A distribution network fault locating method and device based on power quality monitoring data

Abstract

The application provides a distribution network fault positioning method and device based on power quality monitoring data, and relates to the technical field of distribution network fault detection and positioning.The application collects power quality data before and after the occurrence of a fault in a distribution network, pre-processes the data, then performs voltage drop detection on each monitoring area, marks the area exceeding the threshold value as a fault area, monitors the current of the distribution network line in the fault area, detects the sudden increase of the current by using a moving average method to identify the fault line, detects the frequency change and harmonic change of each node in the fault line, obtains the frequency deviation degree and total harmonic distortion rate as main fault indicators, assigns a grade value to each node, calculates the fault correlation value of each node, finally establishes a comprehensive score mechanism, and determines the node with a high score and a large correlation value as the main fault node.The application realizes efficient positioning and identification of the fault node in the distribution network by collecting and analyzing the power quality data of the distribution network.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of distribution network fault detection and positioning, in particular to a distribution network fault positioning method and device based on power quality monitoring data. BACKGROUND

[0002] With the rapid development of power systems, the complexity and load fluctuation of distribution networks are increasing, leading to increasingly serious power quality problems. In the operation of distribution networks, the rapid positioning of faults is an important link to ensure power supply and equipment safety. The quality of power quality directly affects the operating efficiency, stability and safety of power equipment, and thus affects the reliability of the entire power system. Therefore, timely and accurate detection and positioning of distribution network faults have become the key to ensuring the safe and stable operation of power systems.

[0003] Power quality monitoring mainly includes monitoring of parameters such as voltage, current, frequency and harmonics. Through real-time monitoring of these parameters, abnormal conditions in the power system can be detected in a timely manner, especially when a fault occurs. Changes in power quality often provide important fault indications, such as a sharp drop in voltage, a sudden increase in current, fluctuations in frequency, and abnormal harmonics, which reflect potential fault types and locations.

[0004] In the prior art, traditional distribution network fault detection methods mainly rely on calculating the impedance of the two ends of the distribution network line and using the bisection method to locate the specific fault location. However, this method cannot take into account the complexity of the distribution network line, and it fails to consider the transient characteristics of voltage and current when a fault occurs.

[0005] In addition, the implementation of the bisection method requires multiple measurements and judgments, especially in cases where the line is long and the nodes are many. The operation process is too cumbersome, increasing the time and labor costs of fault repair. In particular, in emergency response situations, the complexity of the operation may cause delays, and traditional methods often rely on post-fault analysis, lacking real-time monitoring and alarm mechanisms. This means that when a fault occurs, the distribution network cannot obtain relevant data in a timely manner, resulting in an extended response time and affecting the safety and reliability of the system.

[0006] Therefore, it is necessary to provide a distribution network fault positioning method and device based on power quality monitoring data to solve the above problems.

[0007] The above information disclosed in the background section is only intended to strengthen the understanding of the background of the present disclosure, and therefore it can include information that does not constitute prior art known to those of ordinary skill in the art. SUMMARY

[0008] The present application aims to provide a distribution network fault positioning method and device based on power quality monitoring data to solve the problems raised in the background section.

[0009] To achieve the above object, the present application provides the following technical solutions:

[0010] A power quality monitoring data-based distribution network fault locating method and device, the specific steps comprising:

[0011] Step 1: Define a time detection window before and after the fault occurs, collect the power quality data of each monitoring area in the distribution network in the time detection window, and preprocess the obtained power quality data, the power quality data including voltage data, current data, frequency data and harmonic data, the preprocessing including data cleaning and normalization processing of the power quality monitoring data;

[0012] Step 2: Analyze the voltage data based on the voltage drop threshold to determine the distribution network fault area, in the distribution network fault area, obtain the current surge amplitude according to the moving average method, and determine the distribution network fault line in the distribution network fault area based on the current surge amplitude;

[0013] Step 3: Analyze the frequency data in the distribution network fault line based on the fast Fourier transform to determine the frequency deviation degree before and after the fault of each distribution network node, and take the frequency deviation degree as the first fault indicator;

[0014] Step 4: Analyze the harmonic data in the distribution network fault line based on the fast Fourier transform, extract the fundamental amplitude and each harmonic amplitude from the transform result, obtain the total harmonic distortion rate of each distribution network node based on the fundamental amplitude and each harmonic amplitude, and take the total harmonic distortion rate as the second fault indicator;

[0015] Step 5: Assign a grade value to the fault indicator of each distribution network node in the distribution network fault line, obtain the fault correlation value of each distribution network node according to the calculation method of Spearman rank correlation coefficient, establish a comprehensive score mechanism based on the first and second fault indicators of each distribution network node and the fault correlation value of the distribution network node, sort the comprehensive scores of each distribution network node from large to small, set a score threshold, form a set of distribution network nodes that exceed the score threshold, and select the distribution network node with the smallest fault correlation value in the set as the fault node.

[0016] Further, the obtained power quality data is preprocessed according to the following method:

[0017] The statistical method is used to identify the outliers and repeated data in the power quality data, the outliers and repeated data in the power quality data are deleted, and the mean, median or mode of the power quality data is used to fill in the missing values in the power quality data;

[0018] The normalization processing of the power quality data is the minimum-maximum normalization, which scales the data to The range is such that all power quality data have the same scale, and the formula is:

[0019] ;

[0020] in, It is the power quality data after normalization. is the original power quality data, is the minimum value of the same type of power quality data in the data set, It is the maximum value among the same type of power quality data in the dataset.

[0021] Furthermore, the voltage data of the time detection window is combined with the moving average method to obtain the drop voltage value of each monitoring area before and after the distribution network system fails. The calculation formula is:

[0022] ;

[0023] ;

[0024] ;

[0025] in, It represents the average voltage value in the monitoring area within the time detection window before the fault occurs. It represents the average voltage value in the monitoring area within the time detection window after the fault occurs. Indicates the voltage drop value before and after the fault occurs. Indicates the length of the time detection window, Indicates that in the monitoring area The voltage value at the moment, is the time variable within the time detection window, is the moment when the fault occurs;

[0026] The drop voltage value of each monitoring area in the distribution network system With the preset voltage drop threshold For comparison, if , then the monitoring area is determined to be a distribution network fault area.

[0027] Furthermore, the distribution network fault line is determined based on the current sudden increase magnitude in the marked distribution network fault area, according to the following method:

[0028] Based on the current data in the time detection window and the moving average method, the current surge value of each line in the distribution network fault area is obtained. The formula is:

[0029] ;

[0030] ;

[0031] ;

[0032] wherein, represents the average value of the current in the time detection window before the fault of the distribution network line in the monitoring area, represents the average value of the current in the time detection window after the fault of the distribution network line in the monitoring area, represents the current surge value before and after the fault;

[0033] Set the current surge threshold , compare the current surge value of each line in the distribution network fault area determined and , if , it is determined that this line in the distribution network fault area is a distribution network fault line.

[0034] Further, the frequency data of each distribution network node in the distribution network fault line is analyzed to determine the frequency deviation degree of each distribution network node in the distribution network fault line, and the method is as follows:

[0035] The frequency data in the time detection window before and after the fault is converted by using fast Fourier transform technology, the time domain frequency signal collected is windowed, the preset time detection window size is used as the fixed length signal segment after windowing, the signal segment is changed by using fast Fourier transform technology to obtain a complex array representing the amplitude and phase of each frequency component, and then the modulus of the complex array is calculated to obtain the amplitude of each frequency component. Through the calculation of the frequency resolution, the frequency value corresponding to each frequency component is determined, and finally according to the size of the amplitude, the frequency component with the largest amplitude is selected as the fundamental frequency, the difference between the fundamental frequency values of the two time detection windows before and after the fault of the distribution network node is calculated, and the ratio of the difference between the fundamental frequency values of the two time detection windows to the fundamental frequency value of the time detection window before the fault is taken as the frequency deviation degree of each distribution network node .

[0036] Further, the harmonic data of each distribution network node in the distribution network fault line is analyzed, and according to the amplitude of the frequency domain signal, the total harmonic distortion rate of each distribution network node in the distribution network fault line is determined, and the method is as follows:

[0037] The signal frequency domain is analyzed by using fast Fourier transform, the time domain signal is converted into frequency domain signal, the fundamental wave and its harmonics are identified, and the total harmonic distortion rate of each distribution network node is obtained, and the formula is as follows:

[0038] ;

[0039] in, Indicates the total harmonic distortion rate of the distribution network node, represents the fundamental amplitude, 、 、 Indicates the amplitude of each harmonic, Indicates the total order of harmonics.

[0040] Furthermore, a comprehensive scoring mechanism is established to determine the faulty nodes based on the following method:

[0041] The frequency deviation degree and total harmonic distortion rate of each distribution network node are used as the first fault indicator and the second fault indicator respectively. A level value is assigned to the fault indicator of each distribution network node in the distribution network fault line. The level difference between the first and second fault indicators is calculated, and the fault correlation value of the two fault indicators of each distribution network node is obtained based on the level difference. The formula is as follows:

[0042] ;

[0043] ;

[0044] in, Indicates the The level difference between the first and second fault indicators in the distribution network nodes, 、 They are respectively The level values ​​assigned to the first and second fault indicators in each distribution network node, Indicates the The fault correlation value of the first and second fault indicators in the distribution network nodes, Indicates the total number of distribution network nodes;

[0045] Establish a comprehensive scoring mechanism, set the score range of each fault indicator to 0-100, and assign proportional weights to the first and second fault indicators respectively 、 , the formula for determining the comprehensive score of each distribution network node is:

[0046] ;

[0047] in, Indicates the The comprehensive score of each distribution network node, Indicates the The frequency deviation degree of each distribution network node, Indicates the The total harmonic distortion rate of the distribution network nodes, and =1, ;

[0048] The fault correlation values of the first and second fault indicators of each distribution network node and the comprehensive scores are sorted in ascending order, and the distribution network node with the maximum fault correlation value and the highest comprehensive score is selected as the distribution network fault node. The distribution network fault node.

[0049] The application further provides a distribution network fault positioning device based on power quality monitoring data, which is used for executing the distribution network fault positioning method based on power quality monitoring data.

[0050] A data acquisition and preprocessing module is configured to define a time detection window before and after the fault occurs, acquire power quality data of each monitoring area in the distribution network in the time detection window, and preprocess the acquired power quality data, wherein the power quality data includes voltage data, current data, frequency data and harmonic data, and the preprocessing includes data cleaning and normalization processing of the power quality monitoring data.

[0051] A voltage drop and current surge detection module is configured to analyze the voltage data based on a voltage drop threshold to determine a distribution network fault area, acquire a current surge amplitude in the distribution network fault area according to a moving average method, and determine a distribution network fault line in the distribution network fault area based on the current surge amplitude.

[0052] A frequency change detection module is configured to analyze the frequency data in the distribution network fault line based on a fast Fourier transform to determine a frequency deviation degree before and after the fault of each distribution network node, and take the frequency deviation degree as a first fault indicator.

[0053] A harmonic change detection module is configured to analyze the harmonic data in the distribution network fault line based on a fast Fourier transform, extract a fundamental amplitude and each harmonic amplitude from a transform result, acquire a total harmonic distortion rate of each distribution network node based on the fundamental amplitude and each harmonic amplitude, and take the total harmonic distortion rate as a second fault indicator.

[0054] A fault correlation and comprehensive score module is configured to assign a grade value to the fault indicators of each distribution network node in the distribution network fault line, acquire a fault correlation value of each distribution network node according to a calculation method of a Spearman rank correlation coefficient, establish a comprehensive score mechanism based on the first and second fault indicators of each distribution network node and the fault correlation value of the distribution network node, sort the comprehensive scores of each distribution network node in descending order, set a score threshold, form a set of distribution network nodes that exceed the score threshold, and select the distribution network node with the minimum fault correlation value in the set as a fault node.​

[0055] Compared with the prior art, the present application has the beneficial effects that:

[0056] The present application aims at the deficiencies of traditional distribution network fault detection methods, and significantly improves the efficiency and accuracy of fault detection. Unlike the traditional method which relies on impedance calculation and bisection method for fault positioning, the present application comprehensively considers the complexity of distribution network lines, based on real-time monitoring and analysis of voltage, current, frequency and harmonic data, from the determination of fault area to fault line to fault node, time detection window design is adopted to realize real-time monitoring and rapid response.

[0057] Secondly, the comprehensive score mechanism set up further refines the dimensions of fault judgment by introducing the frequency deviation degree and the total harmonic distortion rate, which not only enhances the recognition ability of different types of faults, but also effectively evaluates the fault correlation of each node through the Spearman rank correlation coefficient, simplifies the operation process, and makes the fault positioning more efficient and intuitive. BRIEF DESCRIPTION OF DRAWINGS

[0058] Figure 1 The present application is a whole method flowchart.

[0059] Figure 2 The present application is a first fault index comprehensive score contribution diagram.

[0060] Figure 3 The present application is a second fault index comprehensive score contribution diagram.

[0061] Figure 4 The present application is a system module flowchart. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical scheme and advantages of the present application clearer and more explicit, the present application will be further described in detail below combined with specific examples.

[0063] It should be noted that the technical terms or scientific terms used in the present application should be understood as the general meaning understood by those skilled in the art to which the present application belongs, unless otherwise defined. The "first", "second" and similar words used in the present application do not represent any order, quantity or importance, but are only used to distinguish different components. "Include" or "contain" and similar words mean that the elements or objects before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connected" or "connected" and similar words are not limited to physical or mechanical connection, but can include electrical connection, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to represent the relative positional relationship, when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0064] Embodiments:

[0065] Please refer to Figure 1 A distribution network fault locating method based on power quality monitoring data, the specific steps comprising:

[0066] Step 1: Define a time detection window before and after the fault occurs, collect the power quality data of each monitoring area in the distribution network in the time detection window, and pre-process the obtained power quality data, the power quality data includes voltage data, current data, frequency data and harmonic data, the preprocessing includes data cleaning and normalization processing of power quality monitoring data;

[0067] Step 2: Analyze the voltage data based on the voltage drop threshold to determine the distribution network fault area, in the distribution network fault area, the current surge amplitude is obtained according to the moving average method, and the distribution network fault line in the distribution network fault area is determined based on the current surge amplitude;

[0068] Step 3: Analyze the frequency data in the distribution network fault line based on fast Fourier transform to determine the frequency deviation degree before and after the fault of each distribution network node, and take the frequency deviation degree as the first fault index;

[0069] Step 4: Analyze the harmonic data in the distribution network fault line based on fast Fourier transform, extract the fundamental amplitude and each harmonic amplitude from the transform result, obtain the total harmonic distortion rate of each distribution network node based on the fundamental amplitude and each harmonic amplitude, and take the total harmonic distortion rate as the second fault index;

[0070] Step 5: Assign a rank value to the fault indicator of each distribution network node in the distribution network fault line, obtain the fault correlation value of each distribution network node according to the calculation method of Spearman rank correlation coefficient, establish a comprehensive score mechanism based on the first and second fault indicators of each distribution network node and the fault correlation value of the distribution network node, sort the comprehensive scores of each distribution network node from large to small, set a score threshold, and form a set of distribution network nodes that exceed the score threshold, and select the distribution network node with the smallest fault correlation value in the set as the fault node.

[0071] It should be noted that by detecting and deleting outliers and repeated data from power quality data, noise and errors in the data can be effectively removed, which is crucial for power grid fault analysis. Outliers may be caused by equipment failure or external interference, and if not handled, it will lead to misjudgment and false fault location, affecting the safety of the system. Using mean, median or mode to fill in missing values can maximize the effectiveness of the data and avoid data deviation caused by missing values, thereby ensuring reliable analysis of power quality changes. The minimum-maximum normalization processing scales all power quality data to a unified scale, allowing different types of data to be compared and analyzed within the same framework.

[0072] Therefore, the acquired power quality data needs to be preprocessed, and the method is as follows:

[0073] The statistical method is used to identify outliers and repeated data in power quality data, delete outliers and repeated data in power quality data, and use the mean, median or mode of power quality data to fill in missing values in power quality data;

[0074] The normalization of power quality data is minimum-maximum normalization, which scales the data to range, so that all power quality data have the same scale, and the formula is:

[0075] ;

[0076] wherein, is the normalized power quality data, is the original power quality data, is the minimum value of the same type of power quality data in the data set, is the maximum value of the same type of power quality data in the data set.

[0077] It should be noted that by analyzing the voltage data of the time detection window before and after the fault occurs based on the voltage drop threshold, the moving average method is used to obtain the voltage average before and after the fault, which can effectively capture the dynamic characteristics of voltage changes and timely identify the distribution network fault area. In addition, with the help of the design of the time detection window, the system can flexibly adjust the length of the time detection window. The value of can be adjusted to adapt to the operating conditions and fault characteristics of different distribution networks, further enhancing the universality and adaptability of fault location.

[0078] Therefore, it is necessary to combine the voltage data of the time detection window with the moving average method to obtain the drop voltage value of each monitoring area before and after the distribution network system fails. The calculation formula is:

[0079] ;

[0080] ;

[0081] ;

[0082] in, It represents the average voltage value in the monitoring area within the time detection window before the fault occurs. It represents the average voltage value in the monitoring area within the time detection window after the fault occurs. Indicates the voltage drop value before and after the fault occurs. Indicates the length of the time detection window, Indicates that in the monitoring area The voltage value at the moment, is the time variable within the time detection window, is the moment when the fault occurs;

[0083] The drop voltage value of each monitoring area in the distribution network system With the preset voltage drop threshold For comparison, if , then the monitoring area is determined to be a distribution network fault area.

[0084] It should be noted that the moving average method is used to analyze the collected current data, and the current average value before and after the fault is calculated in real time to identify the current surge value. By setting the current surge threshold, effective identification of the fault line is achieved, which significantly improves the sensitivity and accuracy of distribution network fault detection.

[0085] Therefore, it is necessary to determine the distribution network fault line based on the current surge magnitude in the marked distribution network fault area. The method is as follows:

[0086] According to the current data of the time detection window, the current surge value of each line in the distribution network fault area is obtained by combining the moving average method, and the formula is:

[0087]

[0088]

[0089]

[0090] , wherein, represents the average current value of the distribution network line in the time detection window before the fault occurs in the monitoring area, represents the average current value of the distribution network line in the time detection window after the fault occurs in the monitoring area, represents the current surge value before and after the fault occurs;

[0091] The current surge threshold is set , the current surge value of each line in the distribution network fault area determined is compared with , if , it is determined that the line in the distribution network fault area is a distribution network fault line.

[0092] It should be noted that the fast Fourier transform technology can accurately extract the characteristics of the frequency signal before and after the fault occurs, identify the change of the fundamental frequency, and obtain the frequency deviation degree, which provides a quantitative basis for fault diagnosis, so that the operation and maintenance personnel can quickly locate the problem node and take effective measures to handle it.

[0093] Therefore, the frequency data of each distribution network node in the distribution network fault line needs to be analyzed to determine the frequency deviation degree of each distribution network node in the distribution network fault line, and the method is:

[0094] The frequency data in the time detection window before and after the fault occurs is converted by using the fast Fourier transform technology, the time domain frequency signal collected is windowed, the preset time detection window size is used as the fixed length signal segment after windowing, the signal segment is changed by using the fast Fourier transform technology to obtain a complex array representing the amplitude and phase of each frequency component, and then the modulus of the complex array is calculated to obtain the amplitude of each frequency component. Through the calculation of the frequency resolution, the frequency value corresponding to each frequency component is determined, and finally the frequency component with the largest amplitude is selected as the fundamental frequency according to the size of the amplitude, the difference between the fundamental frequency values of the two time detection windows before and after the fault occurs in the distribution network node is calculated, and the ratio of the difference between the fundamental frequency values of the two time detection windows to the fundamental frequency value of the time detection window before the fault occurs is taken as the frequency deviation degree of each distribution network node .​​​

[0095] It should be noted that the time domain signal is converted into the frequency domain signal by fast Fourier transform, the fundamental wave and its harmonic components can be effectively identified, and the harmonic distortion degree is quantified, and the total harmonic distortion rate In the formula, The higher the value, the more serious the harmonic distortion, the worse the power quality, is the fundamental amplitude, that is, the amplitude of the lowest frequency component in the signal, the fundamental wave is the main component of the signal, usually refers to the 50Hz or 60Hz alternating current frequency in the power system, and the fundamental amplitude directly affects the effective power of the power system, the harmonic is an integer multiple frequency component of the fundamental wave, usually derived from nonlinear load, is the second harmonic, is the third harmonic, and so on, until The harmonic effective value is compared with the fundamental amplitude, and the ratio is calculated, which indicates the strength of the harmonic component relative to the fundamental wave. The higher the ratio, the greater the influence of the harmonic on the signal, and the worse the power quality.

[0096] Therefore, it is necessary to detect the harmonic change of each distribution network node in the fault line of the distribution network, and determine the total harmonic distortion rate of each distribution network node in the fault line of the distribution network according to the amplitude of the frequency domain signal. The method is as follows:

[0097] The signal frequency domain is analyzed by using fast Fourier transform, the time domain signal is converted into the frequency domain signal, the fundamental wave and its harmonics are identified, and the total harmonic distortion rate of each distribution network node is obtained. The formula is as follows:

[0098]

[0099] Among them, represents the total harmonic distortion rate of the distribution network node, represents the fundamental amplitude, , , represents the amplitude of each harmonic, represents the total order of the harmonic.

[0100] It should be noted that by taking the frequency deviation degree and the total harmonic distortion rate as the key fault indicators, and calculating the grade difference and the fault correlation value, the fault risk of each distribution network node can be effectively evaluated. The setting of the comprehensive score makes different fault indicators quantifiable, which is convenient for sorting the nodes, so as to accurately locate the fault node. This mechanism not only improves the effectiveness of fault diagnosis, but also provides a scientific basis for subsequent fault handling and maintenance, which helps to improve the reliability and stability of the distribution network;

[0101] ​The frequency deviation degree and total harmonic distortion rate are divided into four levels, namely level 1, level 2, level 3 and level 4. Defined as level 1, which is the level value of the first fault indicator ; When the frequency deviation is Defined as level 2, which is the level value of the first fault indicator ; When the frequency deviation is Level 3, the level value of the first fault indicator ; When the frequency deviation exceeds When it is defined as level 4, that is, the level value of the first fault indicator When the total harmonic distortion rate is within the range of When it is defined as level 1, that is, the level value of the second fault indicator ; When the total harmonic distortion rate is within this range When it is defined as level 2, that is, the level value of the second fault indicator ; When the total harmonic distortion rate is within the range of When it is defined as level 3, that is, the level value of the second fault indicator ; When the total harmonic distortion rate exceeds the range When it is defined as level 4, that is, the level value of the second fault indicator .

[0102] It should be noted that when the first fault indicator and the second fault level indicator are both at level 1, the node is determined not to be a distribution network fault node; when the first fault indicator and the second fault level indicator are both at level 4, the node is directly determined to be a distribution network fault node; when the first fault indicator and the second fault level indicator are different and are at level 1 and level 4 at the same time, it is necessary to combine the comprehensive scoring formula for judgment.

[0103] Therefore, it is necessary to establish a comprehensive scoring mechanism to determine the abnormal fault nodes based on the following method:

[0104] The frequency deviation degree and total harmonic distortion rate of each distribution network node are used as the first fault indicator and the second fault indicator respectively. A level value is assigned to the fault indicator of each distribution network node in the distribution network fault line. The level difference between the first and second fault indicators is calculated, and the fault correlation value of the two fault indicators of each distribution network node is obtained based on the level difference. The formula is as follows:

[0105] ;

[0106] ;

[0107] in, Indicates the The level difference between the first and second fault indicators in the distribution network nodes, 、 They are respectively The level values ​​assigned to the first and second fault indicators in each distribution network node, Indicates the The fault correlation value of the first and second fault indicators in the distribution network nodes, Indicates the total number of distribution network nodes; in the above calculation formula of fault correlation value, when the level difference between the first and second fault indicators is The closer, The smaller the fault correlation value The smaller the value, the higher the authenticity of the fault in the distribution network node. is a normalization factor calculated using the number of nodes, so that The value of is kept within a reasonable range to facilitate comparison of fault correlation values ​​between different nodes. The numerator uses the square form of the grade difference This is to prevent the fault-related value from becoming negative.

[0108] As shown in the following table, in the fault correlation value ranking table, we selected 16 distribution network fault nodes, assigned grade values ​​to their fault indicators, and calculated the grade difference and fault correlation value. The chart shows that the fault correlation values ​​range from 0.998786 to 1, indicating that the fault correlation values ​​of most nodes are very close to 1, which means that the fault indicators of these nodes are relatively stable or have a low failure risk. Nodes 1, 6, 11, and 16 have the largest fault correlation values, indicating that their grade difference is 0, which means that their frequency deviation and total harmonic distortion rate levels are exactly the same. At this time, the frequency deviation grade and total harmonic distortion rate grade are determined, and the node with the largest frequency deviation grade and total harmonic distortion rate grade values ​​is selected as the distribution network fault abnormal node.

[0109] Table 1 - Fault correlation value ranking table

[0110]

[0111] Establish a comprehensive scoring mechanism, set the score range of each fault indicator to 0-100, and assign proportional weights to the first and second fault indicators respectively 、 , the formula for determining the comprehensive score of each distribution network node is:

[0112] ;

[0113] in, Indicates the The comprehensive score of each distribution network node, Indicates the The frequency deviation degree of each distribution network node, represents the total harmonic distortion rate of the nth distribution network node, and = 1, ; in the above formula for calculating the comprehensive score of the distribution network node, the weight ratio is set to because the frequency deviation degree has a more direct and significant impact on the stability and safety of the distribution network. The deviation of the frequency will cause the imbalance of the power system operation, which may lead to equipment damage or system failure. Therefore, a higher weight is given to the frequency deviation degree to better reflect its impact on the overall safety of the system; while the total harmonic distortion rate also affects the normal operation of power equipment, but its impact is relatively gradual and potential, and it may take a long time to show serious consequences, so a lower weight is given in the calculation of the comprehensive score;

[0114] A score threshold is set, and the distribution network nodes that exceed the score threshold are grouped into a set, and the distribution network node with the smallest fault correlation value in the set is selected as the distribution network fault node.

[0115] As shown in the following table, in the distribution network node comprehensive score ranking table, we calculate the comprehensive scores of sixteen randomly selected distribution network nodes. It can be clearly seen that when the level value of the first fault indicator and the second fault indicator exceeds level 4, the comprehensive score will significantly increase, and there is a relatively obvious threshold effect in the comprehensive score of the distribution network node after the fault indicator exceeds level 4. That is, the closer or higher the threshold, such as 40%, the higher the sensitivity of the comprehensive score to the change of the indicator. The improvement of the frequency deviation degree and the total harmonic distortion rate should be monitored and managed. In particular, when the indicator approaches or exceeds level 4, appropriate preventive measures should be taken to prevent potential failures.

[0116] Table 2 - Distribution network node comprehensive score ranking table

[0117]

[0118] Please refer to Figure 2 As shown in the first fault indicator comprehensive score contribution diagram, the horizontal axis represents the frequency deviation degree, ranging from 0 to 1, reflecting the frequency deviation of the distribution network node, and the higher the value, the greater the frequency deviation; the vertical axis represents the comprehensive score, ranging from 0 to 100, indicating the overall fault risk and power quality level of the node, the red curve represents the trend relationship between the comprehensive score and the frequency deviation degree, and the black block represents the actual data point; it can be seen that with the increase of the frequency deviation degree, the comprehensive score also shows a clear upward trend, when ​When the lower, close to 0, the comprehensive score is also relatively low, indicating that the state of the distribution network node is relatively normal at this time, when After increasing to a certain extent, the comprehensive score rises rapidly, especially when Close to 0.4 and above, the comprehensive score increases more obviously, which is consistent with the threshold effect described above.

[0119] Please refer to Figure 3 As shown in the second fault indicator comprehensive score contribution diagram, the horizontal axis represents the total harmonic distortion rate, reflecting the harmonic distortion of the distribution network node, and the higher the value, the more serious the harmonic influence, the red curve represents the trend relationship between the comprehensive score and the total harmonic distortion rate, and the black block represents the actual data point, with the increase of the total harmonic distortion rate, the comprehensive score also shows an obvious upward trend, which shows that the increase of the harmonic distortion degree will lead to the increase of the comprehensive score, combined with Figure 3 The fitting conclusion is that the Reduced Chi-Sqr and the COD values are 0.9978 and 0.9924 respectively, indicating that the fitting is good, and the relationship between the comprehensive score and the frequency deviation degree and the total harmonic distortion rate is well described, which means that most of the data points can be reasonably predicted, and the value of the adjusted R square is 0.9983, further verifying the accuracy of the formula, which shows that the fitting result of this formula is suitable for predicting and analyzing the relationship between power quality and fault indicators.

[0120] Please refer to Figure 4 The application further provides a distribution network fault positioning device based on power quality monitoring data, which is used to execute the above-mentioned distribution network fault positioning method based on power quality monitoring data, and comprises:

[0121] A data acquisition and preprocessing module is configured to define a time detection window before and after the fault occurs, acquire power quality data of each monitoring area in the distribution network in the time detection window, and preprocess the acquired power quality data, wherein the power quality data comprises voltage data, current data, frequency data and harmonic data, and the preprocessing comprises data cleaning and normalization processing on the power quality monitoring data;

[0122] A voltage drop and current surge detection module is configured to analyze the voltage data based on a voltage drop threshold to determine a distribution network fault area, and to acquire a current surge amplitude in the distribution network fault area according to a moving average method, and to determine a distribution network fault line in the distribution network fault area based on the current surge amplitude;

[0123] a frequency change detection module, configured to analyze frequency data in the fault line of the power distribution network based on a fast Fourier transform to determine a frequency deviation of each power distribution node before and after the fault, and take the frequency deviation as a first fault indicator;

[0124] a harmonic change detection module, configured to analyze harmonic data in the fault line of the power distribution network based on a fast Fourier transform, extract fundamental amplitude and harmonic amplitudes from the transform result, obtain a total harmonic distortion rate of each power distribution node based on the fundamental amplitude and the harmonic amplitudes, and take the total harmonic distortion rate as a second fault indicator;

[0125] a fault correlation and comprehensive score module, configured to assign a grade value to the fault indicators of each power distribution node in the fault line of the power distribution network, obtain a fault correlation value of each power distribution node according to a calculation method of a Spearman rank correlation coefficient, establish a comprehensive score mechanism based on the first and second fault indicators of each power distribution node and in combination with the fault correlation value of the power distribution node, sort the comprehensive scores of each power distribution node in descending order, set a score threshold, form a set of power distribution nodes that exceed the score threshold, and select a power distribution node with the smallest fault correlation value in the set as a fault node.

[0126] The above formulas are all dimensionless values calculated, the formulas are obtained by collecting a large amount of data to simulate a formula of the most recent real situation, and the preset parameters in the formulas are set by a person skilled in the art according to actual conditions.

[0127] The above embodiments can be realized wholly or partially by software, hardware, firmware or any combination thereof. When realized by software, the above embodiments can be realized wholly or partially in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of the examples described in connection with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized by hardware or software methods depends on the specific application and design constraints of the technical solutions.

[0128] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, which can be located in one place or distributed on multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiments.

[0129] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application.

Claims

1. A distribution network fault location method based on power quality monitoring data, characterized in that: The specific steps include: Step 1: Define a time detection window before and after the fault occurs, collect power quality data from each monitoring area in the distribution network during the time detection window, and preprocess the acquired power quality data. The power quality data includes voltage data, current data, frequency data, and harmonic data. The preprocessing includes data cleaning and normalization of the power quality monitoring data. Step 2: Analyze voltage data based on the drop voltage threshold to determine the distribution network fault area. Within the determined distribution network fault area, obtain the current surge magnitude using a moving average method, and determine the distribution network fault line within the distribution network fault area based on the current surge magnitude. Step 3: Analyze frequency data in the distribution network fault line based on fast Fourier transform to determine the frequency deviation degree of each distribution network node before and after the fault occurs, and use the frequency deviation degree as the first fault indicator; Step 4: Analyze the harmonic data in the distribution network fault line based on the fast Fourier transform. Extract the fundamental wave amplitude and each harmonic amplitude from the transformation results. Calculate the total harmonic distortion rate of each distribution network node based on the fundamental wave amplitude and each harmonic amplitude. Use the total harmonic distortion rate as the second fault indicator. Step 5: Assign a rank value to the fault indicator of each distribution network node in the distribution network fault line. Obtain the fault correlation value of each distribution network node based on the calculation method of the Spearman rank correlation coefficient. Establish a comprehensive scoring mechanism based on the first and second fault indicators of each distribution network node and the fault correlation value of the distribution network node. Sort the comprehensive score of each distribution network node from large to small, and set a score threshold. Group the distribution network nodes that exceed the score threshold into a set, and select the distribution network node with the smallest fault correlation value in the set as the fault node.

2. A distribution network fault location method based on power quality monitoring data according to claim 1, characterized in that: The acquired power quality data is preprocessed according to the following method: Use statistical methods to identify outliers and duplicate data in power quality data, delete outliers and duplicate data in power quality data, and use the mean, median or mode of power quality data to fill missing values ​​in power quality data; The normalization of power quality data is minimum-maximum normalization, which scales the data to The range is such that all power quality data have the same scale, and the formula is: ； in, It is the power quality data after normalization. is the original power quality data, is the minimum value of the same type of power quality data in the data set, It is the maximum value among the same type of power quality data in the dataset.

3. A distribution network fault location method based on power quality monitoring data according to claim 1, characterized in that: The voltage data in the time detection window before and after the fault is analyzed based on the voltage drop threshold, and the distribution network fault area is marked. The method is based on the following: Combining the voltage data of the time detection window with the moving average method, the voltage drop value of each monitoring area before and after the distribution network system failure is obtained. The calculation formula is: ； ； ； in, It represents the average voltage value in the monitoring area within the time detection window before the fault occurs. It represents the average voltage value in the monitoring area within the time detection window after the fault occurs. Indicates the voltage drop value before and after the fault occurs. Indicates the length of the time detection window, Indicates that in the monitoring area The voltage value at the moment, is the time variable within the time detection window, is the moment when the fault occurs; The drop voltage value of each monitoring area in the distribution network system With the preset voltage drop threshold For comparison, if , then the monitoring area is determined to be a distribution network fault area.

4. A distribution network fault location method based on power quality monitoring data according to claim 3, characterized in that: In the marked distribution network fault area, the distribution network fault line is determined based on the current surge magnitude, according to the following method: Based on the current data in the time detection window and the moving average method, the current surge value of each line in the distribution network fault area is obtained. The formula is: ； ； ； in, It indicates the average current value of the distribution network line in the monitoring area within the time detection window before the fault occurs. It indicates the average current value of the distribution network line in the monitoring area within the time detection window after the fault occurs. Indicates the sudden increase in current before and after the fault occurs; Set the current surge threshold , the current surge value of each line in the identified distribution network fault area is calculated and For comparison, if Then it is determined that this line in the distribution network fault area is a distribution network fault line.

5. A distribution network fault location method based on power quality monitoring data according to claim 4, characterized in that: The frequency data of each distribution network node in the distribution network fault line is analyzed to determine the frequency deviation degree of each distribution network node in the distribution network fault line. The method is based on: The fast Fourier transform technology is used to convert the frequency data in the time detection window before and after the fault occurs, and the collected time domain frequency signal is windowed. The preset time detection window size is used as a signal segment of fixed length after windowing. The fast Fourier transform technology is used to change the signal segment to obtain a complex array, which represents the amplitude and phase of each frequency component. The amplitude of each frequency component is obtained by calculating the modulus of the complex array. The frequency value corresponding to each frequency component is determined by calculating the frequency resolution. Finally, according to the size of the amplitude, the frequency component with the largest amplitude is selected as the fundamental frequency. The difference between the fundamental frequency values ​​of the two time detection windows before and after the distribution network node fails is calculated. The ratio of the difference between the fundamental frequency values ​​of the two time detection windows to the fundamental frequency value of the time detection window before the fault occurs is used as the frequency deviation degree of each distribution network node. .

6. A distribution network fault location method based on power quality monitoring data according to claim 5, characterized in that: The harmonic data of each distribution network node in the distribution network fault line is analyzed, and the total harmonic distortion rate of each distribution network node in the distribution network fault line is determined according to the amplitude of the frequency domain signal. The method is based on: Use fast Fourier transform to analyze the signal frequency domain, convert the time domain signal into the frequency domain signal, identify the fundamental wave and its harmonics, and obtain the total harmonic distortion rate of each distribution network node. The formula is as follows: ； in, Indicates the total harmonic distortion rate of the distribution network node, represents the fundamental amplitude, 、 、 Indicates the amplitude of each harmonic, Indicates the total order of harmonics.

7. A distribution network fault location method based on power quality monitoring data according to claim 6, characterized in that: A comprehensive scoring mechanism is established to identify abnormal fault nodes based on the following methods: The frequency deviation degree and total harmonic distortion rate of each distribution network node are used as the first fault indicator and the second fault indicator respectively. A level value is assigned to the fault indicator of each distribution network node in the distribution network fault line. The level difference between the first and second fault indicators is calculated, and the fault correlation value of the two fault indicators of each distribution network node is obtained based on the level difference. The formula is as follows: ； ； in, Indicates the The level difference between the first and second fault indicators in the distribution network nodes, 、 They are respectively The level values ​​assigned to the first and second fault indicators in each distribution network node, Indicates the The fault correlation value of the first and second fault indicators in the distribution network nodes, Indicates the total number of distribution network nodes; Establish a comprehensive scoring mechanism, set the score range of each fault indicator to 0-100, and assign proportional weights to the first and second fault indicators respectively 、 , the formula for determining the comprehensive score of each distribution network node is: ； in, Indicates the The comprehensive score of the distribution network nodes, Indicates the The frequency deviation degree of each distribution network node, Indicates the The total harmonic distortion rate of the distribution network nodes, and =1, ; Sort the fault correlation values ​​and comprehensive scores of the first and second fault indicators of each distribution network node from small to large, and select the node with the largest fault correlation value. And the highest comprehensive score The distribution network node is regarded as the distribution network fault node.

8. A distribution network fault location device based on power quality monitoring data, characterized in that: The distribution network fault locating device is used to execute the distribution network fault locating method based on power quality monitoring data according to any one of claims 1 to 7, comprising: A data acquisition and preprocessing module, which is used to define a time detection window before and after a fault occurs, collect power quality data from each monitoring area in the distribution network during the time detection window, and preprocess the acquired power quality data. The power quality data includes voltage data, current data, frequency data, and harmonic data. The preprocessing includes data cleaning and normalization of the power quality monitoring data. A voltage drop and current surge detection module, configured to analyze voltage data based on a voltage drop threshold to determine a distribution network fault area, obtain a current surge magnitude within the determined distribution network fault area using a moving average method, and determine a distribution network fault line within the distribution network fault area based on the current surge magnitude; a frequency change detection module configured to analyze frequency data in a distribution network fault line based on a fast Fourier transform to determine a frequency deviation degree before and after a fault occurs at each distribution network node, and to use the frequency deviation degree as a first fault indicator; a harmonic change detection module configured to analyze harmonic data in a distribution network fault line based on a fast Fourier transform, extract the fundamental wave amplitude and each harmonic amplitude from the transformation results, obtain the total harmonic distortion rate of each distribution network node based on the fundamental wave amplitude and each harmonic amplitude, and use the total harmonic distortion rate as a second fault indicator; A fault correlation and comprehensive scoring module is used to assign a grade value to the fault index of each distribution network node in the distribution network fault line, obtain the fault correlation value of each distribution network node based on the calculation method of the Spearman rank correlation coefficient, establish a comprehensive scoring mechanism based on the first and second fault indicators of each distribution network node and the fault correlation value of the distribution network node, sort the comprehensive score of each distribution network node from large to small, and establish a score threshold. The distribution network nodes that exceed the score threshold are grouped into a set, and the distribution network node with the smallest fault correlation value in the set is selected as the fault node.

Images