Electric transmission line electrical parameter analysis and verification method based on big data

Abstract

The invention relates to the technical field of data processing, in particular to a power transmission line electrical parameter analysis and verification method based on big data, and the method comprises the steps: carrying out the clustering analysis of standard data, dividing a plurality of clustering clusters, calculating a differentiation coefficient according to the difference quantity of feature parameters between data points in the clustering clusters, determining whether the cluster meets a consistency standard or not according to the differentiation coefficient; if the clustering clusters conform to the consistency standard, features of data points in the clustering clusters are identified, a difference representation coefficient is calculated according to the features to determine whether the clustering clusters conform to the difference standard or not, and the clustering clusters conforming to the difference standard are subjected to priority ranking; selecting the power transmission line corresponding to any maximum difference data point in the plurality of high-priority clustering clusters to perform key analysis and verification, and marking; according to the method, key analysis circuits and comprehensive analysis circuits are rapidly screened out through clustering analysis, and the efficiency and accuracy of analysis and verification of the electrical parameters of the power transmission line are improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method for analyzing and verifying electrical parameters of a power transmission line based on big data. Background Art

[0002] In recent years, the development of big data technology has provided new ideas and methods for the electrical parameter analysis and verification of transmission lines. By collecting the electrical parameter data and environmental parameter data of the transmission lines and using big data analysis technology to process and analyze these data, the operating status of the transmission lines can be understood more comprehensively and accurately. However, in practical applications, there are still some problems that need to be solved, such as how to extract effective features from massive data to improve the accuracy of cluster analysis. Therefore, a transmission line electrical parameter analysis and verification method based on big data is studied. By preprocessing, clustering analysis, analysis and processing of electrical parameter data and environmental parameter data, the accuracy and reliability of transmission line electrical parameter analysis and verification are improved, which is of great significance for ensuring the safe operation of transmission lines and improving the stability and efficiency of power systems.

[0003] For example, Chinese patent application publication number: CN118731582A discloses a method, device and transmission line monitoring system for locating transmission line faults, the method comprising: obtaining a vibration signal of the transmission line; determining the vibration amplitude and frequency of the transmission line based on the vibration signal; determining that the transmission line vibration is abnormal when the vibration amplitude and / or frequency are not within the corresponding predetermined range; obtaining an electrical transmission parameter signal of the transmission line with abnormal vibration; determining whether the transmission line has electrical parameter abnormalities based on the electrical transmission parameter signal, the electrical parameter abnormalities including harmonic abnormalities, transient overcurrent abnormalities and connection hardware failures; calculating the correlation metric value of the abnormal data corresponding to the vibration abnormality and the abnormal data corresponding to the electrical parameter abnormality; when the correlation metric value is greater than a predetermined threshold, determining the fault location range based on the collection locations of the vibration signal and the electrical transmission parameter signal, thereby solving the problem of low transmission line fault location accuracy in the prior art.

[0004] However, the existing technology has the problem that the cluster analysis of the electrical parameter data and environmental parameter data of the transmission line is not accurate enough, resulting in an inability to accurately distinguish the normal state and abnormal state of the transmission line, which in turn leads to deviations in subsequent analysis and verification. Summary of the Invention

[0005] To this end, the present invention provides a method for analyzing and verifying the electrical parameters of transmission lines based on big data, so as to overcome the problem in the prior art that the cluster analysis of the electrical parameter data and environmental parameter data of the transmission lines is not accurate enough, resulting in the inability to accurately distinguish the normal and abnormal states of the transmission lines, and thus causing deviations in subsequent analysis and verification.

[0006] To achieve the above objectives, the present invention provides a method for analyzing and verifying electrical parameters of transmission lines based on big data, comprising:

[0007] Step S1, collecting electrical parameter data and environmental parameter data of the transmission line, and preprocessing the electrical parameter data and environmental parameter data to obtain standard data;

[0008] Step S2, performing cluster analysis on the standard data to divide it into a number of clusters, including extracting electrical parameter characteristics and environmental parameter characteristics from the standard data, comparing the differences in the characteristics of each data point, and combining the data points according to the differences to obtain clusters, wherein the clusters must meet predetermined division criteria;

[0009] Step S3, calculating a differentiation coefficient based on the difference in characteristic parameters between each data point in the cluster, and determining whether the cluster meets the consistency standard based on the differentiation coefficient;

[0010] Step S4: selecting an analysis and processing method for each cluster, including: if the cluster meets the consistency standard, identifying the characteristics of the data points in the cluster, calculating the difference characterization coefficient based on the characteristics to determine whether the cluster meets the difference standard, prioritizing the clusters that meet the difference standard, selecting the transmission lines corresponding to any of the largest difference data points in several clusters in descending order of priority for key analysis and verification, and marking them; if the cluster does not meet the difference standard, performing a comprehensive analysis and verification on the transmission lines corresponding to all data points in the cluster;

[0011] In step S5, the results of the focused analysis and verification and the comprehensive analysis and verification are classified separately, a list of abnormal problems and a list of normal status are established, the problem characteristics in the consistency standard clustering clusters and the inconsistency standard clustering clusters are compared, the accuracy of the previous clustering analysis is determined, and the clustering analysis weight coefficient is re-determined under the condition that the accuracy is less than the preset accuracy.

[0012] Furthermore, preprocessing the electrical parameter data and the environmental parameter data includes removing outliers in the data using the Laida criterion, filling missing data by linear interpolation, and mapping the data to the [0,1] interval using the maximum-minimum normalization method.

[0013] Furthermore, in step S2, the predetermined division criteria include that the average Euclidean distance of data points within a cluster is less than a first preset threshold and the Euclidean distance between any two cluster centers is greater than a second preset threshold.

[0014] Furthermore, in step S3, calculating the differentiation coefficient according to the difference in characteristic parameters between the data points in the clusters includes:

[0015] Calculate the Euclidean distance between any two data points in the cluster to form a distance matrix;

[0016] Calculate the average value of the distance matrix and use it as the average difference between the data points in the cluster;

[0017] Calculate the Euclidean distance from each data point in the cluster to the cluster center to form a center distance vector;

[0018] Calculate the standard deviation of the center distance vector and use it as an indicator of the dispersion of data points around the cluster center;

[0019] The differentiation coefficient is obtained by weighted summing the average difference and the dispersion index.

[0020] Furthermore, in step S3, determining whether the cluster meets the consistency standard according to the differentiation coefficient includes:

[0021] If the differentiation coefficient is less than or equal to a preset differentiation coefficient, determining that the cluster meets the consistency standard;

[0022] If the differentiation coefficient is greater than a preset differentiation coefficient, it is determined that the cluster does not meet the consistency standard.

[0023] Furthermore, the preset differentiation coefficient is determined according to an average value of differentiation coefficients of clusters that have completed cluster analysis for the same transmission line.

[0024] Furthermore, in step S4, calculating the difference characterization coefficient according to the features includes:

[0025] Calculate the difference between the key feature parameters of each data point and the corresponding feature parameters of the cluster center to obtain the difference vector;

[0026] Normalize the difference vector and calculate the norm of the normalized difference vector, which is used as a measure of the difference between the data point and the cluster center;

[0027] Calculate the average value of the difference degree measure of all data points in the cluster and use it as the difference representation coefficient of the cluster.

[0028] Furthermore, in step S4, determining whether the cluster meets the difference standard includes that the difference characterization coefficient of the cluster is greater than or equal to a preset difference characterization coefficient threshold.

[0029] Furthermore, in step S4, prioritizing the clusters that meet the difference criteria includes:

[0030] Calculate the priority score of each cluster, which is obtained by weighted summation of the following factors:

[0031] The size of the difference characterization coefficient;

[0032] The abnormal degree of electrical parameters of data points within the cluster;

[0033] The degree of influence of environmental parameters on data points within the cluster;

[0034] The clusters are sorted in descending order according to their priority scores, with the clusters with higher priority scores at the front.

[0035] Furthermore, in step S5, the accuracy of the preliminary cluster analysis is determined based on the ratio of the number of data points in the consistency standard cluster cluster that are correctly classified into the normal state list to the total number of data points in the cluster cluster, the ratio of the number of data points in the consistency standard cluster cluster that are incorrectly classified into the abnormal problem list to the total number of data points in the cluster cluster, the ratio of the number of data points in the inconsistency standard cluster cluster that are correctly classified into the abnormal problem list to the total number of data points in the cluster cluster, and the ratio of the number of data points in the inconsistency standard cluster cluster that are incorrectly classified into the normal state list to the total number of data points in the cluster cluster.

[0036] Compared with the prior art, the beneficial effect of the present invention lies in that, by performing statistical analysis on the data characteristics of electrical parameters and environmental parameters and setting reasonable first preset thresholds and second preset thresholds, the present invention can make the data points within the cluster more similar in comprehensive characteristics and the differences between different clusters more obvious, which helps to accurately divide the transmission line data into appropriate clusters, improve the accuracy and rationality of clustering, thereby obtaining more valuable clustering results and better reflecting the operating status of the transmission line.

[0037] Furthermore, the present invention calculates the differentiation coefficient by combining the two dimensions of average difference and dispersion index, which can comprehensively and comprehensively evaluate the consistency of data points within the cluster. The average difference reflects the overall degree of difference between data points in the cluster, while the dispersion index reflects the tightness of the distribution of data points around the cluster center. The differentiation coefficient obtained by weighted summation of the two avoids the limitations of single indicator evaluation and more accurately reflects the quality of the cluster. The above method improves the accuracy of cluster analysis of electrical parameter data and environmental parameter data of transmission lines to accurately distinguish between normal and abnormal states of transmission lines and thus screen out key analysis areas.

[0038] Furthermore, the present invention can accurately quantify the degree of difference of each data point relative to the cluster center by calculating the difference vector between each data point and the cluster center, and performing normalization and norm calculation. This method not only takes into account the absolute differences of key characteristic parameters (such as voltage, current, and power), but also eliminates the influence of different parameter dimensions through normalization processing, making the difference measurement more objective and comparable. The average value of the difference degree measurement of all data points in the cluster is taken as the difference characterization coefficient, which can fully reflect the data distribution characteristics of the entire cluster. The higher the difference characterization coefficient, the greater the degree of deviation of the data points in the cluster from the center and the stronger the discreteness of the data; otherwise, it means that the data points are relatively concentrated and the clustering effect is better. In the analysis and verification of transmission lines, clusters that meet the difference standards often indicate the existence of potential problems or abnormal conditions. By prioritizing and focusing on analyzing these clusters, resource allocation can be effectively optimized, and limited manpower and material resources can be concentrated on the transmission lines that need the most attention, thereby improving the efficiency of fault detection and processing.

[0039] Furthermore, the present invention achieves a multi-dimensional comprehensive evaluation of clusters by weighted summing the three factors of difference characterization coefficient, electrical parameter abnormality degree and environmental parameter influence degree. This method not only considers the degree of difference between the data point and the cluster center, but also combines the two key factors of electrical parameter abnormality and environmental parameter influence, making the evaluation results more comprehensive and objective, and can more accurately reflect the potential risks of transmission lines. By calculating the priority score and sorting it in descending order, high-risk clusters that need to be focused on can be accurately identified. The degree of electrical parameter abnormality directly reflects whether the current operating status of the transmission line deviates from the normal range. The degree of environmental parameter influence considers the potential threat of the external environment to the line, and the difference characterization coefficient reveals the stability of the cluster from the perspective of data distribution. The combination of the three helps to discover potential hidden dangers that may cause failures in advance. Through the above method, the transmission line with the largest degree of abnormality can be selected in the key analysis cluster for subsequent analysis, thereby improving the accuracy of cluster analysis of the electrical parameter data and environmental parameter data of the transmission line, as well as the accuracy and efficiency of subsequent transmission line electrical parameter analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a workflow diagram of a method for analyzing and verifying electrical parameters of a power transmission line based on big data according to an embodiment of the present invention;

[0041] Figure 2 This is a flowchart of the process of calculating the differentiation coefficient in the big data-based transmission line electrical parameter analysis and verification method according to an embodiment of the present invention;

[0042] Figure 3 This is a flowchart of the process of calculating the difference characterization coefficient in the big data-based transmission line electrical parameter analysis and verification method according to an embodiment of the present invention;

[0043] Figure 4 This is a workflow diagram for determining whether clusters meet consistency standards in a big data-based transmission line electrical parameter analysis and verification method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention.

[0045] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0046] See also Figure 1-Figure 4 As shown, Figure 1 This is a workflow diagram of a method for analyzing and verifying electrical parameters of a power transmission line based on big data according to an embodiment of the present invention; Figure 2 This is a flowchart of the process of calculating the differentiation coefficient in the big data-based transmission line electrical parameter analysis and verification method according to an embodiment of the present invention; Figure 3 This is a flowchart of the process of calculating the difference characterization coefficient in the big data-based transmission line electrical parameter analysis and verification method according to an embodiment of the present invention; Figure 4 This is a workflow diagram for determining whether clusters meet consistency standards in a big data-based transmission line electrical parameter analysis and verification method according to an embodiment of the present invention.

[0047] The embodiment of the present invention provides a method for analyzing and verifying electrical parameters of a power transmission line based on big data, including:

[0048] Step S1, collecting electrical parameter data and environmental parameter data of the transmission line, and preprocessing the electrical parameter data and environmental parameter data to obtain standard data;

[0049] Step S2, performing cluster analysis on the standard data to divide it into a number of clusters, including extracting electrical parameter characteristics and environmental parameter characteristics from the standard data, comparing the differences in the characteristics of each data point, and combining the data points according to the differences to obtain clusters, wherein the clusters must meet predetermined division criteria;

[0050] Step S3, calculating a differentiation coefficient based on the difference in characteristic parameters between each data point in the cluster, and determining whether the cluster meets the consistency standard based on the differentiation coefficient;

[0051] Step S4: selecting an analysis and processing method for each cluster, including: if the cluster meets the consistency standard, identifying the characteristics of the data points in the cluster, calculating the difference characterization coefficient based on the characteristics to determine whether the cluster meets the difference standard, prioritizing the clusters that meet the difference standard, selecting the transmission lines corresponding to any of the largest difference data points in several clusters in descending order of priority for key analysis and verification, and marking them; if the cluster does not meet the difference standard, performing a comprehensive analysis and verification on the transmission lines corresponding to all data points in the cluster;

[0052] In step S5, the results of the focused analysis and verification and the comprehensive analysis and verification are classified separately, a list of abnormal problems and a list of normal status are established, the problem characteristics in the consistency standard clustering clusters and the inconsistency standard clustering clusters are compared, the accuracy of the previous clustering analysis is determined, and the clustering analysis weight coefficient is re-determined under the condition that the accuracy is less than the preset accuracy.

[0053] In the embodiment of the present invention, the electrical parameter data includes but is not limited to "voltage, current and power (active power and reactive power) of the transmission line", and the environmental parameter data includes but is not limited to "temperature, humidity and air pressure". The preprocessing of the electrical parameter data and environmental parameter data includes using the Laida criterion to remove outliers in the data, filling missing data through linear interpolation, and mapping the data to the [0,1] interval using the maximum-minimum normalization method.

[0054] In the embodiment of the present invention, the key analysis and verification includes detailed inspection of the electrical parameters of the transmission line corresponding to the maximum difference data point in the clusters that meet the difference criteria, such as checking the stability and accuracy of parameters such as voltage, current, and power; evaluating the equipment condition of the transmission line, including changes in electrical parameters such as line resistance, inductance, and capacitance, as well as problems such as aging and damage of equipment; analyzing the impact of environmental parameters on the transmission line, such as the impact of environmental factors such as temperature, humidity, and wind speed on line insulation performance, conductor resistance and other electrical parameters; and checking the operation records of the transmission line, including historical fault records, maintenance records, etc. To understand the operating status and potential problems of the line; comprehensive analysis and verification includes comprehensive testing of the electrical parameters of the transmission line corresponding to all data points in the clusters that do not meet the difference standards, covering parameters such as voltage, current, power, resistance, inductance, and capacitance; comprehensive inspection of the equipment of the transmission line, and evaluation of the operating environment of the transmission line, including the impact of long-term changes in environmental parameters on the line, and the interference of the surrounding environment on the line; in-depth analysis of the operating data of the transmission line, including comparison of operating data in different time periods, and comparison with the operating data of other similar transmission lines, to discover potential problems and patterns.

[0055] Specifically, in step S2, the predetermined division criteria include that the average Euclidean distance of data points within a cluster is less than a first preset threshold and the Euclidean distance between any two cluster centers is greater than a second preset threshold.

[0056] In the embodiment of the present invention, the first preset threshold value and the second preset threshold value can be determined by the following method. First, the collected transmission line electrical parameter data and environmental parameter data are statistically analyzed to understand the distribution range, discreteness and other characteristics of the data. For example, the value range and fluctuation of parameters such as voltage, current, power, temperature, humidity, and air pressure are analyzed. The purpose of clustering is to divide the data into clusters with similar characteristics in order to better analyze the operating status of the transmission line. Therefore, the setting of the threshold value should be able to make the divided clusters have obvious distinction in terms of electrical parameters and environmental parameters, while ensuring the uniformity of the data within the cluster. Consistency; through multiple experiments, try different threshold combinations, observe the quality of clustering results, and use some evaluation indicators, such as cluster compactness (the closeness of data points within a cluster) and separation (the degree of separation between different clusters), to determine the optimal threshold range; for example, suppose that after analyzing a large amount of data, the following data features are obtained: in electrical parameters, the voltage range is [100,220]kV, the current range is [50,500]A, the active power range is [10,1000]MW, and the reactive power range is [-500,500]Mvar; in environmental parameters, the temperature range is [ -20,40]℃, humidity range is [20%, 80%], and air pressure range is [90,110]kPa. After many experiments, it was found that when the first preset threshold is set to 0.5 and the second preset threshold is set to 1.5, a relatively ideal clustering result can be obtained; at this time, the average Euclidean distance of the data points in the cluster is less than 0.5, which means that the data points in the cluster are relatively similar in the comprehensive characteristics of electrical parameters and environmental parameters. For example, in a cluster, the differences in voltage, current, power, temperature, humidity, air pressure and other parameters of each data point are small, indicating that the transmission lines corresponding to these data points are operating at the same level. Under similar conditions; the Euclidean distance between the centers of any two clusters is greater than 1.5: it ensures that there are obvious differences between different clusters. For example, one cluster may represent a transmission line operating under a high temperature, high humidity environment and a high load, while another cluster may represent a transmission line operating under a low temperature, low humidity environment and a low load. Through such threshold settings, transmission lines in different operating states can be effectively distinguished. The first preset threshold is preferably 0.5, and the second preset threshold is preferably 1.5, but the above values ​​are not limited thereto. Those skilled in the art can also adjust the values ​​of the corresponding preset thresholds according to actual needs.

[0057] By performing statistical analysis on the data features of electrical parameters and environmental parameters and setting reasonable first and second preset thresholds, the present invention can make the data points within the clusters more similar in comprehensive features and make the differences between different clusters more obvious. This helps to accurately divide the transmission line data into appropriate clusters, improve the accuracy and rationality of clustering, thereby obtaining more valuable clustering results and better reflecting the operating status of the transmission line.

[0058] Specifically, in step S3, the step of calculating the differentiation coefficient according to the difference in characteristic parameters between the data points in the cluster includes:

[0059] Step S3301, calculating the Euclidean distance between any two data points in the cluster to form a distance matrix;

[0060] Step S3302, calculating the average value of the distance matrix and using it as the average difference value of the data points in the cluster;

[0061] Step S3303, calculating the Euclidean distance from each data point in the cluster to the cluster center to form a center distance vector;

[0062] Step S3304, calculating the standard deviation of the center distance vector and using it as an indicator of the degree of dispersion of data points around the cluster center;

[0063] Step S3305: Perform weighted summation on the average difference and the dispersion index to obtain a differentiation coefficient.

[0064] In the embodiment of the present invention, it is assumed that there is a cluster containing 5 data points, each data point has 3 features (here taking the voltage, current and power in the electrical parameters as an example), and it is assumed that there is a cluster C, in which the data points are: D1 = [100, 200, 300] (indicating that the voltage is 100, the current is 200, the power is 300, and the following are similar); D2 = [120, 220, 320]; D3 = [90, 180, 280]; D4 = [110, 210, 310]; D5 = [105, 205, 305]; the Euclidean distance formula is used to calculate the Euclidean distance between any two data points in the cluster , forming a distance matrix, and calculating other elements in the same way to obtain the distance matrix. The sum of all elements of the distance matrix is ​​348.08, and the number of distance matrix elements is 25, then the average difference is 13.9232. First calculate the cluster center D0 = [105, 205, 305], calculate the distance from other data points to the cluster center, and obtain the center distance vector V = [8.66, 15.81, 22.36, 14.14, 0]. The average value of the center distance vector is 12.194, and the standard deviation is 7.49. Assuming that the weight of the average difference α = 0.4 and the weight of the dispersion index β = 0.6, the differentiation coefficient is 10.06328.

[0065] The weight of the average difference amount and the weight of the dispersion index in the embodiment of the present invention can be determined by the following method: multiple cluster analyses are performed using historical data, and different weight combinations are tried. The degree of matching between the clustering results and the actual situation is observed, and the weight combination that makes the clustering results most consistent with the actual situation is selected. For example, the clustering results can be compared with known transmission line fault records or actual operating status, and indicators such as accuracy and recall rate can be calculated. The weights are determined by optimizing these indicators. For example, historical data of 100 clusters are collected, and each cluster has a corresponding actual operating status label (normal or abnormal). First, the initial weights α=0.5 and β=0.5 are set, and cluster analysis is performed on these 100 clusters and the differentiation coefficient is calculated. , the state of the cluster is judged according to the differentiation coefficient, and then compared with the actual label, and the accuracy is 70%. Then try to adjust the weight to α=0.6, β=0.4, and analyze again. The accuracy is improved to 75%. Continue to adjust the weight to α=0.45, β=0.55, and the accuracy is 72%. After many attempts, it is found that when α=0.6, β=0.4, the accuracy is the highest. Therefore, it is determined that in this scenario, the weight of the average difference is 0.6, and the weight of the dispersion index is 0.4. The value of the weight of the average difference is preferably 0.6, and the value of the weight of the dispersion index is preferably 0.4, but the above values ​​are not limited to this. Those skilled in the art can also adjust the values ​​according to actual needs.

[0066] Specifically, in step S3, when determining whether the cluster meets the consistency standard, it is determined whether the cluster meets the consistency standard according to the comparison result of the differentiation coefficient and the preset differentiation coefficient;

[0067] When the differentiation coefficient is less than or equal to a preset differentiation coefficient, determining that the cluster meets the consistency standard;

[0068] When the differentiation coefficient is greater than a preset differentiation coefficient, it is determined that the cluster does not meet the consistency standard.

[0069] The preset differentiation coefficient in the embodiment of the present invention is the average value of the differentiation coefficients of the clusters that have completed cluster analysis for the same transmission line, but the above value is not limited thereto, and those skilled in the art can also adjust the value according to actual needs.

[0070] The present invention calculates the differentiation coefficient by combining the two dimensions of average difference and dispersion index, which can comprehensively and comprehensively evaluate the consistency of data points within the cluster. The average difference reflects the overall degree of difference between data points in the cluster, while the dispersion index reflects the distribution tightness of data points around the cluster center. The differentiation coefficient obtained by weighted summation of the two avoids the limitations of single indicator evaluation and more accurately reflects the quality of the cluster. The above method improves the accuracy of cluster analysis of electrical parameter data and environmental parameter data of transmission lines, so as to accurately distinguish the normal and abnormal states of transmission lines and thus screen out key analysis areas.

[0071] Specifically, in step S4, the step of calculating the difference characterization coefficient according to the features includes:

[0072] Step S4401, calculating the difference between the key feature parameters of each data point and the corresponding feature parameters of the cluster center to obtain a difference vector;

[0073] Step S4402: normalize the difference vector and calculate the norm of the normalized difference vector, which is used as a measure of the difference between the data point and the cluster center.

[0074] Step S4403 : Calculate the average value of the difference degree measure of all data points in the cluster, and use it as the difference representation coefficient of the cluster.

[0075] The key characteristic parameters of each data point in the embodiment of the present invention include but are not limited to "voltage, current and power". Assume that we have a cluster C containing 4 data points, each of which has 3 key characteristic parameters (here we still take voltage, current and power among electrical parameters as an example). The specific data are as follows: D1 = [110, 220, 330]; D2 = [120, 230, 340]; D3 = [100, 210, 320]; D4 = [130, 240, 350];

[0076] Calculate the cluster center D0, take the average value of each characteristic parameter to get D0 = [115, 225, 335]; calculate the difference between the key characteristic parameter of each data point and the corresponding characteristic parameter of the cluster center, and get the difference vector:

[0077] The difference vector between D1 and D0 is V1 = [110-115, 220-225, 330-335] = [-5, -5, -5];

[0078] The difference vector between D2 and D0 is V2 = [120-115, 230-225, 340-335] = [5, 5, 5];

[0079] The difference vector between D3 and D0 is V3 = [100-115, 210-225, 320-335] = [-15, -15, -15];

[0080] The difference vector between D4 and D0 is V4 = [130-115, 240-225, 350-335] = [15, 15, 15]; the difference vector is normalized and the norm of the normalized difference vector is calculated as a measure of the degree of difference between the data point and the cluster center. For V1, its norm is The normalized vector is The difference degree measure of this data point is 1. Similarly, the difference degree measure of V2 is 1, the difference degree measure of V3 is 1, and the difference degree measure of V4 is 1. Calculate the average value of the difference degree measure of all data points in the cluster and use it as the difference representation coefficient of the cluster. The difference representation coefficient = (1+1+1+1) / 4 = 1.

[0081] Specifically, in step S4, when determining whether the cluster meets the difference standard, it is determined whether the cluster meets the difference standard based on the comparison result of the difference characterization coefficient of the cluster and the preset difference characterization coefficient threshold;

[0082] When the difference characterization coefficient of the cluster is greater than or equal to a preset difference characterization coefficient threshold, it is determined that the cluster meets the difference standard;

[0083] When the difference characterization coefficient of the cluster is less than a preset difference characterization coefficient threshold, it is determined that the cluster does not meet the difference standard.

[0084] The preset difference characterization coefficient threshold in the embodiment of the present invention is the average value of the difference characterization coefficient of each cluster cluster obtained by clustering analysis of several identical transmission lines under different operating conditions, but the above value is not limited thereto, and those skilled in the art can also adjust the value according to actual needs.

[0085] The present invention calculates the difference vector between each data point and the cluster center, and performs normalization and norm calculation, so as to accurately quantify the degree of difference of each data point relative to the cluster center. This method not only takes into account the absolute difference of each key characteristic parameter (such as voltage, current, and power), but also eliminates the influence of different parameter dimensions through normalization processing, making the difference measurement more objective and comparable. The average value of the difference degree measurement of all data points in the cluster is taken as the difference characterization coefficient, which can fully reflect the data distribution characteristics of the entire cluster. The higher the difference characterization coefficient, the greater the degree of deviation of the data points in the cluster from the center and the stronger the discreteness of the data; otherwise, it means that the data points are relatively concentrated and the clustering effect is better. In the analysis and verification of transmission lines, clusters that meet the difference standards often indicate the existence of potential problems or abnormal conditions. By prioritizing these clusters and focusing on analysis, resource allocation can be effectively optimized, and limited manpower and material resources can be concentrated on the transmission lines that need the most attention, thereby improving the efficiency of fault detection and processing.

[0086] Specifically, in step S4, the priority ranking of the clusters that meet the difference criteria includes:

[0087] Calculate the priority score of each cluster, which is obtained by weighted summation of the following factors:

[0088] The size of the difference characterization coefficient;

[0089] The abnormal degree of electrical parameters of data points within the cluster;

[0090] The degree of influence of environmental parameters on data points within the cluster;

[0091] The clusters are sorted in descending order according to their priority scores, with the clusters with higher priority scores at the front.

[0092] In the embodiment of the present invention, the degree of abnormality of the electrical parameters of the data points in the cluster can be determined by comparing with the standard values. For example, for each electrical parameter of the transmission line, such as voltage, current, power, etc., there is a corresponding rated value or normal operating range. The electrical parameters of each data point in the cluster are compared with these standard values. For example, for a transmission line with a rated voltage of 220kV, if the voltage value of a data point deviates from 220kV by a large margin and exceeds the normal fluctuation range (such as ±10%), it means that the voltage parameter of the data point is abnormal. By calculating the degree of deviation of the electrical parameter of each data point from the standard value and comprehensively evaluating the deviation of each electrical parameter, the degree of abnormality of the electrical parameter of the data point is obtained. For the entire cluster, the average value or weighted average value of the abnormality of the electrical parameters of all data points can be calculated to represent the degree of abnormality of the electrical parameters of the data points in the cluster. The degree of influence of environmental parameters on the data points in the cluster can be determined by establishing a model for the influence of environmental parameters (such as temperature, humidity and wind speed) on the electrical parameters. For example, an increase in temperature will cause an increase in the resistance of the wire. The influence of environmental parameters on electrical parameters is analyzed based on these impact models. For each data point, the impact index of environmental parameters on electrical parameters is calculated based on its environmental parameter value. For the cluster, the average or weighted average of the environmental parameter impact indexes of all data points can be calculated to represent the impact of environmental parameters on the data points in the cluster. For example, taking the impact of temperature on wire resistance as an example, the relationship between wire resistance R and temperature T is known to be R=R0(1+α(T-T0)) (R0 is the resistance at reference temperature T0, α is the resistance temperature coefficient). The temperature of a data point is Ti. The impact of temperature change on resistance is calculated based on this formula, which in turn affects parameters such as current and power. The impact of the environmental parameter (temperature) on electrical parameters at the data point is obtained. Taking the influence of other environmental parameters into consideration, the impact of environmental parameters on each data point is obtained, and the impact of environmental parameters on the data points in the cluster is obtained.

[0093] In the embodiment of the present invention, the weight of the difference characterization coefficient, the weight of the degree of abnormality of the electrical parameters of the data points in the cluster, and the weight of the degree of influence of the environmental parameters of the data points in the cluster can be determined by the following method: collect a large amount of historical data of the transmission line, combine it with actual fault records, analyze the correlation between different factors and the occurrence of faults, and statistically analyze the contribution of each factor in the fault case to reversely infer its weight. For example, the proportion of faults directly caused by abnormal electrical parameters and the proportion of faults caused by environmental factors in historical faults are statistically analyzed, as well as the correspondence between abnormal difference characterization coefficients and faults, so as to determine the weight. For example, the transmission line data of a certain area in the past three years were analyzed, and a total of 200 fault events were recorded. According to statistics, 120 faults were directly caused by abnormal electrical parameters, 50 faults were caused by the long-term influence of environmental parameters leading to the decline of equipment performance, and 30 faults were caused by early warning of the difference characterization coefficient before obvious electrical abnormalities occurred. The weight of each factor is calculated based on this: the weight of the degree of electrical parameter abnormality β = 0.6; the weight of the degree of environmental parameter influence γ = 0.25; the weight of the size of the difference characterization coefficient α = 0.15, but the above values ​​are not limited to this, and technical personnel in this field can also adjust the value according to actual needs.

[0094] In an embodiment of the present invention, a regional power company collected electrical parameter data (voltage, current, power, etc.) and environmental parameter data (temperature, humidity, and wind speed) of the transmission lines over a period of time. After data preprocessing and cluster analysis, multiple clusters that meet the difference criteria are obtained. For each cluster, the difference characterization coefficient is calculated according to the method of steps S4401-S4403. For example, the difference characterization coefficient of cluster A is 0.8, and the difference characterization coefficient of cluster B is 0.6. For each data point in the cluster, its electrical parameters are compared with the standard value, and the degree of deviation is calculated. For example, there are 5 data points in cluster A, and the degree of deviation of the voltage, current, and power of each data point from the standard value is calculated respectively. Assume that the voltage deviation of data point 1 is 15%, the current deviation is 10%, and the power deviation is 8%; the voltage deviation of data point 2 is 10%, the current deviation is 8%, and the power deviation is 6%; and so on, to calculate the abnormality of the electrical parameters of each data point, you can use a comprehensive deviation method, such as simply adding up the average value. For example, the abnormality of the electrical parameters of data point 1 is (15% + 10% + 8%) / 3 = 11%; calculate the average value of the abnormality of the electrical parameters of all data points in the cluster. The abnormality of the electrical parameters of the 5 data points in cluster A are 11%, 9%, 10%, 8%, and 12% respectively. Then the abnormality of the electrical parameters of cluster A is (11% + 9% + 10% + 8%). %+12%) / 5=10%; establish a model for the impact of environmental parameters on electrical parameters, such as the impact of temperature on wire resistance (R=R0[1+α×(T-T0)]) and the impact of humidity on the insulation performance of insulators; for each data point in the cluster, calculate the impact index of the environmental parameters on the electrical parameters based on its environmental parameter value. For example, if the temperature of data point 1 is 30°C, the reference temperature T0 is 25°C, the resistance temperature coefficient α is 0.004, and R0 is 10Ω, then the impact of temperature on resistance is R=10×[1+0.004×(30-25)]=10.2Ω, which in turn affects parameters such as current and power, and the degree of influence of the environmental parameter (temperature) of the data point on the electrical parameters is obtained.Considering the comprehensive influence of multiple environmental parameters, the degree of environmental parameter influence of each data point is calculated; the average value of the environmental parameter influence index of all data points in the cluster is calculated. The environmental parameter influence degrees of the five data points in cluster A are 10%, 8%, 9%, 7%, and 11%, respectively. The environmental parameter influence degree of cluster A is (10% + 8% + 9% + 7% + 11%) / 5 = 9%; based on the above statistical analysis, the size weight of the difference characterization coefficient α = 0.15 and the electrical parameter abnormality weight β of the data points in the cluster are determined to be 0.6. The weight of the degree of influence of environmental parameters of data points in the cluster is γ = 0.25; the priority score of each cluster is calculated, the priority score of cluster A = 0.15×0.8+0.6×10%+0.25×9%=0.12+0.06+0.0225=0.2025; the priority score of cluster B can be calculated in the same way; the clusters are sorted in descending order according to the priority score, and the clusters with higher priority scores are placed in front. For example, if the priority score of cluster A is higher than that of cluster B, then cluster A is placed before cluster B.

[0095] The present invention achieves a multi-dimensional comprehensive evaluation of clusters by weighted summing three factors: difference characterization coefficient, degree of electrical parameter anomaly, and degree of environmental parameter influence. This method not only considers the degree of difference between the data point and the cluster center, but also combines two key factors: electrical parameter anomaly and environmental parameter influence, making the evaluation results more comprehensive and objective, and able to more accurately reflect the potential risks of transmission lines. By calculating priority scores and sorting them in descending order, high-risk clusters that need to be focused on can be accurately identified. The degree of electrical parameter anomaly directly reflects whether the current operating status of the transmission line deviates from the normal range. The degree of environmental parameter influence considers the potential threat of the external environment to the line, and the difference characterization coefficient reveals the stability of the cluster from the perspective of data distribution. The combination of the three helps to discover potential hidden dangers that may lead to failures in advance. Through the above method, the transmission line with the largest degree of anomaly can be selected in the key analysis cluster for subsequent analysis, thereby improving the accuracy of cluster analysis of the electrical parameter data and environmental parameter data of the transmission line, as well as the accuracy and efficiency of subsequent transmission line electrical parameter analysis.

[0096] Specifically, in step S5, the accuracy of the preliminary cluster analysis is determined based on the ratio of the number of data points in the consistency standard cluster cluster that are correctly classified into the normal state list to the total number of data points in the cluster cluster, the ratio of the number of data points in the consistency standard cluster cluster that are incorrectly classified into the abnormal problem list to the total number of data points in the cluster cluster, the ratio of the number of data points in the inconsistency standard cluster cluster that are correctly classified into the abnormal problem list to the total number of data points in the cluster cluster, and the ratio of the number of data points in the inconsistency standard cluster cluster that are incorrectly classified into the normal state list to the total number of data points in the cluster cluster.

[0097] In the embodiment of the present invention, the ratio of the number of data points correctly classified into the normal state list in the consistency standard cluster to the total number of data points in the cluster is calculated, which is recorded as P1.

[0098] Calculate the ratio of the number of data points in the consistency standard cluster that are misclassified into the abnormal problem list to the total number of data points in the cluster, which is recorded as P2; calculate the ratio of the number of data points in the inconsistency standard cluster that are correctly classified into the abnormal problem list to the total number of data points in the cluster, which is recorded as P3; calculate the ratio of the number of data points in the inconsistency standard cluster that are misclassified into the normal state list to the total number of data points in the cluster, which is recorded as P4; then, calculate the accuracy K of the cluster analysis using the following formula:

[0099]

[0100] Among them, n1 is the total number of data points in the consistent standard clustering cluster, and n2 is the total number of data points in the inconsistent standard clustering cluster.

[0101] The redetermining of the cluster analysis weight coefficient in the embodiment of the present invention includes redetermining the weight of the average difference amount, the weight of the dispersion index, the weight of the difference characterization coefficient, the weight of the abnormality degree of the electrical parameters of the data points in the cluster cluster, and the weight of the influence degree of the environmental parameters of the data points in the cluster cluster according to the method for determining the weight coefficient in the present invention. The preset time length can be set to ten times the transmission line data collection interval length, and the preset accuracy is the accurate average value after cluster analysis in the same transmission line, but the above value is not limited to this. Those skilled in the art can also adjust the value according to actual needs.

[0102] The present invention determines the accuracy of cluster analysis by calculating the correct classification and incorrect classification ratios of data points in the consistency standard cluster clusters and the inconsistency standard cluster clusters, and can comprehensively and objectively evaluate the accuracy of the clustering results. The calculation of different ratios takes into account the classification of normal states and abnormal problems, so that the accuracy calculation can better reflect the ability of cluster analysis in distinguishing normal and abnormal data, which helps to discover problems in cluster analysis and thus make targeted improvements. When the accuracy is less than the preset accuracy, a method for re-determining the cluster analysis weight coefficient is provided, that is, adjusting the weight according to the influence of different factors (average difference, dispersion index, difference characterization coefficient, electrical parameter abnormality and environmental parameter influence) on the clustering results, so that the cluster analysis can be more in line with actual needs. Different transmission lines may have different operating characteristics. By adjusting the weight coefficient, the influence of key factors on the clustering results can be highlighted, and the adaptability and accuracy of cluster analysis for specific transmission lines can be improved. The above method improves the accuracy of cluster analysis of electrical parameter data and environmental parameter data of transmission lines and the accuracy and efficiency of subsequent electrical parameter analysis of transmission lines.

[0103] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

Claims

1. A method for analyzing and verifying electrical parameters of transmission lines based on big data, characterized in that: include: Step S1, collecting electrical parameter data and environmental parameter data of the transmission line, and preprocessing the electrical parameter data and environmental parameter data to obtain standard data; Step S2, performing cluster analysis on the standard data to divide it into a number of clusters, including extracting electrical parameter characteristics and environmental parameter characteristics from the standard data, comparing the differences in the characteristics of each data point, and combining the data points according to the differences to obtain clusters, wherein the clusters must meet predetermined division criteria; Step S3, calculating a differentiation coefficient based on the difference in characteristic parameters between each data point in the cluster, and determining whether the cluster meets the consistency standard based on the differentiation coefficient; Step S4: selecting an analysis and processing method for each cluster, including: if the cluster meets the consistency standard, identifying the characteristics of the data points in the cluster, calculating the difference characterization coefficient based on the characteristics to determine whether the cluster meets the difference standard, prioritizing the clusters that meet the difference standard, selecting the transmission lines corresponding to any of the largest difference data points in several clusters in descending order of priority for key analysis and verification, and marking them; if the cluster does not meet the difference standard, performing a comprehensive analysis and verification on the transmission lines corresponding to all data points in the cluster; In step S5, the results of the focused analysis and verification and the comprehensive analysis and verification are classified separately, a list of abnormal problems and a list of normal status are established, the problem characteristics in the consistency standard clustering clusters and the inconsistency standard clustering clusters are compared, the accuracy of the previous clustering analysis is determined, and the clustering analysis weight coefficient is re-determined under the condition that the accuracy is less than the preset accuracy.

2. The method for analyzing and verifying electrical parameters of a transmission line based on big data according to claim 1, characterized in that: The preprocessing of the electrical parameter data and the environmental parameter data includes removing outliers in the data using the Laida criterion, filling missing data using a linear interpolation method, and mapping the data to the [0, 1] interval using a maximum-minimum normalization method.

3. The method for analyzing and verifying electrical parameters of a transmission line based on big data according to claim 2, characterized in that: In step S2, the predetermined division criteria include that the average Euclidean distance of data points within a cluster is less than a first preset threshold and the Euclidean distance between any two cluster centers is greater than a second preset threshold.

4. The method for analyzing and verifying electrical parameters of a transmission line based on big data according to claim 3, characterized in that: In step S3, calculating the differentiation coefficient according to the difference in characteristic parameters between the data points in the cluster includes: Calculate the Euclidean distance between any two data points in the cluster to form a distance matrix; Calculate the average value of the distance matrix and use it as the average difference between the data points in the cluster; Calculate the Euclidean distance from each data point in the cluster to the cluster center to form a center distance vector; Calculate the standard deviation of the center distance vector and use it as an indicator of the dispersion of data points around the cluster center; The differentiation coefficient is obtained by weighted summing the average difference and the dispersion index.

5. The method for analyzing and verifying electrical parameters of a power transmission line based on big data according to claim 4, characterized in that: In step S3, determining whether the cluster meets the consistency standard according to the differentiation coefficient includes: If the differentiation coefficient is less than or equal to a preset differentiation coefficient, determining that the cluster meets the consistency standard; If the differentiation coefficient is greater than a preset differentiation coefficient, it is determined that the cluster does not meet the consistency standard.

6. The method for analyzing and verifying electrical parameters of a transmission line based on big data according to claim 5, characterized in that: The preset differentiation coefficient is determined according to an average value of differentiation coefficients of clusters that have completed cluster analysis for the same transmission line.

7. The method for analyzing and verifying electrical parameters of power transmission lines based on big data according to claim 6, characterized in that: In step S4, calculating the difference characterization coefficient according to the features includes: Calculate the difference between the key feature parameters of each data point and the corresponding feature parameters of the cluster center to obtain the difference vector; Normalize the difference vector and calculate the norm of the normalized difference vector, which is used as a measure of the difference between the data point and the cluster center; Calculate the average value of the difference degree measure of all data points in the cluster and use it as the difference representation coefficient of the cluster.

8. The method for analyzing and verifying electrical parameters of power transmission lines based on big data according to claim 7, characterized in that: In step S4, determining that the clusters meet the difference standard includes that the difference characterization coefficient of the clusters is greater than or equal to a preset difference characterization coefficient threshold.

9. The method for analyzing and verifying electrical parameters of a power transmission line based on big data according to claim 8, characterized in that: In step S4, prioritizing the clusters that meet the difference criteria includes: Calculate the priority score of each cluster, which is obtained by weighted summation of the following factors: The size of the difference characterization coefficient; The abnormal degree of electrical parameters of data points within the cluster; The degree of influence of environmental parameters on data points within the cluster; The clusters are sorted in descending order according to their priority scores, with the clusters with higher priority scores at the front.

10. The method for analyzing and verifying electrical parameters of power transmission lines based on big data according to claim 9, characterized in that: In step S5, the accuracy of the preliminary cluster analysis is determined based on the ratio of the number of data points in the consistency standard cluster cluster that are correctly classified into the normal state list to the total number of data points in the cluster cluster, the ratio of the number of data points in the consistency standard cluster cluster that are incorrectly classified into the abnormal problem list to the total number of data points in the cluster cluster, the ratio of the number of data points in the inconsistency standard cluster cluster that are correctly classified into the abnormal problem list to the total number of data points in the cluster cluster, and the ratio of the number of data points in the inconsistency standard cluster cluster that are incorrectly classified into the normal state list to the total number of data points in the cluster cluster.

Images