Power transmission line distributed fault location method based on non-contact detection

Abstract

The application discloses a power transmission line distributed fault positioning method based on non-contact detection and relates to the technical field of power system fault detection; the application adopts a non-contact detection node, integrates an electromagnetic field intensity sensor and an infrared thermal imager, simultaneously collects electromagnetic field intensity and temperature distribution data, analyzes electromagnetic field intensity data from time domain and frequency domain, quantizes statistical characteristics of temperature data, realizes multi-physical quantity and multi-dimension monitoring, can more comprehensively reflect fault characteristics, and solves the problems that false positives or false negatives often occur in the analysis and determination process of power transmission line fault characteristic information in the prior art, the depth mining of fault characteristic information is insufficient, and fault recognition capability is insufficient.

Description

Technical Field

[0001] This invention relates to the field of power system fault detection technology, and in particular to a method for distributed fault location of transmission lines based on non-contact detection. Background Technology

[0002] As the scale of power systems continues to expand, the length and complexity of transmission lines are increasing. As a key link in power transmission, the safe and stable operation of transmission lines is crucial. In the event of a fault, quickly and accurately locating the fault point is of great significance for reducing power outage time and minimizing economic losses.

[0003] However, existing distributed fault location methods for power transmission lines still have the following shortcomings:

[0004] In the process of analyzing and judging the fault characteristics of transmission lines, false alarms or omissions often occur, and there is a lack of in-depth mining of fault characteristics, resulting in insufficient fault identification capabilities.

[0005] Furthermore, the lack of effective use of historical data during fault location means that matching cases cannot be retrieved from historical data based on the analysis results of multi-dimensional data of transmission lines, resulting in a lack of assurance regarding the accuracy of fault location.

[0006] To address this, a distributed fault location method for transmission lines based on non-contact detection has been developed. Summary of the Invention

[0007] This invention aims to provide a distributed fault location method for transmission lines based on non-contact detection, which solves the problems of false alarms or missed alarms often occurring in the analysis and judgment of fault feature information of transmission lines in the existing technology, and the lack of in-depth mining of fault feature information, resulting in insufficient fault identification capability.

[0008] To solve the above-mentioned technical problems, the present invention provides a method for distributed fault location of transmission lines based on non-contact detection, comprising:

[0009] Fault characteristic analysis: Each non-contact detection node collects electromagnetic field intensity and temperature distribution data of the transmission line within the monitoring time window according to the set sampling frequency, and performs anomaly assessment calculation to determine the abnormal nodes and anomaly types within the current monitoring time window. The anomaly types include electromagnetic field anomalies, temperature anomalies, and comprehensive anomalies.

[0010] Fault Association and Localization: Collect abnormal behavior nodes within the current monitoring time window, generate an abnormal node list, and determine the abnormality level of each group of abnormal behavior nodes in the abnormal node list; the abnormality level includes Class I abnormality level and Class II abnormality level; build a knowledge database based on historical fault data, and perform corresponding steps in conjunction with the knowledge database to locate the faults of each group of abnormal behavior nodes in the abnormal node list that belong to Class II abnormality level; the knowledge database stores historical fault data cases for different nodes, including abnormal behavior type, fault location, and maintenance records.

[0011] Preferably, the process of performing anomaly assessment calculations on the electromagnetic field strength of the transmission line within the monitoring time window is as follows:

[0012] Electromagnetic field intensity data for each non-contact detection node within the current monitoring time window are extracted. First, the amplitude standard deviation is calculated using the formula... The standard deviation of the amplitude was calculated. ,in To monitor the average electromagnetic field intensity at each time point within the time window, i is the time point number, and N is the total number of time points;

[0013] Preset and The corresponding weighting coefficients, for the transmission lines within the current monitoring time window. and Normalization is performed separately, and each normalized value is multiplied by the corresponding set weight coefficient and then summed. The summation result is divided by an integer two to obtain the time-domain component evaluation value of each non-contact detection node within the current monitoring time window.

[0014] The signal spectrum of the electromagnetic field intensity signal is extracted using Fast Fourier Transform, and the harmonic distortion, abnormal frequency energy ratio, and spectral entropy are extracted from the signal spectrum.

[0015] The weighting coefficients corresponding to the preset harmonic distortion, abnormal frequency energy ratio and spectral entropy are normalized. After normalization, they are multiplied by the corresponding set weighting coefficients and then summed. The summation result is divided by an integer three to obtain the frequency domain component evaluation value of each non-contact detection node within the current monitoring time window.

[0016] Using the time-domain component evaluation value and the frequency-domain component evaluation value as the two legs of a right triangle, a right triangle is constructed. The area of ​​the right triangle is obtained, and the electromagnetic anomaly index of each non-contact detection node within the current monitoring time window is obtained.

[0017] Preferably, the process of performing anomaly assessment calculations on the temperature distribution data of the transmission line within the monitoring time window is as follows:

[0018] Extract the temperature distribution data of each non-contact detection node within the current monitoring time window, and first calculate the temperature standard deviation using the formula. The standard deviation of the amplitude was calculated. ,in To monitor the average temperature at each time point within the time window;

[0019] Preset and The corresponding weighting coefficients, for the current monitoring time window and Normalization is performed on each node, and then the normalized values ​​are multiplied by the corresponding weighting coefficients and summed to obtain the thermal image index of each non-contact detection node within the current monitoring time window.

[0020] Preferably, determining the abnormal nodes and abnormal behavior types within the current monitoring time window specifically involves:

[0021] The threshold indices corresponding to the preset electromagnetic anomaly index and thermal imaging index are set. If the electromagnetic anomaly index or thermal imaging index of a certain group of non-contact detection nodes is higher than the corresponding preset threshold index within the current monitoring time window, it is determined to be an abnormal node.

[0022] If the electromagnetic anomaly index is higher than the corresponding preset threshold index, the anomaly type is determined to be electromagnetic field anomaly; if the thermal imaging index is higher than the corresponding preset threshold index, the anomaly type is determined to be temperature anomaly; if both the electromagnetic anomaly index and the thermal imaging index are higher than the corresponding preset threshold index, the anomaly type is determined to be comprehensive anomaly.

[0023] Preferably, the determination of the anomaly level for each group of abnormal nodes in the abnormal node list specifically involves:

[0024] M1: Identify the abnormal behavior type of each group of abnormal behavior nodes in the abnormal node list. If the abnormal behavior type is a comprehensive abnormality, it is directly determined to be a Class II abnormality level.

[0025] If the abnormality type is electromagnetic field abnormality or temperature abnormality, the corresponding electromagnetic abnormality index or thermal imaging index is extracted, and the difference between it and the corresponding threshold index is calculated and recorded as the threshold degree difference.

[0026] The threshold level reference difference is preset for electromagnetic field anomalies and temperature anomalies respectively. If the calculated threshold level difference is higher than the corresponding threshold level reference difference, it is directly judged as a Class II anomaly level; otherwise, it is judged as a Class I anomaly level.

[0027] As a preferred option, if the condition is determined to be of an abnormal level, then the following actions are taken:

[0028] M2: If a group of abnormal nodes is determined to be of an abnormal level, the electromagnetic anomaly index or thermal imaging index within the monitoring time window after the determination time point is used as the starting point. If it is still an abnormal node, step M1 is executed again to determine the abnormal level.

[0029] Preferably, the step of combining the knowledge database to perform corresponding steps to locate the faults of each group of abnormal nodes belonging to the second-level abnormality in the abnormal node list is as follows:

[0030] M3: If a group of abnormal nodes is determined to be of Class II anomaly level, the abnormal behavior type of the abnormal node is first identified, and then the electromagnetic anomaly index and thermal imaging index are extracted as reference data. For each group of abnormal nodes with a Class II anomaly level, after identifying the node number, the historical fault cases corresponding to each node number are retrieved in the knowledge database. Cases that match the current abnormal behavior type are selected from the historical fault cases and retained as selected fault cases.

[0031] Construct a Cartesian coordinate system, with the x-axis representing the electromagnetic anomaly index and the y-axis representing the thermal imaging index; construct coordinate points based on the electromagnetic anomaly index and thermal imaging index of the anomaly manifestation node, and plot them in the Cartesian coordinate system as matching points; construct coordinate points based on the electromagnetic anomaly index and thermal imaging index of each group of screened fault cases corresponding to the anomaly manifestation node, and plot them in the Cartesian coordinate system as fault points;

[0032] Draw a circle with the matching point as the center and the preset initial distance as the radius, filter the fault points inside the circle, and calculate the fault location index Tc of each group of fault points inside the circle.

[0033] After calculating the fault location index Tc of each group of fault points within the circle, the fault points with higher fault location index Tc are selected from the fault points within each group of fault points, and the fault locations are identified from the corresponding cases as the fault location results of the abnormal node with the judgment result of Class II abnormality level.

[0034] Preferably, the specific process for calculating the fault location index Tc is as follows:

[0035] Starting from the matching point, connect it to each group of fault points within the circle, and mark the line segment connecting the matching point to each group of fault points within the circle as the fault traction line.

[0036] Obtain the length of each faulty traction line as a similarity value for the fault points within each group of circles;

[0037] Starting from the current time point, count the number of times the fault points in each group of circles are matched within the set time window before the starting point, and use this as the similarity bias value of the fault points in each group of circles.

[0038] The number of matches in the number of fault point matches within each group of circles is analyzed. The number of matches indicates that after matching the corresponding historical fault case, the maintenance personnel go to the corresponding fault location and find that the current fault location matches the fault location of the corresponding historical fault case. If the current match is successful, the current match is considered a match. The proportion of the number of matches in the total number of matches is calculated as the confidence level value of the fault point within each group of circles.

[0039] The similarity value, similarity bias value, and confidence level value of the fault points within each group of circles are respectively labeled as follows: After marking, normalization is performed before inputting into the formula. Weighted calculations were performed to obtain the fault location index Tc for each group of fault points within the circle; where These are the weighting coefficients for similarity value, similarity bias value, and confidence level value, respectively.

[0040] Preferably, the construction process of each non-contact detection node is as follows:

[0041] x non-contact detection nodes are deployed along the transmission line at predetermined intervals; the non-contact detection nodes include electromagnetic field strength sensors and infrared thermal imagers; the value of x is specifically set according to the length of the transmission line.

[0042] Compared with the prior art, the beneficial effects of the present invention are:

[0043] This invention employs a non-contact detection node that integrates an electromagnetic field strength sensor and an infrared thermal imager, simultaneously collecting electromagnetic field strength and temperature distribution data. It analyzes the electromagnetic field strength data in both the time and frequency domains and quantifies it by combining the statistical characteristics of the temperature data. This enables multi-physical quantity and multi-dimensional monitoring, providing a more comprehensive reflection of fault characteristics. It solves the problems in existing technologies where false alarms or missed alarms often occur during the analysis and judgment of fault characteristic information of transmission lines, and where the lack of in-depth mining of fault characteristic information leads to insufficient fault identification capabilities.

[0044] This invention collects electromagnetic field strength and temperature distribution data of transmission lines through various non-contact detection nodes, calculates electromagnetic anomaly index and thermal imaging index using multi-parameter evaluation, and constructs a knowledge database based on historical fault data for fault location. When locating nodes of the second-class anomaly level, a Cartesian coordinate system is constructed, and the fault location index is calculated by comprehensively considering similarity value, similarity bias value and confidence value. Fault points with higher fault location indices are selected to determine the fault location, which greatly improves the accuracy of fault location.

[0045] This invention categorizes abnormal behavior nodes into two levels of anomalies. For the first level of anomalies, the electromagnetic anomaly index or thermal imaging index within the subsequent monitoring time window is continuously analyzed, starting from the judgment time point, to eliminate instantaneous interference or short-term fluctuations and reduce false alarms. For the second level of anomalies, they are directly associated with high-risk faults, which are prioritized to trigger precise location and avoid missed alarms, thus balancing monitoring sensitivity and reliability.

[0046] This invention sets an additional coefficient based on the type of abnormal behavior, and multiplies it by the fault location index to obtain a priority index. The additional coefficient for comprehensive anomalies is greater than that for electromagnetic field anomalies and temperature anomalies, so that comprehensive anomaly nodes can be processed first when their indices are similar, making the operation and maintenance strategy conform to the "risk priority" principle and optimizing the allocation of operation and maintenance resources. Attached Figure Description

[0047] Further details, features, and advantages of this application are disclosed in the following description of exemplary embodiments in conjunction with the accompanying drawings, in which:

[0048] Figure 1 This is a flowchart of the present invention;

[0049] Figure 2 This is a schematic diagram illustrating the construction of the planar rectangular coordinate system in this invention. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] The terminology used in this disclosure is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The singular forms “group,” “class,” and “the” as used in this disclosure and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0052] It should be understood that although the terms first, second, third, etc., may be used in this disclosure to describe various information, such information should not be limited to these terms. These terms are used only to distinguish information of the same type from one another. For example, without departing from the scope of this disclosure, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0053] Example

[0054] Please see Figures 1-2 As shown, the distributed fault location method for transmission lines based on non-contact detection includes:

[0055] Distributed detection construction: x non-contact detection nodes are deployed along the transmission line at set distance intervals; the non-contact detection nodes include electromagnetic field strength sensors and infrared thermal imagers; the value of x is specifically set according to the length of the transmission line, for example, it can be deployed every 5 kilometers.

[0056] In addition, each contactless detection node is also integrated with a wireless communication module, which adopts a hybrid communication mode of LoRa / Wi-Fi / 4G. Under normal conditions, it transmits periodic data in low-power LoRa (with an interval of 1 minute), and in the event of a fault, it switches to the 4G network to upload high-frequency sampling data in real time (sampling rate of 10kHz).

[0057] Fault characteristic analysis: Each non-contact detection node collects electromagnetic field intensity and temperature distribution data of the transmission line within the monitoring time window according to the set sampling frequency, and performs anomaly assessment calculation to determine the abnormal nodes and anomaly types within the current monitoring time window. The anomaly types include electromagnetic field anomalies, temperature anomalies, and comprehensive anomalies.

[0058] Specifically:

[0059] Electromagnetic field intensity data for each non-contact detection node within the current monitoring time window are extracted. First, the amplitude standard deviation is calculated using the formula... The standard deviation of the amplitude was calculated. ,in To monitor the average electromagnetic field intensity at each time point within the time window, i is the time point number, and N is the total number of time points;

[0060] Preset and The corresponding weighting coefficients, for the transmission lines within the current monitoring time window. and Normalization is performed separately, and each normalized value is multiplied by the corresponding set weight coefficient and then summed. The summation result is divided by an integer two to obtain the time-domain component evaluation value of each non-contact detection node within the current monitoring time window.

[0061] The signal spectrum of the electromagnetic field intensity signal is extracted using Fast Fourier Transform, and the harmonic distortion, abnormal frequency energy ratio, and spectral entropy are extracted from the signal spectrum.

[0062] Additional explanation,

[0063] Harmonic distortion: The ratio of harmonic energy to fundamental frequency energy, reflecting the degree of abnormality of non-power frequency components;

[0064] Abnormal frequency energy percentage: The proportion of energy in the target abnormal frequency band (such as 30kHz~100kHz corresponding to corona discharge) to the total spectrum energy;

[0065] Spectral entropy (SE): measures the degree of disorder in frequency distribution. The higher the entropy value, the more complex the signal (abnormal signals usually have higher entropy values).

[0066] The weighting coefficients corresponding to the preset harmonic distortion, abnormal frequency energy ratio and spectral entropy are normalized. After normalization, they are multiplied by the corresponding set weighting coefficients and then summed. The summation result is divided by an integer three to obtain the frequency domain component evaluation value of each non-contact detection node within the current monitoring time window.

[0067] Using the time-domain component evaluation value and the frequency-domain component evaluation value as the two legs of a right triangle, a right triangle is constructed, and the area of ​​the right triangle is obtained to obtain the electromagnetic anomaly index of each non-contact detection node within the current monitoring time window.

[0068] In addition, by using time-domain-frequency domain feature fusion, multi-parameter weighted evaluation, and distributed node collaboration, "accurate detection, quantitative evaluation, and rapid location" of electromagnetic anomalies in transmission lines were achieved.

[0069] Extract the temperature distribution data of each non-contact detection node within the current monitoring time window, and first calculate the temperature standard deviation using the formula. The standard deviation of the amplitude was calculated. ,in To monitor the average temperature at each time point within the time window;

[0070] Preset and The corresponding weighting coefficients, for the current monitoring time window and Normalization is performed separately, and each normalized value is multiplied by the corresponding set weight coefficient and then summed to obtain the thermal image index of each non-contact detection node within the current monitoring time window.

[0071] To elaborate further, the thermal imaging index achieves real-time, accurate, and distributed monitoring of the thermal state of transmission lines by quantitatively modeling the statistical characteristics (mean and standard deviation) of temperature distribution.

[0072] The threshold indices corresponding to the preset electromagnetic anomaly index and thermal imaging index are set. If the electromagnetic anomaly index or thermal imaging index of a certain group of non-contact detection nodes is higher than the corresponding preset threshold index within the current monitoring time window, it is determined to be an abnormal node.

[0073] If the electromagnetic anomaly index is higher than the corresponding preset threshold index, the anomaly type is determined to be electromagnetic field anomaly; if the thermal imaging index is higher than the corresponding preset threshold index, the anomaly type is determined to be temperature anomaly; if both the electromagnetic anomaly index and the thermal imaging index are higher than the corresponding preset threshold index, the anomaly type is determined to be comprehensive anomaly.

[0074] Fault Association and Localization: Collect abnormal behavior nodes within the current monitoring time window, generate an abnormal node list, and determine the abnormality level of each group of abnormal behavior nodes in the abnormal node list; the abnormality level includes Class I abnormality level and Class II abnormality level; build a knowledge database based on historical fault data, and combine the knowledge database to perform corresponding steps to locate the faults of each group of abnormal behavior nodes in the abnormal node list that belong to Class II abnormality level; the knowledge database stores historical fault data cases for different nodes, including abnormal behavior type, fault location, and maintenance records;

[0075] Specifically:

[0076] M1: Identify the abnormal behavior type of each group of abnormal behavior nodes in the abnormal node list. If the abnormal behavior type is a comprehensive abnormality, it is directly determined to be a Class II abnormality level.

[0077] If the abnormality type is electromagnetic field abnormality or temperature abnormality, the corresponding electromagnetic abnormality index or thermal imaging index is extracted, and the difference between it and the corresponding threshold index is calculated and recorded as the threshold degree difference.

[0078] The threshold level reference difference is preset for electromagnetic field anomalies and temperature anomalies respectively. If the calculated threshold level difference is higher than the corresponding threshold level reference difference, it is directly judged as a Class II anomaly level; otherwise, it is judged as a Class I anomaly level.

[0079] M2: If the result of a certain group of abnormal nodes is a Class I abnormality, then take the judgment time point as the starting point and analyze the electromagnetic abnormality index or thermal imaging index within the monitoring time window after the starting point. If it is still an abnormal node, then execute step M1 again to determine the abnormality level.

[0080] Additional explanation: Category II anomaly levels (comprehensive anomaly or single anomaly exceeding the threshold difference) are directly associated with high-risk faults, and will be given priority to trigger accurate positioning based on historical knowledge databases, avoiding misjudgment of single parameters and improving the identification efficiency of complex faults (such as corona discharge accompanied by local heating).

[0081] Class 1 anomaly level (poor degree of single anomaly not exceeding the threshold) eliminates instantaneous interference or short-term fluctuations through continuous monitoring (step M2), reduces false alarms, avoids over-response to minor anomalies, and balances monitoring sensitivity and reliability.

[0082] M3: If a group of abnormal nodes is determined to be of Class II anomaly level, the abnormal behavior type of the abnormal node is first identified, and then the electromagnetic anomaly index and thermal imaging index are extracted as reference data. For each group of abnormal nodes with a Class II anomaly level, after identifying the node number, the historical fault cases corresponding to each node number are retrieved in the knowledge database. Cases that match the current abnormal behavior type are selected from the historical fault cases and retained as selected fault cases.

[0083] Construct a Cartesian coordinate system, with the x-axis representing the electromagnetic anomaly index and the y-axis representing the thermal imaging index; construct coordinate points based on the electromagnetic anomaly index and thermal imaging index of the anomaly manifestation node, and plot them in the Cartesian coordinate system as matching points; construct coordinate points based on the electromagnetic anomaly index and thermal imaging index of each group of screened fault cases corresponding to the anomaly manifestation node, and plot them in the Cartesian coordinate system as fault points;

[0084] Draw a circle with the matching point as the center and the preset initial distance as the radius, filter the fault points inside the circle, start from the matching point and connect with each group of fault points inside the circle, and mark the line segment connecting the matching point and each group of fault points inside the circle as the fault traction line.

[0085] Additional notes: If there are 0 fault points inside the circle, the initial distance will be redrawn according to the set expansion range until the number of fault points inside the circle is greater than or equal to 1.

[0086] Obtain the length of each faulty traction line as a similarity value for the fault points within each group of circles;

[0087] Starting from the current time point, count the number of times the fault points in each group of circles are matched within the set time window before the starting point, and use this as the similarity bias value of the fault points in each group of circles.

[0088] The number of matches in the number of fault point matches within each group of circles is analyzed. The number of matches indicates that after matching the corresponding historical fault case, the maintenance personnel go to the corresponding fault location and find that the current fault location matches the fault location of the corresponding historical fault case. If the current match is successful, the current match is considered a match. The proportion of the number of matches in the total number of matches is calculated as the confidence level value of the fault point within each group of circles.

[0089] The similarity value, similarity bias value, and confidence level value of the fault points within each group of circles are respectively labeled as follows: After marking, normalization is performed before inputting into the formula. Weighted calculations were performed to obtain the fault location index Tc for each group of fault points within the circle; where These are the weighting coefficients for the similarity score, similarity bias score, and confidence score, respectively.

[0090] After calculating the fault location index Tc of each group of fault points within the circle, the fault points with higher fault location index Tc in each group of fault points are selected, and the fault location is identified from the corresponding cases as the fault location result of the abnormal node with the judgment result of Class II abnormality level.

[0091] In addition, by mapping the electromagnetic anomaly index and the thermal imaging index in two dimensions, and combining the electromagnetic-thermal imaging dual-parameter matching of historical fault cases, a more comprehensive fault feature model is constructed; by spatial clustering of dual-parameter coordinate points, the location of similar composite faults in historical cases can be accurately matched, avoiding omissions or misjudgments by a single parameter.

[0092] By filtering, matching, and weighting historical cases in the knowledge database (similarity value, bias value, confidence value), real-time monitoring data is deeply bound to historical fault characteristics. For example, the reliability of historical cases is quantified by the "percentage of coincidences". High-confidence fault points are given priority reference, which significantly improves the accuracy of locating rare or complex faults (such as intermittent discharge accompanied by gradual temperature rise).

[0093] M4: Sort the abnormal nodes in the abnormal node list that belong to the second level of abnormality in descending order according to the fault location index Tc. After sorting, extract the abnormal behavior type of each group of abnormal nodes. Set an additional coefficient for electromagnetic field abnormality, temperature abnormality and comprehensive abnormality. The value range of the additional coefficient is set to 1.139-1.278, and the additional coefficient of comprehensive abnormality > the additional coefficient of electromagnetic field abnormality > the additional coefficient of temperature abnormality.

[0094] The fault location index Tc of each group of abnormal nodes is multiplied by the additional coefficient of the corresponding abnormal behavior type to obtain the priority index of each group of abnormal nodes. The sorting results are then adjusted based on the priority index of each group of abnormal nodes, that is, the sorting is re-sorted according to the priority index.

[0095] In addition, business experience of abnormal behavior types (overall abnormal risk > single abnormality) is converted into a calculable additional coefficient, which is coupled with the fault location index (Tc) to form a priority index that integrates "real-time matching degree" and "risk level".

[0096] For example, when the Tc indices of two groups of nodes are similar, the priority index of the abnormal node is processed first due to its higher additional coefficient, avoiding "risk misjudgment" caused by simply relying on historical matching degree (Tc), and making the operation and maintenance strategy more in line with the "risk priority" handling principle of transmission line faults;

[0097] Fault information push: For each group of abnormal nodes in the abnormal node list that belong to the second level of abnormality, the fault location results will be pushed to the maintenance personnel.

[0098] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for distributed fault location in transmission lines based on non-contact detection, characterized in that, include: Fault characteristic analysis: Each non-contact detection node collects electromagnetic field intensity and temperature distribution data of the transmission line within the monitoring time window according to the set sampling frequency, and performs anomaly assessment calculation to determine the abnormal nodes and anomaly types within the current monitoring time window. The anomaly types include electromagnetic field anomalies, temperature anomalies, and comprehensive anomalies. The process of anomaly assessment calculation for electromagnetic field strength of transmission lines within the monitoring time window is as follows: Electromagnetic field intensity data for each non-contact detection node within the current monitoring time window are extracted. First, the amplitude standard deviation is calculated using the formula... The standard deviation of the amplitude was calculated. ,in To monitor the average electromagnetic field intensity at each time point within the time window, i is the time point number, and N is the total number of time points; Preset and The corresponding weighting coefficients, for the transmission lines within the current monitoring time window. and Normalization is performed separately, and each normalized value is multiplied by the corresponding set weight coefficient and then summed. The summation result is divided by an integer two to obtain the time-domain component evaluation value of each non-contact detection node within the current monitoring time window. The signal spectrum of the electromagnetic field intensity signal is extracted using Fast Fourier Transform, and the harmonic distortion, abnormal frequency energy ratio, and spectral entropy are extracted from the signal spectrum. The weighting coefficients corresponding to the preset harmonic distortion, abnormal frequency energy ratio and spectral entropy are normalized. After normalization, they are multiplied by the corresponding set weighting coefficients and then summed. The summation result is divided by an integer three to obtain the frequency domain component evaluation value of each non-contact detection node within the current monitoring time window. Using the time-domain component evaluation value and the frequency-domain component evaluation value as the two legs of a right triangle, a right triangle is constructed, and the area of ​​the right triangle is obtained to obtain the electromagnetic anomaly index of each non-contact detection node within the current monitoring time window. The process of anomaly assessment calculation for temperature distribution data of transmission lines within the monitoring time window is as follows: Extract the temperature distribution data of each non-contact detection node within the current monitoring time window, and first calculate the temperature standard deviation using the formula. Calculate the temperature standard deviation ,in To monitor the average temperature at each time point within the time window; Preset and The corresponding weighting coefficients, for the current monitoring time window and Normalization is performed separately, and each normalized value is multiplied by the corresponding set weight coefficient and then summed to obtain the thermal image index of each non-contact detection node within the current monitoring time window. The abnormal nodes and their types within the current monitoring time window are identified as follows: The threshold indices corresponding to the preset electromagnetic anomaly index and thermal imaging index are set. If the electromagnetic anomaly index or thermal imaging index of a certain group of non-contact detection nodes is higher than the corresponding preset threshold index within the current monitoring time window, it is determined to be an abnormal node. If the electromagnetic anomaly index is higher than the corresponding preset threshold index, the anomaly type is determined to be electromagnetic field anomaly; if the thermal imaging index is higher than the corresponding preset threshold index, the anomaly type is determined to be temperature anomaly; if both the electromagnetic anomaly index and the thermal imaging index are higher than the corresponding preset threshold index, the anomaly type is determined to be comprehensive anomaly. Fault correlation and localization: Collect abnormal behavior nodes within the current monitoring time window, generate an abnormal node list, and determine the abnormality level of each group of abnormal behavior nodes in the abnormal node list; the abnormality level includes Class I abnormality level and Class II abnormality level. A knowledge database is built based on historical fault data. The corresponding steps are performed in conjunction with the knowledge database to locate faults in each group of abnormal nodes in the abnormal node list that belong to the second level of abnormality. The knowledge database stores historical fault data cases for different nodes, including abnormal behavior type, fault location and maintenance records.

2. The method for distributed fault location of transmission lines based on non-contact detection according to claim 1, characterized in that, The abnormality level of each group of abnormal nodes in the abnormal node list is determined as follows: M1: Identify the abnormal behavior type of each group of abnormal behavior nodes in the abnormal node list. If the abnormal behavior type is a comprehensive abnormality, it is directly determined to be a Class II abnormality level. If the abnormality type is electromagnetic field abnormality or temperature abnormality, the corresponding electromagnetic abnormality index or thermal imaging index is extracted, and the difference between it and the corresponding threshold index is calculated and recorded as the threshold degree difference. The threshold level reference difference is preset for electromagnetic field anomalies and temperature anomalies respectively. If the calculated threshold level difference is higher than the corresponding threshold level reference difference, it is directly judged as a Class II anomaly level; otherwise, it is judged as a Class I anomaly level.

3. The method for distributed fault location of transmission lines based on non-contact detection according to claim 2, characterized in that, If it is determined to be a Class I anomaly, then execute: M2: If a group of abnormal nodes is determined to be of an abnormal level, the electromagnetic anomaly index or thermal imaging index within the monitoring time window after the determination time point is used as the starting point. If it is still an abnormal node, step M1 is executed again to determine the abnormal level.

4. The method for distributed fault location of transmission lines based on non-contact detection according to claim 3, characterized in that, By combining the knowledge database and performing the corresponding steps, fault location is performed on each group of abnormal nodes belonging to the second-level abnormality in the abnormal node list. Specifically: M3: If a group of abnormal nodes is determined to be of Class II anomaly level, the abnormal behavior type of the abnormal node is first identified, and then the electromagnetic anomaly index and thermal imaging index are extracted as reference data. For each group of abnormal nodes with a Class II anomaly level, after identifying the node number, the historical fault cases corresponding to each node number are retrieved in the knowledge database. Cases that match the current abnormal behavior type are selected from the historical fault cases and retained as selected fault cases. Construct a Cartesian coordinate system, with the x-axis representing the electromagnetic anomaly index and the y-axis representing the thermal imaging index; construct coordinate points based on the electromagnetic anomaly index and thermal imaging index of the anomaly manifestation node, and plot them in the Cartesian coordinate system as matching points; construct coordinate points based on the electromagnetic anomaly index and thermal imaging index of each group of screened fault cases corresponding to the anomaly manifestation node, and plot them in the Cartesian coordinate system as fault points; Draw a circle with the matching point as the center and the preset initial distance as the radius, filter the fault points inside the circle, and calculate the fault location index Tc of each group of fault points inside the circle. After calculating the fault location index Tc of each group of fault points within the circle, the fault points with higher fault location index Tc are selected from the fault points within each group of fault points, and the fault locations are identified from the corresponding cases as the fault location results of the abnormal node with the judgment result of Class II abnormality level.

5. The method for distributed fault location of transmission lines based on non-contact detection according to claim 4, characterized in that, The specific process for calculating the fault location index Tc is as follows: Starting from the matching point, connect it to each group of fault points within the circle, and mark the line segment connecting the matching point to each group of fault points within the circle as the fault traction line. Obtain the length of each faulty traction line as a similarity value for the fault points within each group of circles; Starting from the current time point, count the number of times the fault points in each group of circles are matched within the set time window before the starting point, and use this as the similarity bias value of the fault points in each group of circles. The number of matches in the number of fault point matches within each group of circles is analyzed. The number of matches indicates that after matching the corresponding historical fault case, the maintenance personnel go to the corresponding fault location and find that the current fault location matches the fault location of the corresponding historical fault case. If the current match is successful, the current match is considered a match. The proportion of the number of matches in the total number of matches is calculated as the confidence level value of the fault point within each group of circles. The similarity value, similarity bias value, and confidence level value of the fault points within each group of circles are respectively labeled as follows: ; After labeling, normalization is performed before inputting into the formula. Weighted calculations were performed to obtain the fault location index Tc for each group of fault points within the circle; where These are the weighting coefficients for similarity value, similarity bias value, and confidence level value, respectively.

6. The method for distributed fault location of transmission lines based on non-contact detection according to claim 5, characterized in that, The construction process of each non-contact detection node is as follows: x non-contact detection nodes are deployed along the transmission line at predetermined intervals; the non-contact detection nodes include electromagnetic field strength sensors and infrared thermal imagers; the value of x is specifically set according to the length of the transmission line.

Images