise information and useful information in the signal, and the adaptability is insufficient. Moreover, the collected traveling wave signals are easily affected by various external environments of the transmission line, resulting in inconsistent noise interference levels in different traveling wave signals. Therefore, the denoising effect of wavelet denoising is not good, which affects the accuracy of fault ranging. Summary of the Invention

[0004] To solve the above technical problems, the purpose of this application is to provide a method and device for fusing and ranging transmission line faults, and the specific technical solutions adopted are as follows:

[0005] In the first aspect, an embodiment of this application provides a method for fusing and ranging transmission line faults, and the method includes the following steps:

[0006] Collect the current sequence composed of the current traveling wave signals at the moment when a fault occurs on the transmission line;

[0007] Divide the local elements according to the distribution of the wave troughs in the current sequence, and cluster the local elements in the current sequence according to the distribution of the local elements to obtain each fault subsequence;

[0008] Based on the fluctuation of the local element change trend and the difference between local elements in the fault subsequence, obtain the trend fluctuation index of each fault subsequence;

[0009] Based on the difference in the change trend of the peaks of the fault subsequences and the difference in the change trend of the peak order in the fault subsequences, obtain the peak decline difference and the peak interval difference;

[0010] Based on the peak decline difference, the peak interval difference, the average of the trend fluctuation indices of all fault subsequences, and the average of the similarity between fault subsequences, obtain the period chaos coefficient of the current sequence;

[0011] Construct a Hankel matrix of the current sequence, perform singular value decomposition on the Hankel matrix to obtain each singular value, and based on the distribution, average, and difference of the singular values, obtain the proportion of the main components and the difference index of the singular value sequence of the current sequence;

[0012] Based on the correlation between the data in each row of the Hankel matrix, the period chaos coefficient, the proportion of the main components, and the difference index, obtain the signal purity index of the current sequence; perform fault ranging on the transmission line based on the signal purity index.

[0013] Furthermore, the method for obtaining the fault subsequence is as follows:

[0014] Use the peak-valley detection algorithm to obtain all the peaks and valleys in the current sequence, divide the current sequence into each subsequence from each valley value, calculate the range of the elements in each subsequence, and use the ranges of all subsequences as the input of the clustering algorithm to output each clustering cluster;

[0015] Calculate the mean value of the internal elements of each clustering cluster, take the clustering cluster with the largest mean value of the internal elements as the fault clustering cluster, and take the current subsequence corresponding to the internal elements of the fault clustering cluster as the fault subsequence.

[0016] Furthermore, the method for obtaining the trend fluctuation index is as follows:

[0017] Based on the fluctuation of the local element change trend and the difference between local elements in the fault subsequence, obtain the two-side oscillation coefficient and the two-side change coefficient of each fault subsequence;

[0018] Combine the two-side oscillation coefficient and the two-side change coefficient of each fault subsequence to obtain the trend fluctuation index of each fault subsequence. The formula is: ; where is the trend fluctuation index of the i-th fault subsequence; is the two-side oscillation coefficient of the i-th fault subsequence; is the two-side change coefficient of the i-th fault subsequence.

[0019] Further, the method for obtaining the variation coefficient on both sides is as follows:

[0020] For each fault subsequence, obtain the maximum value in each fault subsequence, and use the subsequence composed of the maximum value and all elements to the left of the maximum value as the left subsequence of each fault subsequence, and use the subsequence composed of all elements to the right of the maximum value as the right subsequence of each fault subsequence;

[0021] Obtain the first-order difference sequence of the left subsequence of each fault subsequence, and use the step function to process all elements in the first-order difference sequence to obtain the difference step sequence;

[0022] Calculate the absolute value of the difference between the absolute value of the sum of elements in the difference step sequence and the length of the first-order difference sequence of the left subsequence as the oscillation index of the left subsequence of each fault subsequence;

[0023] Adopt the same method as the left subsequence to obtain the oscillation index of the right subsequence of each fault subsequence; calculate the product of the absolute value of the difference and the mean value between the oscillation indices of the left subsequence and the right subsequence as the oscillation coefficient on both sides of each fault subsequence;

[0024] Calculate the variance of the elements in the left subsequence and the variance of the elements in the right subsequence of each fault subsequence, and use the mean value of the variances of the left subsequence and the right subsequence as the variation coefficient on both sides of each fault subsequence.

[0025] Further, the method for obtaining the peak interval difference is as follows:

[0026] Sort all the peaks in all fault subsequences according to their sequence positions in the current sequence to construct a peak sequence; sort the sequence positions of all peaks in the fault subsequence according to their sizes to construct a peak sequence position sequence;

[0027] Obtain the first-order difference sequences of the peak sequence and the peak sequence position sequence, and use the variance of all elements in the first-order difference sequence of the peak sequence as the peak decline difference; use the variance of all elements in the first-order difference sequence of the peak sequence position sequence as the peak interval difference.

[0028] Further, the method for obtaining the period chaos coefficient is as follows:

[0029] Calculate the mean value of the trend fluctuation indices of all fault subsequences as the average fluctuation index;

[0030] Obtain the mean value of the Pearson similarity coefficients between each fault subsequence and all other fault subsequences as the sequence similarity of each fault subsequence;

[0031] The formula for the period chaos coefficient is: ; where; D is the period chaos coefficient of the current sequence; They are the peak drop difference and the peak interval difference; It is the average fluctuation index; P is the mean of the sequence similarities of all fault subsequences; It is a preset parameter adjustment coefficient.

[0032] Furthermore, the method for obtaining the proportion of the main components and the difference index is as follows:

[0033] Sort all singular values in descending order to construct a singular value sequence;

[0034] Take the singular value sequence as the input of the Otsu threshold method. The output of the Otsu threshold method is the threshold T. Take all singular values greater than or equal to the threshold T as important singular values, and take all singular values less than the threshold T as minor singular values; Take the ratio of the number of important singular values to the total number of all singular values as the proportion of the main components of the singular value sequence;

[0035] Calculate the mean and variance of all important singular values in the singular value sequence, and take the product of the mean and variance as the fluctuation index of the important singular values; Use the same method as the fluctuation index of the important singular values to calculate the fluctuation index of the minor singular values, and take the absolute value of the difference between the fluctuation index of the important singular values and the fluctuation index of the minor singular values as the difference index of the singular value sequence.

[0036] Furthermore, the method for obtaining the signal purity index is as follows:

[0037] Calculate the absolute value of the Pearson similarity coefficient between each row of data in the Hankel matrix and the data of other rows, and take the mean of all the absolute values of the Pearson similarity coefficients of each row of data as the similarity coefficient of each row of data; Take the mean of the similarity coefficients of all row data in the Hankel matrix as the self-similarity coefficient of the Hankel matrix;

[0038] The formula for the signal purity index is: ; where H is the signal purity index of the current sequence; G is the proportion of the main components of the singular value sequence; L is the difference index of the singular value sequence; O is the self-similarity coefficient of the Hankel matrix; D is the period chaos coefficient of the current sequence; It is a preset parameter adjustment factor.

[0039] Furthermore, the method for ranging the transmission line fault based on the signal purity index includes:

[0040] Obtain the default threshold for wavelet denoising based on the signal purity index of the current sequence. The formula is: ; where is the default threshold for wavelet denoising; n is the number of elements in the current sequence, H is the signal purity index of the current sequence, is the logarithmic function with base 10; is a preset parameter adjustment index;

[0041] Denoise the current traveling wave signal according to the default threshold of wavelet denoising, and calculate the fault distance in the transmission line using the traveling wave ranging method.

[0042] In a second aspect, an embodiment of the present application further provides a transmission line fault fusion ranging device, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of the method described in any one of the above are implemented.

[0043] The present application has at least the following beneficial effects:

[0044] The present application constructs the trend fluctuation index of the subsequence, calculates the oscillation degree of the subsequence and the monotonic degree on both sides of the maximum value; then constructs the period chaos coefficient, calculates the similarity degree of the fault subsequence in the current sequence, thereby evaluating the degree of noise interference received by the current traveling wave signal; finally constructs the signal purity index, further evaluates the noise content in the current traveling wave signal, and calculates the threshold of wavelet denoising based on this.

[0045] The beneficial effect is that by constructing the signal purity index to evaluate the degree of noise interference in the current traveling wave signal, and then dynamically adjusting the threshold of wavelet denoising, the problem that the existing technology has poor denoising effect due to inconsistent noise interference degrees in different traveling wave signals is solved, the denoising effect of the current traveling wave signal is improved, and the accuracy of fault ranging is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0047] Figure 1 is a flowchart of the steps of a transmission line fault fusion ranging method provided by an embodiment of the present application;

[0048] Figure 2 is an undenosed waveform diagram of current traveling wave data provided by an embodiment of the present application;

[0049] Figure 3 is a schematic diagram of the existing filtering denoising effect provided by an embodiment of the present application;

[0050] Figure 4 is a schematic diagram of the improved filtering denoising effect provided by an embodiment of the present application. Detailed implementation manners

[0051] To further elaborate on the technical means and effects adopted by this application to achieve the intended invention purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation manners, structures, features, and effects of a transmission line fault fusion ranging method and device proposed according to this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0052] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this application belongs.

[0053] The following specifically describes the specific solutions of a transmission line fault fusion ranging method and device provided by this application with reference to the accompanying drawings.

[0054] Please refer to Figure 1 , which shows a step flowchart of a transmission line fault fusion ranging method provided by an embodiment of this application. The method includes the following steps:

[0055] Step S1, obtain relevant data and process it.

[0056] This application uses Rogowski coil traveling wave sensors installed on the transmission line to collect the current traveling wave signals that appear instantaneously when a fault occurs on the transmission line. The data acquisition frequency is 300KHz, and the acquisition duration each time is 0.1s. The collected current traveling wave signals are converted into numerical data through an analog-to-digital converter, and a current sequence is constructed according to the time sequence of data acquisition.

[0057] Step S2, construct a trend fluctuation index to evaluate the degree of data oscillation inside the current traveling wave signal when a fault occurs on the transmission line; construct a period chaos coefficient to evaluate whether the fault subsequence in the current sequence still has similarity; finally, construct a signal purity index to evaluate the degree of noise interference received by the current sequence.

[0058] When no fault occurs, the current signal has strong periodicity; if a fault occurs, the current signal will almost suddenly increase simultaneously when the fault occurs, and the data fluctuation becomes several times larger; however, although the current signal has a sudden increase, after the sudden increase, the current signal still has periodicity, the intervals between peaks are the same, and the peaks and faults both decrease steadily to a certain extent over time.

[0059] If the noise interference level in the current traveling wave signal is relatively low, the fluctuation data in the current sequence still has a certain degree of periodicity; if the noise interference level is very high, the periodicity of the fluctuation data in the current sequence will be destroyed; therefore, the noise interference level can be evaluated by calculating whether the fluctuation data in the current sequence still has periodicity after the fault occurs.

[0060] Since the data collected is when the transmission line fails, and the trend difference and data fluctuation difference between the current data after the fault and before the fault are both relatively large, but at the same time, the current data after the fault and before the fault both have a certain degree of internal similarity, so the abnormal current data with large fluctuations after the fault can be extracted by clustering.

[0061] All the peaks and valleys in the current sequence are obtained through the peak-valley detection algorithm, and the current sequence is segmented into multiple current subsequences according to the valleys. Calculate the range of each current subsequence, and use the ranges of all current subsequences as the input of the k-means clustering algorithm. Set the k value in the k-means clustering algorithm to 2, and the output of the k-means clustering algorithm is 2 clustering clusters. The k-means algorithm is a well-known technology and will not be elaborated here.

[0062] Calculate the mean value of the internal elements of each clustering cluster. Since the data fluctuation of the current after the fault is much larger than that before the fault, and the range of the current subsequence after the fault is larger, the clustering cluster with the largest mean value of the internal elements is used as the fault clustering cluster. The current subsequence corresponding to the internal elements of the fault clustering cluster is used as the fault subsequence.

[0063] For each fault subsequence, obtain the maximum value in each fault subsequence. The subsequence composed of the maximum value and all the elements on the left side of the maximum value is used as the left subsequence of each fault subsequence, and the subsequence composed of all the elements on the right side of the maximum value is used as the right subsequence of each fault subsequence.

[0064] If the current traveling wave signal is less affected by noise interference, the data fluctuation within each fault subsequence should be relatively smooth, the change between data is relatively small and orderly. Therefore, the left subsequence and the right subsequence should respectively show an increasing and a decreasing trend, and the degree of increase or decrease each time is relatively consistent; if the noise interference level is very high, the data fluctuation will be large and disorderly.

[0065] Obtain the first-order difference sequence of the left subsequence of each fault subsequence, and through the function judges and modifies all the elements in the first-order difference sequence to obtain the difference step sequence. The modification rule is as follows: if the element value is greater than 0, the element is set to 1; if the element value is 0, the element remains unchanged; if the element value is less than 0, the element is set to -1.

[0066] Take the absolute value of the difference between the sum of the elements in the differential step sequence and the length of the first-order difference sequence as the oscillation index of the left subsequence of each fault subsequence; when the oscillation index is 0, it proves that the left subsequence has strict monotonicity. The larger the oscillation index, the worse the monotonicity of the left subsequence. By the function, it is possible to judge the monotonicity of the sequence without considering the magnitude of the element values.

[0067] In the same way, obtain the oscillation index of the right subsequence of each fault subsequence; calculate the absolute value and the mean value of the difference between the oscillation indices of the left subsequence and the right subsequence, and take the product of the absolute value of the difference and the mean value as the two-side oscillation coefficient of each fault subsequence.

[0068] Calculate the variance of the elements in the left subsequence and the variance of the elements in the right subsequence of each fault subsequence respectively, and take the mean value of the variances of the left subsequence and the right subsequence as the two-side change coefficient of each fault subsequence. The smaller the two-side change coefficient, the more consistent the degree of change of the data on both sides of the maximum value in the fault subsequence each time, and the relatively stable the fluctuation of the data on both sides.

[0069] Furthermore, in order to evaluate the data smoothing degree of each fault subsequence, construct the trend fluctuation index of each fault subsequence, and the formula is: . In the formula, is the trend fluctuation index of the i-th fault subsequence; is the two-side oscillation coefficient of the i-th fault subsequence; is the two-side change coefficient of the i-th fault subsequence.

[0070] If the two-side oscillation coefficient is smaller, it indicates that in the i-th fault subsequence, both the left subsequence and the right subsequence have strong monotonicity, and the degree of monotonicity is relatively consistent, and the possibility of data oscillation is small; if the two-side change coefficient is smaller, it indicates that the change amplitude of the internal data in the left subsequence and the right subsequence is relatively consistent each time; therefore, if the trend fluctuation index is smaller, it indicates that the internal data change trend of the i-th fault subsequence is smoother, and the change amplitude of the data each time is more consistent, and the influence of noise interference is smaller.

[0071] If the two-side oscillation coefficient is larger, it indicates that the monotonicity of the left subsequence and the right subsequence is weaker, and the possibility of large data oscillation is greater; if the two-side change coefficient is larger, it indicates that the data change amplitude in the left subsequence and the right subsequence is also larger; therefore, if the trend fluctuation index The larger it is, the greater the amplitude of each change in the numbers in the $i$-th fault subsequence, and the greater the influence of noise interference.

[0072] Thus, the trend fluctuation index of each fault subsequence is obtained.

[0073] Sort all the peaks in all fault subsequences according to their ordinal positions in the current sequence to construct a peak sequence; sort the ordinals of all peaks in the fault subsequence according to their magnitudes to construct a peak ordinal sequence.

[0074] If the degree of noise interference during the transmission line fault is very small, then the degree of decline between adjacent peaks in the post-fault current sequence is relatively consistent, and the intervals between adjacent peaks are relatively consistent, so the differences between adjacent data in the peak sequence and the peak ordinal sequence are relatively consistent.

[0075] Obtain the first-order difference sequence of the peak sequence, and take the variance of the first-order difference sequence as the peak decline difference; similarly, take the variance of the first-order difference sequence of the peak ordinal sequence as the peak interval difference.

[0076] Furthermore, calculate the mean of the trend fluctuation indices of all fault subsequences as the average fluctuation index.

[0077] Obtain the mean of the Pearson similarity coefficients between the $i$-th fault subsequence and all other fault subsequences as the sequence similarity of the $i$-th fault subsequence.

[0078] Furthermore, in order to evaluate the degree of noise interference received by the current traveling wave signal, construct the period chaos coefficient of the current sequence, and the formula is: . Where; $D$ is the period chaos coefficient of the current sequence; are the peak decline difference and the peak interval difference respectively; is the average fluctuation index; $P$ is the mean of the sequence similarities of all fault subsequences; is the preset tuning parameter coefficient. To avoid the denominator being zero, in this embodiment, it takes the value of 1.

[0079] If the peak decline difference and the peak interval difference are both small, it indicates that after the fault occurs, the peaks in each fault subsequence still have strong regularity, the peaks decline with time, and the degree of decline is consistent and the intervals between the peaks are consistent; if the average fluctuation index is small, it indicates that the data within each fault subsequence is relatively smooth and the data change amplitudes are relatively consistent; if the mean $P$ of the sequence similarities is large, it indicates that there is a certain similarity between the fault subsequences and the data trends are relatively consistent; therefore, if the period chaos coefficient $D$ is smaller, it indicates that the degree of noise interference is smaller and there is still a certain similarity between the fault subsequences.

[0080] If the peak drop difference and the peak interval difference are both large, it indicates that the peak fluctuations between the fault subsequences are significant, and they no longer exhibit the regularities of consistent decline and consistent intervals. If the average fluctuation index is larger, it means that the data variation degrees within each fault subsequence are more inconsistent, and the data oscillation degree is greater. If the average value P of the sequence similarity is smaller, it indicates that the correlations between the fault subsequences are poorer, and the data change trends are more inconsistent. Therefore, if the period chaos coefficient D is larger, it means that the current traveling wave signal is more severely affected by noise interference.

[0081] Thus, the period chaos coefficient of the current sequence is obtained.

[0082] Due to the uneven line parameters and unpredictable factors such as external interference on the hardware during the acquisition process, the collected traveling wave signal will contain a large amount of high-frequency noise. A Hankel matrix is constructed from the current sequence and singular value decomposition is performed, and the decomposed singular values are arranged in descending order to construct a singular value sequence. The construction of the Hankel matrix and singular value decomposition are well-known techniques and will not be elaborated here.

[0083] In the current traveling wave signal, larger singular values are usually related to the main components of the traveling wave signal, such as the transient response and fundamental frequency components during a fault, while smaller singular values often correspond to noise components.

[0084] By further analyzing the decomposed singular values, if the noise content in the current traveling wave signal is very low, the rows in the constructed Hankel matrix will be relatively similar, and several larger singular values in the decomposed singular values will correspond to the main components of the current traveling wave signal, and the differences between the larger singular values are significant. If the noise content in the current traveling wave signal is extremely high, the differences between the rows in the constructed Hankel matrix are large. Although there will be larger singular values in the decomposed singular values, since the main components of the signal are masked by noise, the number of larger singular values will be small. On the contrary, the number of smaller singular values will increase, and the singular values corresponding to noise are relatively consistent.

[0085] The singular value sequence is used as the input of the Otsu threshold method. The output of the Otsu threshold method is the threshold T. All singular values greater than or equal to the threshold T are regarded as important singular values, and all singular values less than the threshold T are regarded as minor singular values. The ratio of the number of important singular values to the total number of all singular values is used as the proportion G of the main components in the singular value sequence. The larger the proportion of the main components, the more the number of larger singular values, and the better the signal quality.

[0086] Calculate the mean and variance of all important singular values in the singular value sequence, and take the product of the mean and variance as the fluctuation index of the important singular values; similarly, calculate the fluctuation index of the secondary singular values, and take the absolute value of the difference between the fluctuation index of the important singular values and the fluctuation index of the secondary singular values as the difference index L of the singular value sequence; the larger the difference index, the greater the fluctuation difference and mean difference between the important singular values and the secondary singular values, indicating that the main components of the signal corresponding to the larger singular values are more obvious and the degree of noise interference is smaller.

[0087] In the Hankel matrix, calculate the absolute value of the Pearson similarity coefficient between each row of data and the data of other rows, and take the mean of all the absolute values of the Pearson similarity coefficients of each row of data as the similarity coefficient of each row of data; take the mean of the similarity coefficients of all row data in the Hankel matrix as the self-similarity coefficient O of the Hankel matrix.

[0088] Furthermore, in order to reflect the degree of noise interference of the current traveling wave signal, a signal purity index is constructed, and the formula is: ; where H is the signal purity index of the current sequence; G is the proportion of the main components of the singular value sequence; L is the difference index of the singular value sequence; O is the self-similarity coefficient of the Hankel matrix; D is the period chaos coefficient of the current sequence; is a preset tuning factor, and in order to avoid the denominator being 0, it takes the value of 1 in this embodiment.

[0089] If the proportion of the main components G is larger, it means that the number of larger singular values decomposed from the Hankel matrix constructed by the current sequence is more, and the main components in the corresponding current traveling wave signal are more; if the difference index L is larger, it means that the difference between the important singular values and the secondary singular values is greater; if the self-similarity coefficient O is larger, it means that the data of each row in the Hankel matrix constructed by the current sequence are more similar and the noise content is less; if the period chaos coefficient D is smaller, it means that the degree of noise interference is smaller; therefore, if the signal purity index H is larger, it means that the degree of noise interference of the current traveling wave signal is smaller.

[0090] If the proportion of the main components G is smaller, it means that the number of important singular values is less; if the difference index L is smaller, it means that the difference between the important singular values and the secondary singular values is smaller, the singular value fluctuation is smaller, and the noise content is larger; if the self-similarity coefficient O is smaller, it means that the similarity between each row of data in the Hankel matrix is smaller; if the period chaos coefficient D is larger, it means that the degree of noise interference is larger; therefore, if the signal purity index H is smaller, it means that the degree of noise interference of the current traveling wave signal is larger.

[0091] Step S3, denoise the current traveling wave signal and implement fault location.

[0092] In the method for estimating the threshold of wavelet denoising, the default threshold determination model determines the threshold according to the intensity of the noise. Therefore, in this application, the default threshold of wavelet denoising is calculated through the signal purity index obtained by calculation to obtain the threshold , where is the default threshold of wavelet denoising; n is the number of elements in the current sequence, H is the signal purity index of the current sequence, is the logarithmic function with base 10; is the preset tuning parameter index. To avoid the denominator being 0, the value in this embodiment is 1.

[0093] Furthermore, the collected current traveling wave signal is denoised according to the threshold thr. Among them, the waveform diagram of the un-denoised current traveling wave data is as Figure 2 shown, and the schematic diagram of the existing filtering denoising effect is as Figure 3 shown. Although the existing filtering denoising algorithm has a good denoising effect, the denoising effect is too smooth, and the useful information in the denoised signal is lost, making it difficult to distinguish when the fault occurred; the schematic diagram of the improved filtering denoising effect is as Figure 4 shown. It not only denoises the original data but also does not cause the loss of useful information in the signal. It can be clearly found that the current waveform changes abnormally at 0.025 seconds. Then, for the denoised traveling wave signal, the fault ranging formula in the traveling wave ranging method is used to calculate the fault distance in the transmission line, realizing a method and device for fusing fault ranging of transmission lines.

[0094] Based on the same inventive concept as the above method, an embodiment of this application also provides a device for fusing fault ranging of transmission lines, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above methods for fusing fault ranging of transmission lines.

[0095] It should be noted that: the above sequence of embodiments of this application is only for description and does not represent the advantages or disadvantages of the embodiments. And the above description of specific embodiments of this specification is provided. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0096] Each embodiment in this application is described in a progressive manner. The same or similar parts among the embodiments can be referred to each other, and the key points of each embodiment are the differences from other embodiments.

[0097] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present application shall be included within the protection scope of the present application.

Claims

1. A transmission line fault fusion distance measurement method, characterized in that: The method comprises the following steps: Collect the current sequence composed of the current traveling wave signal at the moment of fault on the transmission line; The local elements are divided according to the distribution of the troughs in the current sequence, and the local elements in the current sequence are clustered according to the distribution of the local elements to obtain each fault subsequence; Based on the fluctuation of the change trend of local elements in the fault subsequence and the difference between the local elements, the trend fluctuation index of each fault subsequence is obtained; Based on the difference in the change trend of the peak value of the fault subsequence and the difference in the change trend of the position sequence of the peak value in the fault subsequence, the peak drop difference and the peak interval difference are obtained; Based on the peak drop difference, the peak interval difference, the average trend fluctuation index of all fault subsequences, and the average similarity between fault subsequences, the period chaos coefficient of the current sequence is obtained; Construct the Hankel matrix of the current sequence, perform singular value decomposition on the Hankel matrix to obtain each singular value, and obtain the main component proportion and difference index of the singular value sequence of the current sequence based on the distribution, average and difference of the singular values; The signal purity index of the current sequence is obtained based on the correlation, periodic chaos coefficient, main component proportion and difference index among the data rows in the Hankel matrix; the transmission line fault is located based on the signal purity index.

2. A transmission line fault fusion distance measurement method as claimed in claim 1, characterized in that: The method for obtaining the fault subsequence is: Use the peak and valley detection algorithm to obtain all peaks and valleys in the current sequence, divide the current sequence into subsequences from each valley, calculate the range of elements in each subsequence, use the range of all subsequences as the input of the clustering algorithm, and output each clustering cluster; The internal element mean of each cluster is calculated, and the cluster with the largest internal element mean is taken as the fault cluster, and the current subsequence corresponding to the internal elements of the fault cluster is taken as the fault subsequence.

3. A transmission line fault fusion distance measurement method as claimed in claim 1, characterized in that: The method for obtaining the trend fluctuation index is: Based on the fluctuation of the change trend of the local elements in the fault subsequence and the difference between the local elements, the oscillation coefficients on both sides and the change coefficients on both sides of each fault subsequence are obtained; Combining the oscillation coefficients on both sides and the variation coefficients on both sides of each fault subsequence, the trend fluctuation index of each fault subsequence is obtained. The formula is: ; In the formula, is the trend fluctuation index of the i-th fault subsequence; is the oscillation coefficient on both sides of the i-th fault subsequence; is the coefficient of variation on both sides of the i-th fault subsequence.

4. A transmission line fault fusion distance measurement method as claimed in claim 3, characterized in that: The method for obtaining the variation coefficients on both sides is: For each fault subsequence, obtain the maximum value in each fault subsequence, take the subsequence consisting of the maximum value and all elements to the left of the maximum value as the left subsequence of each fault subsequence, and take the subsequence consisting of all elements to the right of the maximum value as the right subsequence of each fault subsequence; Obtain the first-order difference sequence of the left subsequence of each fault subsequence, and use the step function to process all elements in the first-order difference sequence to obtain a differential step sequence; Calculate the absolute value of the difference between the absolute value of the sum of the elements in the differential step sequence and the length of the first-order difference sequence of the left subsequence as the oscillation index of the left subsequence of each fault subsequence; Using the same method as the left subsequence, obtain the oscillation index of the right subsequence of each fault subsequence; Calculate the product of the absolute value and the mean of the difference between the oscillation indexes of the left subsequence and the right subsequence as the oscillation coefficient on both sides of each fault subsequence; The variance of the elements in the left subsequence and the variance of the elements in the right subsequence of each fault subsequence are calculated, and the mean of the variance of the left subsequence and the right subsequence is taken as the coefficient of variation on both sides of each fault subsequence.

5. A transmission line fault fusion distance measurement method as claimed in claim 2, characterized in that: The method for obtaining the peak interval difference is: All peaks in all fault subsequences are sorted according to their position in the current sequence to construct a peak sequence; the positions of all peaks in the fault subsequences are sorted according to their size to construct a peak position sequence; The first-order difference sequence of the peak sequence and the peak position sequence is obtained, and the variance of all elements in the first-order difference sequence of the peak sequence is taken as the peak drop difference; the variance of all elements in the first-order difference sequence of the peak position sequence is taken as the peak interval difference.

6. A transmission line fault fusion distance measurement method as claimed in claim 1, characterized in that: The method for obtaining the periodic disorder coefficient is: Calculate the average of the trend fluctuation indexes of all fault subsequences as the average fluctuation index; Obtaining the mean of the Pearson similarity coefficients between each fault subsequence and all other fault subsequences as the sequence similarity of each fault subsequence; The formula for the periodic disorder coefficient is: ; in; D is the periodic disorder coefficient of the current sequence; is the peak drop difference and peak interval difference; is the average fluctuation index; P is the mean of the sequence similarity of all fault subsequences; is the preset tuning coefficient.

7. A transmission line fault fusion distance measurement method as claimed in claim 1, characterized in that: The method for obtaining the main component proportion and difference index is: Sort all singular values ​​in descending order to construct a singular value sequence; The singular value sequence is used as the input of the Otsu threshold method, and the output of the Otsu threshold method is the threshold T. All singular values ​​greater than or equal to the threshold T are regarded as important singular values, and all singular values ​​less than the threshold T are regarded as secondary singular values. The ratio of the number of important singular values ​​to the total number of all singular values ​​is regarded as the main component ratio of the singular value sequence. Calculate the mean and variance of all important singular values ​​in the singular value sequence, and take the product of the mean and variance as the fluctuation index of the important singular value; use the same method as the fluctuation index of the important singular value to calculate the fluctuation index of the secondary singular value, and take the absolute value of the difference between the fluctuation index of the important singular value and the fluctuation index of the secondary singular value as the difference index of the singular value sequence.

8. A transmission line fault fusion distance measurement method as claimed in claim 1, characterized in that: The method for obtaining the signal purity index is: Calculate the absolute value of the Pearson similarity coefficient between each row of data in the Hankel matrix and other rows of data, and use the average of the absolute values ​​of all the Pearson similarity coefficients of each row of data as the similarity coefficient of each row of data; use the average of the similarity coefficients of all rows of data in the Hankel matrix as the self-similarity coefficient of the Hankel matrix; The formula for the signal purity index is: ; In the formula, H is the signal purity index of the current sequence; G is the main component ratio of the singular value sequence; L is the difference index of the singular value sequence; O is the self-similarity coefficient of the Hankel matrix; D is the periodic chaos coefficient of the current sequence; It is the preset parameter factor.

9. A transmission line fault fusion distance measurement method as claimed in claim 1, characterized in that: The method of measuring the distance of a transmission line fault based on a signal purity index comprises: The default threshold of wavelet denoising is obtained based on the signal purity index of the current sequence. The formula is: ; In the formula, is the default threshold of wavelet denoising; n is the number of elements in the current sequence, H is the signal purity index of the current sequence, is a logarithmic function with base 10; is the preset parameter index; The current traveling wave signal is denoised according to the default threshold of wavelet denoising, and the fault distance in the transmission line is calculated using the traveling wave ranging method.

10. A transmission line fault fusion ranging device, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the transmission line fault fusion ranging method as described in any one of claims 1-9 are implemented.

Images