Data analysis method for hidden danger monitoring of distribution network cable

By building an adaptive sliding window and removing noise points, the problem of underfitting or overfitting caused by the fixed window size in traditional cable sheath circulation data warning is solved, and the accuracy and response speed of the cable monitoring system are improved.

CN119025872AActive Publication Date: 2024-11-26ZAOZHUANG POWER SUPPLY COMPANY OF STATE GRID SHANDONG ELECTRIC POWER

Patent Information

Application Number
CN202411513951.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-29
Publication Date
2024-11-26
Estimated Expiration
2044-10-29

AI Technical Summary

Technical Problem

In the traditional method of warning of the cable sheath circulation data of the distribution network, the window size is fixed, resulting in underfitting or overfitting of the data points, resulting in untimely early warning response of the monitoring system, low monitoring accuracy, and inability to accurately obtain abnormal data.

Method used

By obtaining the cable sheath circulation data within the historical period, an adaptive sliding window is built, and the window length is adjusted according to the density and differences of the data points in the window, the final fit window is obtained, data fitting is performed, noise points are eliminated, and the degree of abnormality is calculated. Early warning.

Benefits of technology

It improves the accuracy and response speed of the distribution cable monitoring system, reduces fitting errors, and enhances the safety monitoring capability of the cable operating status.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119025872A_ABST
    Figure CN119025872A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of data processing, in particular to a data analysis method for hidden danger monitoring of a distribution network cable, and the method comprises the steps: constructing a scatter diagram according to a cable sheath circulation value in a historical time period; a sliding window is constructed in the scatter diagram, the complexity is obtained according to the density of the data points in the sliding window, the length of the sliding window is adjusted according to the complexity and the difference between every two data points in the sliding window, and a final fitting window is obtained; according to the sliding step length, the step of obtaining the final fitting window is repeated until the final fitting window contains data of the last sampling moment in the scatter diagram; and fitting the data points in each final fitting window to obtain a final fitting value, obtaining the abnormal degree of the real-time cable sheath ring current value according to the real-time cable sheath ring current value and the final fitting value, and carrying out early warning on the cable according to the abnormal degree, thereby improving the accuracy of hidden danger monitoring on the cable.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a data analysis method for monitoring potential hazards in distribution network cables. Background Technology

[0002] Distribution cables are typically composed of several or groups of stranded conductors, each insulated from the others, and covered with a high-density insulating layer. In the overall power system, they primarily function to transmit and distribute power. Cable lines offer advantages such as high power supply reliability, good safety, and minimal floor space requirements, making them a major focus in urban power distribution network construction and increasingly widely used in the power grids of industrial and mining enterprises such as steel, petrochemical, and coal mines. When a cable is normally energized, a circulating current is induced in its metallic sheath. Under normal circumstances, this sheath circulating current is very small. However, when problems such as decreased insulation of the cable sheath or poor grounding occur, the sheath circulating current will abnormally increase, threatening the cable's operational safety. Sheath circulating current sensors are installed on the grounding wire of the metal shielding layer at the cable terminal to collect the sheath circulating current signal. This signal is then sent to a monitoring and positioning device. By monitoring the long-term trend of the sheath circulating current, trend alarms are triggered to ensure the safe operation of the cable.

[0003] In traditional methods, when issuing early warnings based on sheath circulation data, the sheath circulation data from the past few days is fitted with a fixed-size window to obtain abnormal data of the sheath circulation. However, if the window is too large or too small, it will cause underfitting of the data points, resulting in inaccurate acquisition of abnormal data, untimely early warning response of the monitoring system, and low monitoring accuracy.

[0004] Therefore, how to obtain an adaptive window and fit the sheath circulation data in the adaptive window to obtain abnormal data, so as to improve the accuracy of the monitoring system in monitoring the safety of cables, has become an urgent problem to be solved. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a data analysis method for monitoring potential hazards in distribution network cables, in order to solve the problem of how to obtain an adaptive window and fit abnormal data based on the sheath circulation data in the adaptive window, so as to improve the accuracy of the monitoring system in monitoring the safety of cables.

[0006] This invention provides a data analysis method for monitoring potential hazards in distribution network cables, the method comprising the following steps: The cable sheath circulation current value of the distribution network cable at each sampling time within a preset number of cycles is obtained within a historical period, and the sheath circulation current data sequence corresponding to each cycle is obtained. The sheath circulation current data sequence corresponding to each cycle is mapped to the same scatter plot, where the horizontal axis of the scatter plot is the sampling time and the vertical axis is the cable sheath circulation current value. A sliding window of a preset size is constructed in the scatter plot. The complexity of the sliding window is obtained based on the density of data points in the sliding window. The length of the sliding window is adjusted based on the complexity and the difference between every two data points in the sliding window to obtain the final fitting window. Using the length of the final fitting window as the sliding step, the steps of obtaining the final fitting window are repeated until the final fitting window contains the data points of the last sampling time in the scatter plot, thus obtaining all final fitting windows. The data points contained in each of the final fitting windows are fitted to obtain the final fitted value for each sampling time in each of the final fitting windows. The real-time cable sheath circulation current value is obtained. Based on the real-time cable sheath circulation current value and the final fitted value at the sampling time where the real-time cable sheath circulation current value is located, the degree of abnormality of the real-time cable sheath circulation current value is obtained. The potential danger warning is given to the distribution network cable based on the degree of abnormality.

[0007] Furthermore, determining the complexity of the sliding window based on the density of data points within the sliding window includes: Within the sliding window, a minimum convex polygon containing all data points within the sliding window is constructed, and the ratio between the area of ​​the minimum convex polygon and the area of ​​the sliding window is calculated to obtain a first ratio. Obtain the number of all data points contained within the sliding window, calculate the ratio between the number of all data points and the area of ​​the smallest convex polygon, and obtain a second ratio. The complexity of the sliding window is obtained by weighted summation of the first ratio and the second ratio.

[0008] Furthermore, adjusting the length of the sliding window based on the complexity and the difference between every two data points in the sliding window to obtain the final fitting window includes: If the complexity is less than or equal to a preset complexity threshold, the length of the sliding window is increased by a preset length value to obtain the final fitting window; If the complexity is greater than the preset complexity threshold, then all data points in the sliding window are fitted to obtain the fitting point corresponding to each data point contained in the sliding window. Based on the difference between every two fitting points in the sliding window, the length of the sliding window is adjusted to obtain the final fitting window.

[0009] Furthermore, adjusting the length of the sliding window based on the difference between every two fitting points within the sliding window to obtain the final fitting window includes: For any data point in the sliding window, obtain the left and right neighboring fitting points of the fitting point corresponding to the data point, calculate the absolute value of the slope between the fitting point corresponding to the data point and the left neighboring fitting point to obtain the first absolute slope, calculate the absolute value of the slope between the fitting point corresponding to the data point and the right neighboring fitting point to obtain the second absolute slope, and calculate the absolute value of the slope between the left and right neighboring fitting points to obtain the third absolute slope. Among the first absolute slope and the second absolute slope, the maximum value is selected as the maximum slope value. The absolute value of the difference between the maximum slope value and the third absolute slope value is calculated. The absolute value of the difference is normalized to obtain the fitting difference degree of the data points. The fitting difference degree of each data point in the sliding window is obtained. Based on the fitting difference degree of each data point and the fitting point corresponding to each data point, the noise points contained in the sliding window are obtained. The length of the sliding window is adjusted according to the number of noise points to obtain the final fitting window.

[0010] Furthermore, obtaining the noise points contained in the sliding window based on the degree of fitting difference of each data point and the fitting point corresponding to each data point includes: Calculate the absolute value of the difference between each data point in the sliding window and its corresponding fitting point to obtain the fitting error. Normalize the fitting error corresponding to each data point in the sliding window to obtain the corresponding normalized fitting error value. For any data point in the sliding window, the normalized value of the fitting error of the data point and the degree of fitting difference of the data point are weighted and summed to obtain the noise probability index of the data point. If the noise probability index of the data point is greater than the preset noise probability threshold, the data point is regarded as a noise point.

[0011] Furthermore, adjusting the length of the sliding window based on the number of noise points to obtain the final fitting window includes: Obtain the total number of all data points contained in the sliding window, calculate the ratio between the number of noise points and the total number of all data points, and obtain the noise ratio of the sliding window; If the noise ratio is within the preset noise ratio range, then the sliding window is used as the final fitting window; If the noise ratio is less than the minimum value in the preset noise ratio range, then the length of the sliding window is reduced by the preset length value to obtain the final fitting window; If the noise ratio is greater than the maximum value in the preset noise ratio range, then the length of the sliding window is increased by the preset length value to obtain the final fitting window.

[0012] Furthermore, the step of fitting the data points contained in each of the final fitting windows to obtain the final fitted value for each sampling time within each of the final fitting windows includes: Noise points in each of the final fitting windows are removed, and data points in each of the final fitting windows after noise point removal are fitted to obtain the final fitted value for each sampling time contained in each of the final fitting windows.

[0013] Furthermore, the step of determining the degree of anomaly of the real-time cable sheath circulating current value based on the real-time cable sheath circulating current value and the final fitted value at the sampling time of the real-time cable sheath circulating current value includes: The absolute value of the difference between the final fitted value at the sampling time of the real-time cable sheath circulation current value and the real-time cable sheath circulation current value is calculated to obtain the real-time difference degree. Within the period of the real-time cable sheath circulation current value, for any sampling time before the sampling time of the real-time cable sheath circulation current value, calculate the absolute value of the difference between the cable sheath circulation current value at the sampling time and the final fitted value at the sampling time to obtain the historical difference degree. Obtain the historical difference degree corresponding to each sampling time before the sampling time of the real-time cable sheath circulation current value to obtain the average historical difference degree. The absolute value of the difference between the real-time difference level and the average historical difference level is calculated to obtain the degree of abnormality of the real-time cable sheath circulation value.

[0014] Furthermore, the step of providing early warning of potential hazards to the distribution network cable based on the degree of abnormality includes: If the degree of abnormality is greater than the preset abnormality warning parameter, an abnormality warning will be issued for the distribution network cable.

[0015] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows: This invention acquires the cable sheath circulation current value of a distribution network cable at each sampling time within a preset number of cycles over a historical period, obtaining a sheath circulation current data sequence corresponding to each cycle. This data sequence is then mapped onto the same scatter plot, where the horizontal axis represents the sampling time and the vertical axis represents the cable sheath circulation current value. A sliding window of a preset size is constructed within the scatter plot. The complexity of the sliding window is determined based on the density of data points within it. The length of the sliding window is adjusted according to this complexity and the difference between any two data points within the window, resulting in a final fitting window. The length of the final fitting window is used as the sliding step size. The steps of obtaining the final fitting window are repeated until the final fitting window contains the data point of the last sampling time in the scatter plot, thus obtaining all final fitting windows. The data points contained in each final fitting window are fitted to obtain the final fitted value for each sampling time within each final fitting window, thereby obtaining the real-time cable sheath circulation current value. Based on the real-time cable sheath circulation current value and the final fitted value of the sampling time in which the real-time cable sheath circulation current value is located, the degree of anomaly of the real-time cable sheath circulation current value is obtained. Based on the degree of anomaly, a potential hazard warning is issued for the distribution network cable. Specifically, a sliding window of a preset size is constructed from historical sheath circulation current data. The length of the sliding window is adjusted according to the complexity of the data points in the sliding window to obtain the final fitting window. This increases the number of data points in the sliding window with lower complexity and reduces the fitting error caused by noise points in the sliding window with higher complexity, resulting in a better fitting effect. The data points within the final fitting window are fitted, and the degree of anomaly of the real-time sheath circulation current value is calculated based on the fitting result, thereby issuing anomaly warnings for the distribution network cable and improving the accuracy of cable safety monitoring. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of a data analysis method for monitoring potential hazards in distribution network cables, provided in Embodiment 1 of the present invention. Detailed Implementation

[0018] Embodiments of this disclosure are described in detail below, with examples of these embodiments illustrated in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this disclosure, and should not be construed as limiting it.

[0019] It should be noted that the terms "first," "second," etc., used in this disclosure and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure.

[0020] To illustrate the technical solution of the present invention, specific embodiments are described below.

[0021] See Figure 1 This is a flowchart of a data analysis method for monitoring potential hazards in distribution network cables, provided in Embodiment 1 of the present invention. Figure 1 As shown, the method may include: Step S101: Obtain the cable sheath circulation current value of the distribution network cable at each sampling time within a preset number of cycles during the historical time period, obtain the sheath circulation current data sequence corresponding to each cycle, and map the sheath circulation current data sequence corresponding to each cycle onto the same scatter plot, wherein the horizontal axis of the scatter plot is the sampling time and the vertical axis is the cable sheath circulation current value.

[0022] Since the magnitude of the sheath circulation current in distribution network cables is affected by a variety of factors, including environmental conditions such as temperature and humidity, the circulation current data for a certain period of time is not valuable when issuing early warnings based on the circulation current data for the day. Therefore, it is necessary to comprehensively consider the historical sheath circulation current data over the past few days. In order to facilitate the subsequent fitting of the historical sheath circulation current data, the historical sheath circulation current data collected by the sensor over the past few days is mapped onto the same scatter plot.

[0023] The preset quantity is set to 3, and each cycle is 1 day. There is no limit here. Implementers can set it according to the implementation scenario to obtain the cable sheath circulation current value for each day in the past 3 days, obtain the corresponding sheath circulation current data sequence for each day, and map the sheath circulation current data sequence for the 3 days to the same scatter plot. The vertical axis of the scatter plot is the cable sheath circulation current value, and the horizontal axis is the sampling time for each day.

[0024] Step S102: Construct a sliding window of a preset size in the scatter plot. Determine the complexity of the sliding window based on the density of data points in the sliding window. Adjust the length of the sliding window based on the complexity and the difference between every two data points in the sliding window to obtain the final fitting window.

[0025] In traditional methods, fitting data points to a fixed-size window often results in underfitting due to a small window size or few data points, or overfitting due to high complexity of the data points. Therefore, in this invention, the window size is adaptively changed according to the distribution of data points in the scatter plot, making the anomaly warning of real-time sheath circulation data more accurate.

[0026] First, a sliding window of a preset size is constructed in the scatter plot. The length of the sliding window is set to 5, which is not limited here. The implementer can adjust the height of the sliding window according to the implementation scenario. The height of the sliding window is the absolute value of the difference between the maximum and minimum values ​​in the collected sheath circulation data. The sliding window contains the data points at the first sampling time in the scatter plot. Then, the size of the sliding window is adjusted according to the data points in the sliding window to obtain the final fitting window.

[0027] The specific process of adjusting the size of the sliding window based on the data points in the sliding window to obtain the final fitting window is as follows: (1) The complexity of the sliding window is obtained based on the density of data points in the sliding window.

[0028] Specifically, within the sliding window, a minimum convex polygon containing all data points within the sliding window is constructed, and the ratio between the area of ​​the minimum convex polygon and the area of ​​the sliding window is calculated to obtain a first ratio. Obtain the number of all data points contained within the sliding window, calculate the ratio between the number of all data points and the area of ​​the smallest convex polygon, and obtain a second ratio. The complexity of the sliding window is obtained by weighted summation of the first ratio and the second ratio.

[0029] In one implementation, the formula for calculating the complexity of the sliding window is:

[0030] Where D represents the complexity of the sliding window, This represents the area of ​​the smallest convex polygon. Let n represent the area of ​​the sliding window, and n represent the number of data points contained within the sliding window. Indicates the first weight. This indicates the second weight.

[0031] It should be noted that the settings The value is 0.5. The value is 0.5, and there is no limit here. Implementers can set it according to the implementation scenario. The larger the value, the larger the area of ​​the minimum convex polygon, and thus... The larger the value of D, the more dispersed the data points are within the sliding window; thus, the larger the value of D, the more complex the sliding window. The larger the value of n, the more data points are contained within the sliding window; thus, the larger the value of D, the more complex the sliding window.

[0032] (2) Adjust the length of the sliding window according to the complexity and the difference between every two data points in the sliding window to obtain the final fitting window.

[0033] If the complexity is less than or equal to a preset complexity threshold, the length of the sliding window is increased by a preset length value to obtain the final fitting window; If the complexity is greater than the preset complexity threshold, then all data points in the sliding window are fitted to obtain the fitting point corresponding to each data point contained in the sliding window. Based on the difference between every two fitting points in the sliding window, the length of the sliding window is adjusted to obtain the final fitting window.

[0034] In one implementation, a preset complexity threshold of 0.4 and a preset length of 2 are set, without limitation, and implementers can set them according to the implementation scenario. If the complexity of the sliding window is less than or equal to 0.4, it indicates that the number of data points in the sliding window is small and the distribution is relatively concentrated, resulting in strong stability. In this case, underfitting may occur. To obtain a better fitting effect, it is necessary to increase the number of data points in the sliding window to increase the complexity of the sliding window. Therefore, the length of the sliding window is increased by 2 along the positive x-axis of the scatter plot, and the length of the sliding window is 7. At this time, the number of data points in the sliding window increases, and the sliding window at this time is used as the final fitting window.

[0035] If the complexity of the sliding window is greater than 0.4, it indicates that there are many data points within the sliding window, their distribution is relatively discrete, and the probability of them containing noise points is high. Therefore, it is necessary to perform polynomial fitting on each data point within the sliding window using the least squares method to obtain the fitting point corresponding to each data point in the sliding window, as well as the fitting value of each fitting point. Based on the difference between any two fitting points in the sliding window, the length of the sliding window is adjusted to obtain the final fitting window. The least squares method for polynomial fitting is a current technique and will not be elaborated upon here. The method for obtaining the final fitting window is as follows: For any data point in the sliding window, obtain the left and right neighboring fitting points of the fitting point corresponding to the data point, calculate the absolute value of the slope between the fitting point corresponding to the data point and the left neighboring fitting point to obtain the first absolute slope, calculate the absolute value of the slope between the fitting point corresponding to the data point and the right neighboring fitting point to obtain the second absolute slope, and calculate the absolute value of the slope between the left and right neighboring fitting points to obtain the third absolute slope. Among the first absolute slope and the second absolute slope, the maximum value is selected as the maximum slope value. The absolute value of the difference between the maximum slope value and the third absolute slope value is calculated. The absolute value of the difference is normalized to obtain the fitting difference degree of the data points. The fitting difference degree of each data point in the sliding window is obtained. Based on the fitting difference degree of each data point and the fitting point corresponding to each data point, the noise points contained in the sliding window are obtained. The length of the sliding window is adjusted according to the number of noise points to obtain the final fitting window.

[0036] In one embodiment, since the fitted curve may fluctuate sharply or change direction due to noise, the slope of the fitted point corresponding to each data point in its neighborhood is calculated to determine the degree of difference between the slope of the fitted point and the slope of the overall fitted curve, so as to filter out noise points based on the degree of difference.

[0037] For the i-th data point within the sliding window, the formula for calculating the degree of fit difference for the i-th data point is:

[0038] in, This indicates the degree of difference in fit for the i-th data point. This represents the slope between the fitted point corresponding to the previous sampling time (the left neighbor of the fitted point corresponding to the i-th data point) and the fitted point corresponding to the i-th data point. This represents the slope between the fitted point corresponding to the i-th data point and the fitted point corresponding to the next sampling time after the i-th data point (the right neighbor of the fitted point corresponding to the i-th data point). This represents the slope between the left and right neighboring fitted points of the fitted point corresponding to the i-th data point. Represents the absolute value symbol. This indicates taking the maximum value. This indicates calculating the difference. This represents the linear normalization function.

[0039] It should be noted that, This is the first absolute slope, representing the slope of the fitted point corresponding to the i-th data point in its left neighborhood. This is the second absolute slope, representing the slope of the fitted point corresponding to the i-th data point in its right neighborhood. This is the third absolute slope, representing the slope of the overall fitted curve for the i-th data point. This represents the maximum of the first and second absolute slopes, i.e., the maximum slope value. The greater the difference between the maximum slope value and the third absolute slope value, the more abrupt a fluctuation or turning point occurs at the fitted point corresponding to the i-th data point. The larger the value, the greater the probability that the i-th data point is a noise point.

[0040] Similarly, the degree of fitting difference for each data point in the sliding window is obtained. Further, based on the degree of fitting difference for each data point and the fitting value of the corresponding fitting point, the noise points included in the sliding window are obtained. The method for obtaining the noise points in the sliding window is as follows: Calculate the absolute value of the difference between each data point in the sliding window and its corresponding fitting point to obtain the fitting error. Normalize the fitting error corresponding to each data point in the sliding window to obtain the corresponding normalized fitting error value. For any data point in the sliding window, the normalized value of the fitting error of the data point and the degree of fitting difference of the data point are weighted and summed to obtain the noise probability index of the data point. If the noise probability index of the data point is greater than the preset noise probability threshold, the data point is regarded as a noise point.

[0041] In one embodiment, the formula for calculating the normalized value of the fitting error of the i-th data point within the sliding window is as follows:

[0042] in, This represents the normalized value of the fitting error for the i-th data point. This represents the y-coordinate value of the i-th data point. Let represent the fitted value of the fitted point corresponding to the i-th data point, and K represent the total number of data points in the sliding window. Represents the absolute value symbol.

[0043] It should be noted that, This is the fitting error. The larger the value, the greater the difference between the fitted point corresponding to the i-th data point and the i-th data point itself. The larger the value, the greater the probability that the i-th data point is a noise point.

[0044] Based on the normalized fitting error value of the i-th data point and the degree of fitting difference of the i-th data point, the noise probability index of the i-th data point is obtained. The formula for calculating the noise probability index of the i-th data point is as follows:

[0045] in, This represents the noise probability index for the i-th data point. This represents the normalized value of the fitting error for the i-th data point. This indicates the degree of difference in fit for the i-th data point. This represents the first weighting coefficient. This represents the second weighting coefficient.

[0046] It should be noted that the settings The value is 0.5. The value is 0.5, and there is no limit here. Implementers can set it according to the implementation scenario. The larger the value, the greater the difference between the fitted point corresponding to the i-th data point and the i-th data point itself. The larger the value, the greater the probability that the i-th data point is a noise point; The larger the value, the greater the likelihood of a sharp fluctuation or turning point at the fitted point corresponding to the i-th data point, and thus... The larger the value, the greater the probability that the i-th data point is a noise point.

[0047] The preset noise point probability threshold is set to 0.7. There is no restriction here. Implementers can set it according to the implementation scenario. If the noise probability index of the i-th data point is greater than 0.7, then the i-th data point is regarded as a noise point.

[0048] Similarly, the noise probability index of each data point included in the sliding window is obtained to get all the noise points in the sliding window. Then, the length of the sliding window is adjusted according to the number of noise points to obtain the final fitting window. The specific adjustment method is as follows: Obtain the total number of all data points contained in the sliding window, calculate the ratio between the number of noise points and the total number of all data points, and obtain the noise ratio of the sliding window; If the noise ratio is within the preset noise ratio range, then the sliding window is used as the final fitting window; If the noise ratio is less than the minimum value in the preset noise ratio range, then the length of the sliding window is reduced by the preset length value to obtain the final fitting window; If the noise ratio is greater than the maximum value in the preset noise ratio range, then the length of the sliding window is increased by the preset length value to obtain the final fitting window.

[0049] In one embodiment, the noise ratio of the sliding window is obtained by calculating the ratio between the number of noise points contained in the sliding window and the total number of data points contained in the sliding window. A preset noise ratio range is set to [0.1, 0.3], which is not limited here and can be adjusted by the implementer according to the implementation scenario. If the noise ratio of the sliding window is within [0.1, 0.3], that is, the noise ratio of the sliding window is greater than or equal to 0.1 and less than or equal to 0.3, the length of the sliding window is not adjusted, and the sliding window is used as the final fitting window. If the noise ratio of the sliding window is less than 0.1, it indicates that there are fewer noise points in the sliding window and the data points in the sliding window are relatively stable. In order to amplify the local features of the data points, the length of the sliding window needs to be reduced by 2 to obtain the final fitting window. If the noise ratio of the sliding window is greater than 0.3, it indicates that there are more noise points in the sliding window and fewer real data points. In order to increase the amount of real data, the length of the sliding window needs to be increased by 2 to obtain the final fitting window.

[0050] Step S103: Using the length of the final fitting window as the sliding step, repeat the steps of obtaining the final fitting window until the final fitting window contains the data point of the last sampling time in the scatter plot, thus obtaining all final fitting windows.

[0051] According to step S102, the first final fitting window in the scatter plot is obtained. Then, the length of the first final fitting window is used as the sliding step size. The sliding window is moved using the sliding step size, and the second final fitting window is obtained in the same way as the first final fitting window. Then, the length of the second final fitting window is used as the new sliding step size. The sliding window is moved using the new sliding step size, and the third final fitting window is obtained in the same way as the first final fitting window. This process continues until the final fitting window contains the data points at the last sampling time in the scatter plot. At this point, the traversal of the sliding window is completed, and multiple final fitting windows containing all data points in the scatter plot are obtained. These windows are used to fit the data points within each final fitting window, thereby improving the accuracy of the fitting.

[0052] Specifically, when the sliding window contains the last data point in the scatter plot, the length of the sliding window is not adjusted, and the sliding window is used as the final fitting window.

[0053] Step S104: Fit the data points contained in each of the final fitting windows to obtain the final fitting value for each sampling time contained in each of the final fitting windows, obtain the real-time cable sheath circulation current value, obtain the abnormality degree of the real-time cable sheath circulation current value based on the real-time cable sheath circulation current value and the final fitting value at the sampling time where the real-time cable sheath circulation current value is located, and provide a hidden danger warning for the distribution network cable based on the abnormality degree.

[0054] After obtaining all the final fitting windows, the least squares method can be used to perform polynomial fitting on the data points in each final fitting window to obtain the final fitting value for each sampling time within each final fitting window. The specific method for obtaining the final fitting value for each sampling time is as follows:

[0055] Noise points in each of the final fitting windows are removed, and data points in each of the final fitting windows after noise point removal are fitted to obtain the final fitted value for each sampling time contained in each of the final fitting windows.

[0056] It should be noted that since noise points within the final fitting window can increase the fitting error, the data points in each final fitting window after removing noise points are fitted to obtain a better fitting result.

[0057] Furthermore, after obtaining the final fitted value at each sampling time, this final fitted value can be used as a standard for anomaly detection to detect real-time anomalies in distribution network cables. The method for detecting real-time anomalies in distribution network cables is as follows: obtain the real-time cable sheath circulating current value; based on the final fitted value at the sampling time where the real-time cable sheath circulating current value is located, determine the degree of anomaly in the real-time cable sheath circulating current value; set a preset anomaly degree warning parameter of 1.5 (this is not limited and can be set by the implementer according to the implementation scenario). If the degree of anomaly in the real-time cable sheath circulating current value is greater than 1.5, an anomaly warning is issued for the distribution network cable.

[0058] The specific method for obtaining the degree of anomaly in the real-time cable sheath circulating current value is as follows: The absolute value of the difference between the final fitted value at the sampling time of the real-time cable sheath circulation current value and the real-time cable sheath circulation current value is calculated to obtain the real-time difference degree. Within the period of the real-time cable sheath circulation current value, for any sampling time before the sampling time of the real-time cable sheath circulation current value, calculate the absolute value of the difference between the cable sheath circulation current value at the sampling time and the final fitted value at the sampling time to obtain the historical difference degree. Obtain the historical difference degree corresponding to each sampling time before the sampling time of the real-time cable sheath circulation current value to obtain the average historical difference degree. The absolute value of the difference between the real-time difference level and the average historical difference level is calculated to obtain the degree of abnormality of the real-time cable sheath circulation value.

[0059] In one embodiment, assuming the real-time cable sheath circulating current value is at the j-th sampling time of the day, the formula for calculating the anomaly degree of the real-time cable sheath circulating current value is:

[0060] in, This indicates the degree of anomaly in the real-time cable sheath circulating current value, which is also the degree of anomaly at the j-th sampling time of the day. This represents the cable sheath circulating current value at the j-th sampling time of the day, which is also the real-time cable sheath circulating current value. This represents the final fitted value at the j-th sampling time, which is also the final fitted value at the sampling time where the real-time cable sheath circulation current value is located. This represents the cable sheath circulating current value at the t-th sampling time prior to the current sampling time. This represents the final fitted value at the t-th sampling time prior to the sampling time where the real-time cable sheath circulation current value is located. Represents the absolute value symbol.

[0061] It should be noted that, That is, the degree of real-time difference. That is, the degree of historical differences. This refers to the average historical difference. The greater the difference between the real-time difference and the average historical difference, the greater the probability of an anomaly occurring at the j-th sampling time of the day, meaning a greater probability of an anomaly in the real-time cable sheath circulation current value. The larger the value, the greater the degree of abnormality in the real-time cable sheath circulation current value.

[0062] It is worth noting that the focus of this invention is on how to adaptively adjust the size of a preset window based on the collected historical cable sheath circulation current data to obtain a final fitting window, and then fit the data in the final fitting window to obtain the degree of anomaly in the real-time cable sheath circulation current data. Using the degree of anomaly in real-time cable sheath circulation current data to provide early warning of potential hazards in distribution network cables is existing technology and will not be elaborated upon here.

[0063] In summary, this invention acquires the cable sheath circulation current value of the distribution network cable at each sampling time within a preset number of cycles over a historical period, obtaining a sheath circulation current data sequence corresponding to each cycle. This data sequence is then mapped onto the same scatter plot, where the horizontal axis represents the sampling time and the vertical axis represents the cable sheath circulation current value. A sliding window of a preset size is constructed within the scatter plot. The complexity of the sliding window is determined based on the density of data points within it. The length of the sliding window is adjusted according to this complexity and the difference between any two data points within the window, resulting in a final fitting window. The length of the final fitting window is used as the sliding step size. The steps of obtaining the final fitting window are repeated until the final fitting window contains the data point of the last sampling time in the scatter plot, thus obtaining all final fitting windows. The data points contained in each final fitting window are fitted to obtain the final fitted value for each sampling time within each final fitting window, thereby obtaining the real-time cable sheath circulation current value. Based on the real-time cable sheath circulation current value and the final fitted value of the sampling time in which the real-time cable sheath circulation current value is located, the degree of anomaly of the real-time cable sheath circulation current value is obtained. Based on the degree of anomaly, a potential hazard warning is issued for the distribution network cable. Specifically, a sliding window of a preset size is constructed from historical sheath circulation current data. The length of the sliding window is adjusted according to the complexity of the data points in the sliding window to obtain the final fitting window. This increases the number of data points in the sliding window with lower complexity and reduces the fitting error caused by noise points in the sliding window with higher complexity, resulting in a better fitting effect. The data points within the final fitting window are fitted, and the degree of anomaly of the real-time cable sheath circulation current value is calculated based on the fitting result, thereby issuing an anomaly warning for the distribution network cable and improving the accuracy of cable safety monitoring.

[0064] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A data analysis method for monitoring hidden dangers of distribution network cables, characterized in that: The data analysis method for monitoring hidden dangers of distribution network cables comprises: Obtain the cable sheath circulating current value of the distribution network cable at each sampling moment within a preset number of cycles in a historical period, obtain the sheath circulating current data sequence corresponding to each cycle, and map the sheath circulating current data sequence corresponding to each cycle to the same scatter plot, where the abscissa of the scatter plot is the sampling moment and the ordinate is the cable sheath circulating current value; Constructing a sliding window of a preset size in the scatter plot, obtaining the complexity of the sliding window according to the density of data points in the sliding window, and adjusting the length of the sliding window according to the complexity and the difference between every two data points in the sliding window to obtain a final fitting window; The length of the final fitting window is used as a sliding step length, and the step of obtaining the final fitting window is repeated until the final fitting window contains the data point at the last sampling time in the scatter plot, thereby obtaining all the final fitting windows; Fit the data points contained in each of the final fitting windows to obtain the final fitting value at each sampling moment contained in each of the final fitting windows, obtain the real-time cable sheath circulating current value, and obtain the abnormality degree of the real-time cable sheath circulating current value according to the real-time cable sheath circulating current value and the final fitting value at the sampling moment of the real-time cable sheath circulating current value, and issue a hidden danger warning for the distribution network cable according to the abnormality degree.

2. A data analysis method for monitoring hidden dangers of distribution network cables according to claim 1, characterized in that: The obtaining the complexity of the sliding window according to the density of the data points in the sliding window includes: In the sliding window, construct a minimum convex polygon including all data points in the sliding window, and calculate the ratio between the area of ​​the minimum convex polygon and the area of ​​the sliding window to obtain a first ratio; Obtaining the number of all data points contained in the sliding window, calculating the ratio between the number of all data points and the area of ​​the minimum convex polygon, and obtaining a second ratio; The first ratio and the second ratio are weightedly summed to obtain the complexity of the sliding window.

3. A data analysis method for monitoring hidden dangers of distribution network cables according to claim 1, characterized in that: The step of adjusting the length of the sliding window according to the complexity and the difference between every two data points in the sliding window to obtain a final fitting window comprises: If the complexity is less than or equal to a preset complexity threshold, the length of the sliding window is increased by a preset length value to obtain a final fitting window; If the complexity is greater than the preset complexity threshold, all data points in the sliding window are fitted to obtain fitting points corresponding to each data point contained in the sliding window. According to the difference between every two fitting points in the sliding window, the length of the sliding window is adjusted to obtain a final fitting window.

4. A data analysis method for monitoring hidden dangers of distribution network cables according to claim 3, characterized in that: The step of adjusting the length of the sliding window according to the difference between every two fitting points in the sliding window to obtain a final fitting window comprises: For any data point in the sliding window, obtain the left adjacent fitting point and the right adjacent fitting point of the fitting point corresponding to the data point, calculate the absolute value of the slope between the fitting point corresponding to the data point and the left adjacent fitting point to obtain a first absolute value slope, calculate the absolute value of the slope between the fitting point corresponding to the data point and the right adjacent fitting point to obtain a second absolute value slope, and calculate the absolute value of the slope between the left adjacent fitting point and the right adjacent fitting point to obtain a third absolute value slope; Among the first absolute value slope and the second absolute value slope, select the maximum value as the maximum slope value, calculate the absolute value of the difference between the maximum slope value and the third absolute value slope, normalize the absolute value of the difference, and obtain the fitting difference degree of the data point; The fitting difference degree of each data point in the sliding window is obtained, and the noise points contained in the sliding window are obtained according to the fitting difference degree of each data point and the fitting point corresponding to each data point. The length of the sliding window is adjusted according to the number of the noise points to obtain a final fitting window.

5. A data analysis method for monitoring hidden dangers of distribution network cables according to claim 4, characterized in that: The obtaining of the noise points contained in the sliding window according to the fitting difference degree of each of the data points and the fitting point corresponding to each of the data points comprises: Calculating the absolute value of the difference between each data point in the sliding window and the corresponding fitting point to obtain a fitting error, and normalizing the fitting error corresponding to each data point in the sliding window to obtain a corresponding normalized fitting error value; For any data point in the sliding window, a weighted sum is performed on the normalized value of the fitting error of the data point and the degree of fitting difference of the data point to obtain a noise possibility index of the data point. If the noise possibility index of the data point is greater than a preset noise possibility threshold, the data point is regarded as a noise point.

6. A data analysis method for monitoring hidden dangers of distribution network cables according to claim 4, characterized in that: The step of adjusting the length of the sliding window according to the number of noise points to obtain a final fitting window includes: Obtaining the number of all data points contained in the sliding window, calculating the ratio between the number of the noise points and the number of all data points, and obtaining the noise ratio of the sliding window; If the noise ratio is within a preset noise ratio range, taking the sliding window as a final fitting window; If the noise ratio is less than the minimum value in the preset noise ratio range, reducing the length of the sliding window by the preset length value to obtain a final fitting window; If the noise ratio is greater than the maximum value in the preset noise ratio range, the length of the sliding window is increased by the preset length value to obtain a final fitting window.

7. A data analysis method for monitoring hidden dangers of distribution network cables according to claim 4, characterized in that: The step of fitting the data points contained in each of the final fitting windows to obtain a final fitting value at each sampling moment contained in each of the final fitting windows comprises: The noise points in each of the final fitting windows are removed, and the data points in each of the final fitting windows after the noise points are removed are fitted to obtain the final fitting value at each sampling moment contained in each of the final fitting windows.

8. The data analysis method for monitoring hidden dangers of distribution network cables according to claim 1 is characterized in that: The method of obtaining the abnormality degree of the real-time cable sheath circulating current value according to the real-time cable sheath circulating current value and the final fitting value at the sampling time of the real-time cable sheath circulating current value comprises: Calculate the absolute value of the difference between the final fitting value at the sampling time of the real-time cable sheath circulating current value and the real-time cable sheath circulating current value to obtain the real-time difference degree; In the cycle where the real-time cable sheath circulating current value is located, for any sampling moment before the sampling moment where the real-time cable sheath circulating current value is located, calculate the absolute value of the difference between the cable sheath circulating current value at the sampling moment and the final fitting value at the sampling moment to obtain the historical difference degree, obtain the historical difference degree corresponding to each sampling moment before the sampling moment where the real-time cable sheath circulating current value is located, and obtain the average historical difference degree; The absolute value of the difference between the real-time difference degree and the average historical difference degree is calculated to obtain the abnormality degree of the real-time cable sheath circulating current value.

9. A data analysis method for monitoring hidden dangers of distribution network cables according to claim 1, characterized in that: The step of providing a hidden danger warning to the distribution network cable according to the abnormality degree includes: If the abnormality level is greater than a preset abnormality level warning parameter, an abnormality warning is issued for the distribution network cable.

Citation Information

Patent Citations

  • Power data efficient processing method and system based on digital twinning

    CN117057517A

  • Systems and methods for detecting connection anomalies

    US20230139081A1

Cited By

  • Four-fusion integrated terminal fault detection system based on sensor analysis

    CN119619827A