A power transmission line fault prediction method and system based on prior knowledge

Abstract

The present application relates to the field of power transmission line fault prediction, and specifically discloses a power transmission line fault prediction method and system based on prior knowledge, the method comprising: obtaining a plurality of prediction sections of a power transmission line and section basic information of each prediction section; generating a channel fuel load label, a line insulation exposure label and a disposal accessibility label of the prediction section according to the section basic information; and generating a basic fire point threat value of the prediction section according to fire point information and meteorological information. The present application can hierarchically evaluate the basic threat of forest fire, the possibility of fire point spreading to the line corridor, the possibility of line failure affected by forest fire and the operation and maintenance priority, thereby improving the accuracy, interpretability and operation and maintenance applicability of the power transmission line forest fire fault prediction result.

Description

Technical Field

[0001] This invention belongs to the field of transmission line fault prediction technology, specifically relating to a method and system for predicting transmission line faults based on prior knowledge. Background Technology

[0002] Transmission lines, as crucial facilities in the power system responsible for long-distance power transmission, are typically characterized by long routes, complex terrain crossings, significant environmental variations along the lines, and substantial susceptibility to external environmental influences on their operation. Some transmission lines traverse mountainous areas, forest regions, agroforestry zones, canyons, and slopes. During dry seasons, periods of high temperature and low humidity, strong winds, or frequent human-caused fires, wildfires are prone to occur along the line corridors and in their vicinity. The flames, smoke, heat, and burning particles generated by wildfires can reduce air insulation strength, affecting the safe operation of conductors, insulators, hardware, and the line corridor. In severe cases, this can lead to discharges, tripping, or line faults.

[0003] A search revealed a method for dynamic prediction and early warning of wildfires along power transmission lines based on time-series changes, disclosed in publication number CN105787602B. This method constructs an elliptical model of wildfire spread based on the initial location of the wildfire point, wind direction, wind speed, and slope information. It also calculates the closest distance between the fire point and the power transmission line, as well as the spread speed of the fire point towards the power transmission line, thereby determining the early warning level for wildfires along power transmission lines. This method can dynamically predict the spread of wildfires towards the power transmission line, but its focus remains on the early warning judgment between the fire point location, spread speed, and distance to the power line.

[0004] The risk of transmission line fire faults is usually not determined by the distance of the fire point or the direction of fire spread. Rather, it includes at least several levels, such as whether the fire point itself poses a basic threat, whether the fire point can spread along continuous combustibles to the line corridor, whether the fire point will pose a line fault threat due to insufficient line clearance, insufficient distance between conductors and ground, insufficient distance between conductors and tree obstacles, or abnormal insulation condition after it approaches the line corridor, and how to determine the urgency of operation and maintenance when risks exist in multiple sections at the same time.

[0005] If existing methods for predicting the risk of wildfires on transmission lines combine the basic threat of the fire point, the possibility of the wildfire spreading to the line corridor, the possibility of the line being affected by the wildfire and the urgency of operation and maintenance into a single risk result, it is easy to lead to unclear boundaries between the judgment of line fault risk and the judgment of priority of disposal, which will affect the accurate generation of section-level operation and maintenance strategies. Summary of the Invention

[0006] The purpose of this invention is to provide a transmission line fault prediction method and system based on prior knowledge, in order to solve the problem that existing transmission line wildfire fault prediction methods mentioned in the background art tend to mix the basic threat of the fire point, the possibility of the wildfire spreading to the line corridor, the possibility of the line being affected by the wildfire and the urgency of operation and maintenance, resulting in unclear boundaries between line fault risk judgment and priority judgment, and the prediction results being difficult to accurately support section-level operation and maintenance.

[0007] To achieve the above objectives, this invention provides a method for predicting transmission line faults based on prior knowledge, comprising the following steps: Obtain multiple predicted sections of the transmission line and basic information of each predicted section; Based on the basic information of the section, generate channel fuel load tags, line insulation exposure tags, and accessibility tags for the predicted section; Based on wildfire ignition point information and meteorological information, a basic fire threat value for the predicted section is generated; The corridor propagation threat value is obtained by correcting the base fire point threat value using the channel fuel load tag; By using line insulation exposure tags to correct corridor propagation threat values, the effective threat value of line faults can be obtained; Based on the effective threat value of the line fault, the importance of the line, and the accessibility of the handling, a handling priority value is generated; Output the line fault risk level based on the effective threat value of the line fault, and output the handling priority level based on the handling priority value; The accessibility tag is used to adjust the priority value of the handling and does not participate in increasing the effective threat value of line faults.

[0008] In one embodiment, obtaining multiple predicted segments of a transmission line includes: The transmission line can be divided into multiple prediction sections according to the line corridor between adjacent towers, or the transmission line corridor can be divided into multiple prediction sections according to a preset length. For route segments that traverse forest areas, mountains, canyons, slopes, agroforestry transition zones, areas with a high incidence of historical fires, or where road accessibility is lower than the preset accessibility conditions, the corresponding route segments will be separately classified as independent prediction segments. Each predicted segment is associated with at least one of the following: segment number, starting and ending towers, segment centerline, segment corridor boundary, segment buffer zone, line voltage level, line importance, and segment accessibility.

[0009] In one embodiment, generating a predicted segment's channel fuel load tag, line insulation exposure tag, and disposal accessibility tag includes: Based on at least two of the following factors: the flammability of vegetation around the predicted section, the density of combustibles, the length of the continuous combustible zone between the fire point and the line corridor, the frequency of historical fire points, and the recent rainfall interval, a channel fuel load label is generated. Based on at least two of the following factors in the predicted section: line corridor clearance, conductor-to-ground distance, conductor-to-tree distance, degree of forest crossing, insulator condition, and historical wildfire trip records, generate line insulation exposure labels. Based on at least two of the following factors: road accessibility of the predicted section, estimated arrival time, terrain complexity, historical firefighting time, and distance to available emergency resources, a response accessibility label is generated.

[0010] Preferably, the generation of basic fire threat values ​​for the predicted segment includes: Obtain the location, intensity, duration, confidence level, wind direction, wind speed, temperature, humidity, and rainfall interval of the fire. Determine the distance relationship and buffer zone relationship between the fire point and the predicted section based on the fire point location; The approach trend of the fire point to the predicted section is determined based on the fire point location, the predicted section, and the wind direction; A base fire threat value is generated based on distance relationships, buffer zone relationships, fire intensity, fire duration, fire confidence, approach trend, and environmental dryness determined by temperature, humidity, and rainfall intervals.

[0011] Preferably, determining the approach trend of the fire point to the predicted segment includes: Determine the fire approach direction from the fire location to the nearest point in the prediction section or the center line of the prediction section; Calculate the angle between the approach direction of the fire point and the current wind direction, and determine this angle as the wind approach angle; When the wind approach angle is less than the preset approach angle threshold, it is determined that the fire point has a tendency to approach the predicted section. When the wind approach angle is greater than the preset deviation angle threshold, it is determined that the fire point has a tendency to deviate from the predicted section. When the wind approach angle is between the preset approach angle threshold and the preset divergence angle threshold, the approach trend of the fire point to the predicted section is determined to be uncertain.

[0012] In a preferred embodiment, the corridor propagation threat value is obtained by correcting the base fire point threat value using the corridor fuel load tag, including: When the channel fuel load label indicates that there is a continuous combustible zone between the fire point and the line corridor, and the basic fire point threat value reaches the preset threat threshold, the basic fire point threat value is positively corrected to obtain the corridor propagation threat value. When the fuel load label of the channel indicates that the continuous combustible zone between the fire point and the line corridor is interrupted by a road, water body, bare ground or firebreak, reduce the positive correction of the base fire threat value or make a negative correction of the base fire threat value. When the corridor fuel load label indicates that the density of combustibles around the predicted section is higher than a preset density threshold, and the fire point has a tendency to approach the predicted section, the corridor propagation threat value is increased.

[0013] In a preferred embodiment, the corridor propagation threat value is corrected using line insulation exposure tags to obtain an effective threat value for line faults, including: When the line insulation exposure label indicates that the line corridor clearance of the predicted section is lower than the preset clearance threshold, or the conductor-to-ground distance is lower than the preset safe distance threshold, or the conductor-to-tree distance is lower than the preset tree distance threshold, the corridor propagation threat value is positively corrected. When the line insulation exposure label indicates that the predicted section has historical wildfire tripping records, abnormal insulator conditions, or the degree of crossing forest areas exceeds the preset crossing threshold, the corridor propagation threat value is positively corrected. When the line insulation exposure label characterizes the predicted section, the line corridor clearance, conductor-to-ground distance, and conductor-to-tree distance all meet the preset safety conditions, the positive correction magnitude of the corridor propagation threat value is reduced, or the corridor propagation threat value is kept unchanged. Generate effective threat values ​​for line faults based on the revised corridor propagation threat values.

[0014] In a preferred embodiment, generating a disposal priority value includes: Generate fault risk components based on the effective threat value of line faults; Generate line influence components based on the line's importance; A processing timeliness component is generated based on the processing accessibility tag; A handling priority value is generated based on the fault risk component, the line impact component, and the handling timeliness component. Specifically, when the estimated arrival time of the predicted segment represented by the accessibility tag is longer than the preset arrival time limit, or when the distance of the available emergency resources is greater than the preset resource distance, the disposal priority value is increased.

[0015] In a preferred embodiment, after the output line fault risk level and the handling priority level, the following is also included: Section-level handling strategies are generated based on the line fault risk level and handling priority level; When the line fault risk level is low and the handling priority level is low, a continuous monitoring or remote review strategy is generated. When the risk level of a line fault is medium, or the priority level of handling is increased, a strategy of video re-verification, drone patrol or on-site inspection is generated. When the line fault risk level is high and the handling priority level is high, generate on-site duty, passage clearing, local emergency linkage or operation and maintenance emergency work order strategy. When the line fault risk level is extremely high, or the predicted section is an important transmission channel and the accessibility of handling is lower than the preset accessibility conditions, a dispatch-side risk warning, key personnel arrival, or continuous status tracking strategy is generated.

[0016] This invention also provides a transmission line fault prediction system based on prior knowledge, comprising: The segment acquisition module is used to acquire multiple predicted segments of the transmission line and the basic segment information of each predicted segment; The prior label generation module is used to generate channel fuel load labels, line insulation exposure labels, and accessibility labels for predicted sections based on the basic information of the sections. The fire threat generation module is used to generate basic fire threat values ​​for the predicted area based on wildfire fire point information and meteorological information. The corridor propagation threat generation module is used to correct the basic fire point threat value using the channel fuel load label to obtain the corridor propagation threat value; The effective threat generation module is used to correct corridor propagation threat values ​​using line insulation exposure tags to obtain effective threat values ​​for line faults. The priority generation module is used to generate priority values ​​based on the effective threat value of the line fault, the importance of the line, and the accessibility tag for handling. The results output module is used to output the line fault risk level based on the effective threat value of the line fault, and to output the handling priority level based on the handling priority value. The accessibility tag is used to adjust the priority value of the handling and does not participate in increasing the effective threat value of line faults.

[0017] This invention provides a method and system for predicting transmission line faults based on prior knowledge. Compared with existing technologies, it has the following advantages: This invention divides transmission lines into multiple prediction sections and generates channel fuel load tags, line insulation exposure tags, and disposal accessibility tags for each prediction section. This enables transmission line fault prediction in wildfire scenarios to be segmented by combining the channel environment, line exposure status, and disposal conditions of specific sections. This avoids using the distance of the entire line or a single fire point as the basis for judgment, and improves the location accuracy and interpretability of fault prediction results.

[0018] This invention decomposes the impact of wildfires on transmission lines into three levels: basic fire point threat, corridor propagation threat, and effective line fault threat. First, a basic fire point threat value is generated based on wildfire fire point information and meteorological information. Then, the corridor propagation threat value is obtained by correcting it with corridor fuel load tags. Finally, the effective line fault threat value is obtained by correcting it with line insulation exposure tags. This invention can separately determine the degree of danger of the fire point itself, the possibility of the fire point propagating to the line corridor, and the possibility of the fire point causing discharge, tripping, or insulation safety risks after approaching the line. This avoids directly equating the risk of wildfire spread with the risk of transmission line faults.

[0019] This invention, through the layered correction of channel fuel load tags and line insulation exposure tags, can distinguish between situations where the fire point is close but difficult to spread to the line corridor due to obstruction by roads, water bodies, bare ground, or firebreaks, and high-risk situations where the fire point is relatively far away but there is a continuous combustible zone and the line corridor clearance, conductor-to-ground distance, or conductor-to-tree distance are insufficient. This reduces misjudgments caused by relying solely on the distance to the fire point for early warning and improves the accuracy of wildfire fault risk assessment.

[0020] This invention uses accessibility tags to generate priority values ​​for handling, and clarifies that they do not participate in increasing the effective threat value of line faults. This allows for separate evaluation of fault occurrence risk and urgency of maintenance handling. The system can output the line fault risk level and handling priority level separately, which facilitates maintenance personnel to reasonably arrange handling measures such as remote verification, drone patrol, on-site inspection, passage clearing, on-site duty, local emergency response, or risk alerts from the dispatch side when multiple predicted sections have the risk of wildfires at the same time. Attached Figure Description

[0021] Figure 1 This is a system architecture diagram of a transmission line fault prediction system based on prior knowledge.

[0022] Figure 2 This is a flowchart of a transmission line fault prediction method based on prior knowledge.

[0023] Figure 3 A flowchart for generating channel fuel load tags, line insulation exposure tags, and disposal accessibility tags.

[0024] Figure 4 This is a flowchart illustrating the hierarchical correction process for base fire threat value, corridor propagation threat value, and effective line fault threat value.

[0025] Figure 5 A flowchart mapping line fault risk levels and handling priority levels to section-level handling strategies. Detailed Implementation

[0026] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. The described embodiments are only some embodiments of the present invention, and not all embodiments.

[0027] Based on the embodiments of the present invention, other embodiments obtained by those skilled in the art without creative effort are all within the scope of protection of the present invention.

[0028] In this invention, the transmission line can be an overhead transmission line, and the predicted section can be a line corridor section between adjacent towers, or a line corridor section divided according to a preset length, preset grid, or preset risk boundary; the wildfire hotspot information can come from satellite hotspot monitoring platforms, fixed video surveillance devices, drone inspection equipment, manual reporting terminals, third-party fire monitoring platforms, or power grid operation and maintenance platforms, and the source of the hotspot information does not constitute a limitation on the implementation of this invention.

[0029] Passage fuel load tags, line insulation exposure tags, and accessibility tags are all prior knowledge tags used for wildfire fault prediction on transmission lines. They can be stored in server databases, edge computing devices, transmission line operation and maintenance platforms, or cloud knowledge bases. Basic fire point threat value, corridor propagation threat value, effective line fault threat value, and response priority value can be represented by score value, level value, probability value, normalized value, or interval value.

[0030] Passage fuel load tags, line insulation exposure tags, and accessibility tags are all prior knowledge tags used for wildfire fault prediction on transmission lines. They can be stored in server databases, edge computing devices, transmission line operation and maintenance platforms, or cloud knowledge bases. Basic fire point threat value, corridor propagation threat value, effective line fault threat value, and response priority value can be represented by score value, level value, probability value, normalized value, or interval value.

[0031] Example 1: As Figure 1 and Figure 2 As shown, this embodiment provides a transmission line fault prediction method based on prior knowledge. This method can be executed by a transmission line fault prediction system, which can be deployed in a power grid operation and maintenance master station, a transmission line online monitoring platform, a transmission line inspection management platform, a cloud server, an edge computing gateway, or a computing environment formed by the collaboration of the above devices. The system receives basic data of transmission lines, wildfire hotspot data, meteorological data, channel environment data, and operation and maintenance resource data, and outputs the line fault risk level and handling priority level.

[0032] The system first acquires multiple predicted sections of the transmission line and basic information about each section. Predicted sections can be formed by the system based on tower coordinates, line direction, line corridor boundaries, and line management areas. Basic information about a section includes at least the section number, starting and ending towers, section centerline, section corridor boundaries, section buffer zone, line voltage level, line importance, corridor environment information, and section accessibility information. The section centerline can be formed by connecting the coordinates of adjacent towers. The section corridor boundaries can be formed by extending the line centerline outwards to both sides according to a preset corridor width. The section buffer zone can be formed by extending the section corridor boundaries outwards by a preset distance. The line importance can be determined based on information such as line voltage level, importance of the power supply area, whether it is an important transmission channel, whether it supplies critical loads, and whether there are alternative transmission paths. Section accessibility can be determined based on information such as road distance, terrain slope, vehicle accessibility, walking distance for personnel, historical arrival time, and emergency resource location.

[0033] Based on the basic information of the section, the system generates channel fuel load tags, line insulation exposure tags, and disposal accessibility tags for the predicted section. The channel fuel load tag is used to characterize the fuel conditions for the fire point to spread to the line corridor of the predicted section. The line insulation exposure tag is used to characterize the possibility of the fire point causing discharge, tripping, or insulation safety risks after spreading to or near the line corridor. The disposal accessibility tag is used to characterize the ease with which maintenance personnel or emergency resources can reach the corresponding predicted section and carry out disposal. In one specific implementation, the channel fuel load tag, line insulation exposure tag, and disposal accessibility tag are set to three levels: low, medium, and high, respectively. In another specific implementation, the above tags are set to normalized scores from 0 to 100.

[0034] Based on wildfire ignition information and meteorological information, the system generates basic fire threat values ​​for the predicted area. Wildfire ignition information includes ignition location, intensity, duration, and confidence level. Meteorological information includes wind direction, wind speed, temperature, humidity, and rainfall interval. The ignition location can be latitude and longitude coordinates or planar coordinates in a projected coordinate system. The ignition intensity can be determined by the ignition thermal radiation intensity, ignition area, ignition temperature, video flame recognition level, or manually reported level. The ignition confidence level can be determined by the ignition data source, the confidence level of the recognition model, the cross-validation results of multiple data sources, or the results of manual review.

[0035] The basic fire threat value represents the basic degree of threat posed by the fire point itself and its surrounding meteorological conditions to the predicted section before the introduction of channel fuel load tags and line insulation exposure tags. In one specific implementation, the system calculates the basic fire threat value based on the distance score from the fire point to the predicted section, the fire point intensity score, the fire point duration score, the fire point confidence score, the approach trend score, and the environmental dryness score. For example, the basic fire threat value B can be calculated as B=aD+bF+cT+dC+eM+fE; where D is the distance score, F is the fire point intensity score, T is the fire point duration score, C is the fire point confidence score, M is the approach trend score, E is the environmental dryness score, and a, b, c, d, e, and f are preset weights, and a+b+c+d+e+f=1.

[0036] In the above calculation process, the distance score D can be determined according to the shortest distance between the fire point and the boundary of the predicted corridor. In one example, when the distance from the fire point to the boundary of the predicted corridor is less than the first distance threshold, D is 80 to 100; when the distance from the fire point to the boundary of the predicted corridor is between the first and second distance thresholds, D is 50 to 80; when the distance from the fire point to the boundary of the predicted corridor is greater than the second distance threshold, D is 0 to 50. The fire point intensity score F can be determined according to the fire point thermal radiation intensity, fire point area, or artificial fire situation. The fire intensity level is determined as follows: when the fire intensity is high, F is 80 to 100; when the fire intensity is medium, F is 50 to 80; when the fire intensity is low, F is 0 to 50. The fire duration score T can be determined according to the length of time the fire is continuously monitored. When the fire duration is greater than the first duration threshold, T is 80 to 100; when the fire duration is between the first and second duration thresholds, T is 50 to 80; when the fire duration is less than the second duration threshold, T is 0 to 50.

[0037] The fire point confidence score C can be determined based on the credibility of the fire point source, the confidence of identification, and the results of multi-source verification. When the fire point confidence score is greater than or equal to 0.8, C is 80 to 100; when the fire point confidence score is greater than or equal to 0.5 and less than 0.8, C is 50 to 80; when the fire point confidence score is less than 0.5, C is 0 to 50. The environmental dryness score E can be determined based on temperature, humidity, and rainfall interval. When the temperature is higher than a preset temperature threshold, the humidity is lower than a preset humidity threshold, and the rainfall interval is longer than a preset rainfall interval threshold, E is 80 to 100; when some of the above conditions are met, E is 40 to 80; when none of the above conditions are met, E is 0 to 40. The above thresholds and weights can be obtained from historical wildfire data, preset by power grid operation and maintenance experience, or adjusted according to different regions, seasons, and voltage levels of lines.

[0038] When determining the approach trend score M, the system can first determine the approach direction of the fire point from its location to the nearest point or centerline of the predicted section, then calculate the angle between the approach direction and the current wind direction, and define this angle as the wind approach angle. When the wind approach angle is less than the preset approach angle threshold, the system determines that the fire point has a tendency to approach the predicted section, and M can be between 80 and 100. When the wind approach angle is greater than the preset divergence angle threshold, the system determines that the fire point has a tendency to diverge from the predicted section, and M can be between 0 and 40. When the wind approach angle is between the preset approach angle threshold and the preset divergence angle threshold, the system determines that the approach trend of the fire point to the predicted section is uncertain, and M can be between 40 and 70. The preset approach angle threshold can be 30 degrees, 45 degrees, or 60 degrees, and can also be set according to the local wildfire spread characteristics, terrain conditions, and wind speed. The preset divergence angle threshold can be 120 degrees, 135 degrees, or 150 degrees, and can also be adjusted according to the application area.

[0039] After generating a basic fire threat value, the system uses the channel fuel load label to correct the basic fire threat value, thus obtaining the corridor propagation threat value. The corridor propagation threat value represents the probability that a fire point will propagate to the predicted section of the line corridor under the current channel fuel conditions. In one specific implementation, when the channel fuel load label indicates that there is a continuous combustible zone between the fire point and the line corridor, and the basic fire threat value reaches a preset threat threshold, the system performs a positive correction on the basic fire threat value. The continuous combustible zone can refer to a continuous or nearly continuous distribution of trees, shrubs, grasslands, crops, or other combustibles between the fire point location and the line corridor. The preset threat threshold can be a medium-risk threshold corresponding to the basic fire threat value, or it can be a numerical threshold set by the system based on historical data.

[0040] In one specific implementation, the corridor propagation threat value P can be calculated as P=B×K1; where K1 is the corridor fuel load correction coefficient; when the corridor fuel load label is high and there is a continuous combustible strip between the fire point and the corridor, K1 is 1.2 to 1.5; when the corridor fuel load label is medium, K1 is 1.0 to 1.2; when the corridor fuel load label is low, or the continuous combustible strip is interrupted by roads, water bodies, bare ground, or firebreaks, K1 is 0.6 to 1.0; the system can limit P to the range of 0 to 100, and when the calculation result is greater than 100, 100 is taken as the corridor propagation threat value; in another implementation, the corridor propagation threat value P can also be calculated as P=B+α, where α is the corridor fuel load correction term.

[0041] When the fuel load label of the channel indicates that the density of combustibles around the predicted section is higher than the preset density threshold, and the fire point has a tendency to approach the predicted section, the system increases the corridor propagation threat value. Through the above processing, the system can distinguish whether the fire point itself is threatening from whether the fire point is easy to propagate to the line corridor, avoiding the direct judgment of the transmission line fault risk based solely on the distance of the fire point.

[0042] After obtaining the corridor propagation threat value, the system corrects the corridor propagation threat value using line insulation exposure tags to obtain the effective threat value of the line fault. The effective threat value of the line fault represents the effective threat level of the fire point propagating to or near the line corridor, causing discharge, tripping, or insulation safety risks in the corresponding predicted section. In one specific implementation, the effective threat value R of the line fault can be calculated according to R=P×K2, where K2 is the line insulation exposure correction coefficient. When the line insulation exposure tag is high, K2 is 1.2 to 1.5; when the line insulation exposure tag is medium, K2 is 1.0 to 1.2; when the line insulation exposure tag is low, and the line corridor clearance, conductor-to-ground distance, and conductor-to-tree distance all meet the preset safety conditions, K2 is 0.7 to 1.0. The system can limit R to the range of 0 to 100, and when the calculation result is greater than 100, 100 is taken as the effective threat value of the line fault. In another implementation, the effective threat value R of the line fault can also be calculated according to R=P+β, where β is the line insulation exposure correction term.

[0043] When the line insulation exposure tag indicates that the line corridor clearance in the predicted section is lower than the preset clearance threshold, the conductor-to-ground distance is lower than the preset safety distance threshold, or the conductor-to-tree distance is lower than the preset tree distance threshold, the system increases the effective threat value of the line fault. When the line insulation exposure tag indicates that the predicted section has a historical wildfire tripping record, an abnormal insulator condition, or a forest crossing degree higher than the preset crossing threshold, the system increases the effective threat value of the line fault. When the line corridor clearance, conductor-to-ground distance, and conductor-to-tree distance in the predicted section all meet the preset safety conditions, the system reduces the positive correction magnitude of the corridor propagation threat value or keeps the corridor propagation threat value unchanged. Through these two levels of correction, the basic fire point threat value, the corridor propagation threat value, and the effective threat value of the line fault correspond to the threat of the fire point itself, the threat of the fire point propagating to the line corridor, and the effective threat of the fire point causing a line fault, respectively.

[0044] The system generates a disposal priority value based on the effective threat value of the line fault, the importance of the line, and the accessibility label for disposal. The disposal priority value is used to indicate the order of disposal for each predicted section when multiple predicted sections have the risk of wildfire faults at the same time. In one specific implementation, the disposal priority value Q can be calculated according to Q=gR+hL+iS; where R is the fault risk component, L is the line impact component, S is the disposal timeliness component, and g, h, and i are preset weights, and g+h+i=1. The line impact component L can be determined according to the line voltage level, the importance of the power supply area, the degree of load impact, whether it is an important transmission channel, and whether there is an alternative power supply path. The disposal timeliness component S can be determined by the disposal accessibility label. When the disposal accessibility label is low accessibility, S is 80 to 100; when the disposal accessibility label is medium accessibility, S is 50 to 80; and when the disposal accessibility label is high accessibility, S is 0 to 50.

[0045] The accessibility tag is only used to adjust the handling priority value and does not participate in increasing the effective threat value of line faults; the handling timeliness component S only participates in the calculation of the handling priority value Q and does not participate in the calculation of the corridor propagation threat value P or the effective threat value of line faults R; even if a certain predicted section is difficult to handle, the system will not increase the effective threat value of line faults R, but only increase the handling priority value Q, so as to avoid conflating the evaluation of fault occurrence risk with the urgency of operation and maintenance handling.

[0046] The system outputs the line fault risk level based on the effective threat value of the line fault and the handling priority level based on the handling priority value. When R is greater than or equal to 80, the system outputs extremely high risk; when R is greater than or equal to 60 and less than 80, the system outputs high risk; when R is greater than or equal to 40 and less than 60, the system outputs medium risk; and when R is less than 40, the system outputs low risk. The system outputs the handling priority level based on the handling priority value Q. When Q is greater than or equal to 80, the system outputs emergency priority; when Q is greater than or equal to 60 and less than 80, the system outputs high priority; when Q is greater than or equal to 40 and less than 60, the system outputs medium priority; and when Q is less than 40, the system outputs low priority. The above thresholds are only examples, and the system can be adjusted according to the line voltage level, regional wildfire frequency, historical fault data, operation and maintenance resource configuration, and power grid operation requirements.

[0047] For example, for a certain predicted section, when the effective threat value of the line fault is high but the distance to the disposal resources is relatively short, the system can output high risk and medium priority; for another predicted section, when the effective threat value of the line fault is medium risk but the line is of high importance and the accessibility of disposal is poor, the system can output medium risk and high priority, thereby making the output results more consistent with the actual disposal logic in the operation and maintenance of transmission lines.

[0048] Example 2: As Figures 2-5As shown, this embodiment further explains the prediction segment division, the generation of three types of prior labels, and the segment-level handling strategy mapping based on embodiment 1.

[0049] When dividing the prediction sections, the transmission line can be divided into multiple prediction sections according to the line corridor between adjacent towers. For example, the conductor corridor between tower A and tower B can be considered as one prediction section. Alternatively, the transmission line corridor can be divided into multiple prediction sections according to a preset length, such as 500 meters, 1000 meters, 2000 meters, or other set lengths. For areas with significant changes in terrain or risk environment, a non-equidistant method can be used to divide the prediction sections. For example, for line sections that cross forest areas, mountains, canyons, slopes, agroforestry transition zones, historically high-risk fire zones, or where road accessibility is lower than the preset accessibility conditions, the system can divide the corresponding line sections into independent prediction sections.

[0050] Each predicted segment can be associated with at least one of the following information: segment number, starting and ending towers, segment centerline, segment corridor boundary, segment buffer zone, line voltage level, line importance, and segment accessibility. The segment buffer zone can include a primary buffer zone, a secondary buffer zone, and a tertiary buffer zone. The primary buffer zone can be used to represent a high-concern area where the fire point is close to the line corridor, the secondary buffer zone can be used to represent a medium-concern area where the fire point needs to be continuously tracked, and the tertiary buffer zone can be used to represent a peripheral concern area where the fire point may affect the line. The distance between the above buffer zones can be determined based on the line voltage level, terrain conditions, vegetation type, and historical fire point spread rate. After the predicted segments are divided, the predicted segments will be used as the basic calculation unit for subsequent tag generation, fire point matching, threat value calculation, and response strategy generation.

[0051] When generating the three types of prior labels, basic information of the section can be read from the basic line database, remote sensing vegetation database, geographic information database, historical fire point database, operation and maintenance work order database, and emergency resource database. For the channel fuel load label, the vegetation combustibility, combustible density, continuous combustible belt length between the fire point and the line corridor, historical fire point frequency, and recent rainfall interval can be read from the surrounding vegetation combustibility, combustible density, continuous combustible belt length, and recent rainfall interval. Different scores can be preset according to the types such as coniferous forest, shrub, grassland, crops, broad-leaved forest, and bare land. Combustible density can be determined by remote sensing vegetation index, forestry survey data, inspection records, or manually labeled data. The continuous combustible belt length can be identified by the geographic information system as the distance of combustible connection between the fire point and the line corridor that is not cut off by roads, water bodies, bare land, or firebreaks. The historical fire point frequency can be obtained by statistically analyzing the fire point records in the predicted section and its neighboring areas within a preset time range. The recent rainfall interval can represent the time length since the most recent effective rainfall.

[0052] In one specific implementation, the system generates a channel fuel load label based on at least two of the following factors: vegetation combustibility, combustible density, continuous combustible strip length, historical fire frequency, and recent rainfall interval. When the vegetation type surrounding the predicted section is coniferous forest, shrubs, or dry grassland, and the combustible density is higher than a first density threshold, the system increases the channel fuel load score. When a continuous combustible strip exists between the fire point and the line corridor, and the length of the continuous combustible strip is greater than a first length threshold, the system increases the channel fuel load score. When the combustible strip between the fire point and the line corridor is interrupted by roads, water bodies, bare land, or firebreaks, the system decreases the channel fuel load score. When the historical fire frequency in the corresponding predicted section or its vicinity within a preset time range is higher than a first frequency threshold, or the recent rainfall interval is longer than a first rainfall interval threshold, the system increases the channel fuel load score.

[0053] The system can generate a channel fuel load score based on the above data and map the score to low, medium, and high channel fuel load labels. For example, the channel fuel load score V can be calculated as V=p1V1+p2V2+p3V3+p4V4+p5V5, where V1 is the vegetation combustibility score, V2 is the combustible density score, V3 is the continuous combustible zone length score, V4 is the historical fire frequency score, V5 is the recent rainfall interval score, and p1 to p5 are preset weights. When V is greater than or equal to 80, the system will mark the predicted section as a high channel fuel load label; when V is greater than or equal to 50 and less than 80, the system will mark the predicted section as a medium channel fuel load label; when V is less than 50, the system will mark the predicted section as a low channel fuel load label. The above thresholds can be adjusted according to different regional vegetation types, historical wildfire data, and route corridor management requirements.

[0054] For line insulation exposure tags, the system can read the line corridor clearance, conductor-to-ground distance, conductor-to-tree distance, degree of forest crossing, insulator status, and historical wildfire trip records. The line corridor clearance can be determined by inspection measurement data, lidar point cloud data, UAV image recognition results, or manual inspection records. The conductor-to-ground distance can be determined by line design data, sag calculation data, on-site distance measurement data, or a 3D line model. The conductor-to-tree distance can be determined by vegetation height recognition results, line spatial model, and on-site inspection records. Historical wildfire trip records can be determined by dispatch accident records, protection action records, and line fault databases. The insulator status can include at least one of the following: insulator pollution level, infrared temperature rise, leakage current, discharge traces, damage records, or defect work orders. When the insulator pollution level is higher than the preset pollution level, the infrared temperature rise is abnormal, the leakage current is abnormal, there are discharge traces, there are damage records, or there are defect work orders, the system increases the line insulation exposure score.

[0055] In one specific implementation, the system generates a line insulation exposure label based on at least two of the following factors: line corridor clearance, conductor-to-ground distance, conductor-to-tree distance, degree of forest crossing, insulator condition, and historical wildfire trip records. When the line corridor clearance is lower than a preset clearance threshold, the conductor-to-ground distance is lower than a preset safety distance threshold, or the conductor-to-tree distance is lower than a preset tree distance threshold, the system increases the line insulation exposure score. When the predicted section has historical wildfire trip records, abnormal insulator condition, or a degree of forest crossing exceeding a preset crossing threshold, the system increases the line insulation exposure score.

[0056] The system can generate a line insulation exposure score based on the above data and map the score to low, medium, and high line insulation exposure labels. For example, the line insulation exposure score I can be calculated as I = q1I1 + q2I2 + q3I3 + q4I4 + q5I5 + q6I6, where I1 is the line corridor clearance score, I2 is the conductor-to-ground distance score, I3 is the conductor-to-tree barrier distance score, I4 is the forest crossing score, I5 is the insulator status score, I6 is the historical wildfire tripping record score, and q1 to q6 are preset weights. When I is greater than or equal to 80, the system will mark the predicted section as a high line insulation exposure label; when I is greater than or equal to 50 and less than 80, the system will mark the predicted section as a medium line insulation exposure label; and when I is less than 50, the system will mark the predicted section as a low line insulation exposure label.

[0057] For accessibility tags, the system can read road accessibility, estimated arrival time, terrain complexity, historical firefighting time, and distance to available emergency resources. Road accessibility can be determined based on road grade, road distance, passable vehicle types, and historical road closure records. Estimated arrival time can be estimated based on the location of the maintenance outpost, road distance, terrain obstacles, traffic conditions, and historical arrival times. Terrain complexity can be determined based on factors such as slope, altitude changes, forest density, and the presence of canyons or waterways. The distance to available emergency resources can be determined based on the location of drone airports, emergency vehicles, fire brigades, maintenance teams, or local emergency resources.

[0058] In one specific implementation, the system generates a response accessibility label based on at least two of the following factors: road accessibility, estimated arrival time, terrain complexity, historical firefighting time, and distance to available emergency resources. When the predicted road accessibility is poor, the estimated arrival time is longer than the preset arrival time limit, the terrain complexity is high, the historical firefighting time is longer than the preset response time limit, or the distance to available emergency resources is greater than the preset resource distance, the system lowers the response accessibility evaluation result. The system can generate a response accessibility score based on the above data and map the score to a high, medium, or low accessibility label. For example, the response accessibility score S can be calculated as S = r1S1 + r2S2 + r3S3 + r4S4 + r5S5; where S1 is the road accessibility score, S2 is the estimated arrival time score, S3 is the terrain complexity score, S4 is the historical firefighting time score, S5 is the distance to available emergency resources score, and r1 to r5 are preset weights. When S is greater than or equal to 80, the system marks the predicted segment as a highly reachable segment; when S is greater than or equal to 50 and less than 80, the system marks the predicted segment as a medium reachable segment; when S is less than 50, the system marks the predicted segment as a low reachable segment.

[0059] In another implementation, if a certain label can only obtain one original factor, the system can call the default compensation factor of the area where the predicted segment is located to complete the remaining scoring items. The default compensation factor can be determined by historical wildfire data of the same area, statistical data of similar line segments, or manually preset empirical values. The above labels can be updated periodically or triggered after acquiring new remote sensing data, inspection records, fire point records, or changes in operation and maintenance resources. Through the above scoring rules, the three types of prior labels are not simply text labels, but callable calculation results formed by rule calculation of line segment data, environmental data, historical data, and operation and maintenance data.

[0060] After outputting the line fault risk level and handling priority level, the system can generate a segment-level handling strategy based on the combination of the two. The system can set up a handling strategy library, which stores the mapping relationship between the line fault risk level, handling priority level, and segment-level handling strategy. The strategy templates in the handling strategy library can be pre-configured by the operation and maintenance platform or updated based on historical handling results. When generating a handling strategy, the system can output the execution object, execution action, execution time limit, review content, and feedback requirements, so that the handling strategy can be called by the operation and maintenance platform, inspection work order system, mobile terminal, drone inspection platform, or dispatch monitoring system.

[0061] When the line fault risk level is low and the handling priority level is low, the system invokes the continuous monitoring strategy, which includes updating the fire location, fire intensity, wind direction and speed, and predicted section status according to the first refresh cycle; when the line fault risk level is low but the handling priority level is medium or high, the system invokes the remote verification strategy, which includes verifying the fire status by calling fixed video surveillance, drone imagery, or satellite hotspot refresh data.

[0062] When the line fault risk level is medium, or the handling priority level is increased, a video redirection verification, drone inspection, or on-site verification strategy is generated. The video redirection verification strategy includes sending a redirection command to the rotatable camera to make the camera face the fire point or the predicted section corridor. The drone inspection strategy includes sending the predicted section coordinates, fire point coordinates, inspection route, shooting angle, and feedback requirements to the drone inspection platform. The on-site verification strategy includes generating an operation and maintenance work order, which includes the handling location, handling object, estimated arrival time, verification content, and feedback requirements.

[0063] When the line fault risk level is high and the handling priority level is high, a work order strategy is generated for on-site duty, passage clearing, local emergency response, or maintenance emergency response. The on-site duty strategy includes sending the predicted section location, fire location, risk level, handling priority level, and on-site monitoring requirements to maintenance personnel. The passage clearing strategy includes checking combustibles, tree obstructions, temporary storage materials, and external obstacles affecting passage safety within the line corridor. The local emergency response strategy includes pushing the fire location, line name, predicted section, risk level, and handling suggestions to fire, forestry, or local emergency departments.

[0064] When the line fault risk level is extremely high, or the predicted section is an important transmission channel and the accessibility for handling is lower than the preset accessibility conditions, a dispatch-side risk alert, key personnel deployment, or continuous status tracking strategy is generated. The dispatch-side risk alert includes the line name, predicted section, fire point number, line fault risk level, handling priority level, and possible impact range. The key personnel deployment strategy includes notifying the operation and maintenance manager, emergency team, or relevant professionals to arrive at the designated location. The continuous status tracking strategy includes shortening the fire point data refresh cycle, increasing the frequency of drone patrols, continuously monitoring wind direction changes, and continuously updating the effective threat value of line faults.

[0065] Example 3: Figure 1 As shown, this embodiment provides a transmission line fault prediction system based on prior knowledge. The system includes a section acquisition module, a priori tag generation module, a fire point threat generation module, a corridor propagation threat generation module, an effective threat generation module, a handling priority generation module, and a result output module.

[0066] The segment acquisition module is used to acquire multiple predicted segments of the transmission line and the basic information of each predicted segment. Specifically, the segment acquisition module receives the coordinates of the transmission line towers, the line corridor boundary, the line voltage level, historical line data, and geographic information data, and generates multiple predicted segments based on adjacent tower intervals, preset lengths, or risk boundaries. The prior label generation module is connected to the segment acquisition module and is used to generate channel fuel load labels, line insulation exposure labels, and accessibility labels for the predicted segments based on the segment basic information. Specifically, the prior label generation module receives data from the line basic database, remote sensing vegetation database, geographic information database, historical fire point database, maintenance work order database, and emergency resource database, and generates three types of prior labels according to preset label generation rules.

[0067] The fire point threat generation module generates a basic fire point threat value for the predicted section based on wildfire fire point information and meteorological information. Specifically, the fire point threat generation module receives fire point location, fire point intensity, fire point duration, fire point confidence level, wind direction, wind speed, temperature, humidity, and rainfall interval, and calculates the distance relationship, buffer relationship, and approach trend between the fire point and the predicted section to generate the basic fire point threat value. The corridor propagation threat generation module is connected to the fire point threat generation module and the prior label generation module. It is used to correct the basic fire point threat value using the corridor fuel load label to obtain the corridor propagation threat value. Specifically, the corridor propagation threat generation module receives the basic fire point threat value and the corridor fuel load label, and corrects the basic fire point threat value based on conditions such as the existence of a continuous combustible zone, whether the continuous combustible zone is interrupted, and whether the density of surrounding combustibles exceeds a preset density threshold.

[0068] The effective threat generation module is connected to the corridor propagation threat generation module and the prior tag generation module. It is used to correct the corridor propagation threat value using line insulation exposure tags to obtain the effective threat value for the line fault. Specifically, the effective threat generation module receives the corridor propagation threat value and the line insulation exposure tag, and corrects the corridor propagation threat value based on conditions such as line corridor clearance, conductor-to-ground distance, conductor-to-tree distance, historical wildfire tripping records, insulator status, and the extent of forest crossing. The disposal priority generation module is connected to the effective threat generation module and the prior tag generation module. It is used to generate a disposal priority value based on the effective threat value for the line fault, the line importance, and the disposal accessibility tag. Specifically, the disposal priority generation module receives the effective threat value for the line fault, the line importance, and the disposal accessibility tag, and generates a fault risk component, a line impact component, and a disposal timeliness component, respectively. The disposal priority value is then generated from these components.

[0069] The result output module is connected to both the effective threat generation module and the handling priority generation module. It outputs the line fault risk level based on the effective threat value and the handling priority level based on the handling priority value. The result output module can also output the line fault risk level and handling priority level to the transmission line operation and maintenance platform, dispatch monitoring system, inspection work order system, mobile terminal, or drone inspection platform. These modules can be implemented by the same processor executing different program segments stored in memory, or they can be implemented collaboratively by multiple processors, servers, edge computing gateways, transmission line monitoring terminals, inspection management platforms, or mobile terminals. Data can be transmitted between these modules via wired communication networks, wireless communication networks, dedicated power grids, the Internet, or internal data buses.

[0070] Example 4: This example describes the system's data storage, rule invocation, operation mode, and typical application process.

[0071] In one specific implementation, the system includes a basic line database, a geographic information database, a fire point monitoring database, a meteorological database, a historical fault database, a tag database, a rule database, and a response strategy database. The basic line database stores tower coordinates, line alignment, line voltage level, line importance, line corridor boundaries, and conductor spatial parameters; the geographic information database stores vegetation type, road distribution, water body distribution, bare land distribution, slope aspect, and forest area range; the fire point monitoring database stores fire point location, fire point intensity, fire point duration, fire point confidence level, and fire point source; and the meteorological database stores wind direction, wind speed, temperature, humidity, rainfall, and rainfall interval.

[0072] The historical fault database stores historical wildfire trip records, protection action records, insulator anomaly records, and historical handling records; the tag database stores channel fuel load tags, line insulation exposure tags, and handling accessibility tags for each predicted section; the rule database stores basic fire threat value calculation rules, channel fuel load correction rules, line insulation exposure correction rules, handling priority generation rules, risk level classification rules, and priority level classification rules; and the handling strategy database stores handling strategy templates corresponding to different combinations of line fault risk levels and handling priority levels.

[0073] When the system is running, the segment acquisition module first obtains line data from the line basic database and geographic information database and generates predicted segments; the prior label generation module then reads label data from the label library or related databases. When there is no label for the corresponding predicted segment in the label library, or when the label update time exceeds the preset update cycle, the prior label generation module recalculates the channel fuel load label, line insulation exposure label, and disposal accessibility label.

[0074] The fire threat generation module obtains real-time fire point information and meteorological information from the fire point monitoring database and meteorological database, and calculates the basic fire point threat value for each predicted section affected by a fire point; the corridor propagation threat generation module calls the channel fuel load correction rule in the rule base and combines it with the channel fuel load label to generate the corridor propagation threat value; the effective threat generation module calls the line insulation exposure correction rule in the rule base and combines it with the line insulation exposure label to generate the effective threat value of the line fault; the disposal priority generation module calls the disposal priority generation rule in the rule base and combines it with the line importance and disposal accessibility label to generate the disposal priority value; the result output module calls the risk level classification rule and priority level classification rule to generate the line fault risk level and disposal priority level, and can generate the section-level disposal strategy according to the strategy template in the disposal strategy library.

[0075] In a typical application, a predicted section A traverses a forest area. A continuous forest belt exists between the fire point and the power line corridor, resulting in insufficient corridor clearance and poor road accessibility. Based on the basic information of predicted section A, the system generates a high-channel fuel load label, a high-line insulation exposure label, and a low-response accessibility label. When the system detects a fire point located within the outer buffer zone of predicted section A, and that the fire intensity is high, the duration is long, and the wind direction is towards predicted section A, the system generates a high basic fire point threat value. Because predicted section A has a high-channel fuel load label, the system corrects the basic fire point threat value to a high corridor propagation threat value. Because predicted section A has a high-line insulation exposure label, the system further corrects the corridor propagation threat value to a high effective line fault threat value. Because predicted section A has a high line importance and low response accessibility, the system generates a high response priority value. The system outputs predicted section A as having a high line fault risk level and a high response priority level, and generates on-site monitoring, corridor clearing, local emergency response coordination, or maintenance emergency work order strategies.

[0076] In another typical application, the predicted fire point near section B is relatively close to the line, but there is water and bare ground between the fire point and the line corridor, and the line corridor has sufficient clearance. Although the system generates a certain basic fire point threat value based on the fire point distance and fire point intensity, the system reduces the corridor propagation threat value because the continuous combustible zone represented by the channel fuel load label is interrupted. Due to the low line insulation exposure label, the system does not significantly increase the effective line fault threat value. The system can output that the predicted section B is at a low or medium line fault risk level, and output the corresponding disposal priority level according to the accessibility of disposal and the importance of the line. Through the above process, the present invention can avoid misjudging the predicted section B as a high fault risk simply because the fire point is close.

[0077] The scoring values, weights, thresholds, correction coefficients, and strategy templates in the above embodiments can all be adjusted according to the voltage level of the transmission line, historical data of wildfires in the region, configuration of operation and maintenance resources, meteorological conditions, and management requirements of the line corridor. As long as the hierarchical processing logic of basic fire point threat value, corridor propagation threat value, effective threat value of line fault, and handling priority value is still adopted, they all belong to the implementation of this invention.

[0078] This invention divides transmission lines into multiple prediction sections and generates channel fuel load tags, line insulation exposure tags, and disposal accessibility tags for each prediction section. This enables transmission line fault prediction in wildfire scenarios to be segmented by combining the channel environment, line exposure status, and disposal conditions of specific sections. This avoids using the distance of the entire line or a single fire point as the basis for judgment, and improves the location accuracy and interpretability of fault prediction results.

[0079] This invention decomposes the impact of wildfires on transmission lines into three levels: basic fire point threat, corridor propagation threat, and effective line fault threat. First, a basic fire point threat value is generated based on wildfire fire point information and meteorological information. Then, the corridor propagation threat value is obtained by correcting it with corridor fuel load tags. Finally, the effective line fault threat value is obtained by correcting it with line insulation exposure tags. This invention can separately determine the degree of danger of the fire point itself, the possibility of the fire point propagating to the line corridor, and the possibility of the fire point causing discharge, tripping, or insulation safety risks after approaching the line. This avoids directly equating the risk of wildfire spread with the risk of transmission line faults.

[0080] This invention, through the layered correction of channel fuel load tags and line insulation exposure tags, can distinguish between situations where the fire point is close but difficult to spread to the line corridor due to obstruction by roads, water bodies, bare ground, or firebreaks, and high-risk situations where the fire point is relatively far away but there is a continuous combustible zone and the line corridor clearance, conductor-to-ground distance, or conductor-to-tree distance are insufficient. This reduces misjudgments caused by relying solely on the distance to the fire point for early warning and improves the accuracy of wildfire fault risk assessment.

[0081] This invention uses accessibility tags to generate priority values ​​for handling, and clarifies that they do not participate in increasing the effective threat value of line faults. This allows for separate evaluation of fault occurrence risk and urgency of maintenance handling. The system can output the line fault risk level and handling priority level separately, which facilitates maintenance personnel to reasonably arrange handling measures such as remote verification, drone patrol, on-site inspection, passage clearing, on-site duty, local emergency response, or risk alerts from the dispatch side when multiple predicted sections have the risk of wildfires at the same time.

[0082] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for predicting transmission line faults based on prior knowledge, characterized in that, Includes the following steps: Obtain multiple predicted sections of the transmission line and basic information of each predicted section; Based on the basic information of the section, generate channel fuel load tags, line insulation exposure tags, and accessibility tags for the predicted section; Based on wildfire ignition point information and meteorological information, a basic fire threat value for the predicted section is generated; The corridor propagation threat value is obtained by correcting the base fire point threat value using the channel fuel load tag; By using line insulation exposure tags to correct corridor propagation threat values, the effective threat value of line faults can be obtained; Based on the effective threat value of the line fault, the importance of the line, and the accessibility of the handling, a handling priority value is generated; Output the line fault risk level based on the effective threat value of the line fault, and output the handling priority level based on the handling priority value; The accessibility tag is used to adjust the priority value of the handling and does not participate in increasing the effective threat value of line faults.

2. The transmission line fault prediction method based on prior knowledge according to claim 1, characterized in that, Obtain multiple predicted sections of the transmission line, including: The transmission line can be divided into multiple prediction sections according to the line corridor between adjacent towers, or the transmission line corridor can be divided into multiple prediction sections according to a preset length. For route segments that traverse forest areas, mountains, canyons, slopes, agroforestry transition zones, areas with a high incidence of historical fires, or where road accessibility is lower than the preset accessibility conditions, the corresponding route segments will be separately classified as independent prediction segments. Each predicted segment is associated with at least one of the following: segment number, starting and ending towers, segment centerline, segment corridor boundary, segment buffer zone, line voltage level, line importance, and segment accessibility.

3. The transmission line fault prediction method based on prior knowledge according to claim 1, characterized in that, Generate channel fuel load tags, line insulation exposure tags, and disposal accessibility tags for the predicted sections, including: Based on at least two of the following factors: the flammability of vegetation around the predicted section, the density of combustibles, the length of the continuous combustible zone between the fire point and the line corridor, the frequency of historical fire points, and the recent rainfall interval, a channel fuel load label is generated. Based on at least two of the following factors in the predicted section: line corridor clearance, conductor-to-ground distance, conductor-to-tree distance, degree of forest crossing, insulator condition, and historical wildfire trip records, generate line insulation exposure labels. Based on at least two of the following factors: road accessibility of the predicted section, estimated arrival time, terrain complexity, historical firefighting time, and distance to available emergency resources, a response accessibility label is generated.

4. The transmission line fault prediction method based on prior knowledge according to claim 1, characterized in that, Generate the base fire threat value for the predicted section, including: Obtain the location, intensity, duration, confidence level, wind direction, wind speed, temperature, humidity, and rainfall interval of the fire. Determine the distance relationship and buffer zone relationship between the fire point and the predicted section based on the fire point location; The approach trend of the fire point to the predicted section is determined based on the fire point location, the predicted section, and the wind direction; A base fire threat value is generated based on distance relationships, buffer zone relationships, fire intensity, fire duration, fire confidence, approach trend, and environmental dryness determined by temperature, humidity, and rainfall intervals.

5. The transmission line fault prediction method based on prior knowledge according to claim 4, characterized in that, Determine the approach trend of the fire point to the predicted section, including: Determine the fire approach direction from the fire location to the nearest point in the prediction section or the center line of the prediction section; Calculate the angle between the approach direction of the fire point and the current wind direction, and determine this angle as the wind approach angle; When the wind approach angle is less than the preset approach angle threshold, it is determined that the fire point has a tendency to approach the predicted section. When the wind approach angle is greater than the preset deviation angle threshold, it is determined that the fire point has a tendency to deviate from the predicted section. When the wind approach angle is between the preset approach angle threshold and the preset divergence angle threshold, the approach trend of the fire point to the predicted section is determined to be uncertain.

6. The transmission line fault prediction method based on prior knowledge according to claim 1, characterized in that, By correcting the baseline fire point threat value using the corridor fuel load label, the corridor propagation threat value is obtained, including: When the channel fuel load label indicates that there is a continuous combustible zone between the fire point and the line corridor, and the basic fire point threat value reaches the preset threat threshold, the basic fire point threat value is positively corrected to obtain the corridor propagation threat value. When the fuel load label of the channel indicates that the continuous combustible zone between the fire point and the line corridor is interrupted by a road, water body, bare ground or firebreak, reduce the positive correction of the base fire threat value or make a negative correction of the base fire threat value. When the corridor fuel load label indicates that the density of combustibles around the predicted section is higher than a preset density threshold, and the fire point has a tendency to approach the predicted section, the corridor propagation threat value is increased.

7. The transmission line fault prediction method based on prior knowledge according to claim 1, characterized in that, By using line insulation exposure tags to correct corridor propagation threat values, effective line fault threat values ​​are obtained, including: When the line insulation exposure label indicates that the line corridor clearance of the predicted section is lower than the preset clearance threshold, or the conductor-to-ground distance is lower than the preset safe distance threshold, or the conductor-to-tree distance is lower than the preset tree distance threshold, the corridor propagation threat value is positively corrected. When the line insulation exposure label indicates that the predicted section has historical wildfire tripping records, abnormal insulator conditions, or the degree of crossing forest areas exceeds the preset crossing threshold, the corridor propagation threat value is positively corrected. When the line insulation exposure label characterizes the predicted section, the line corridor clearance, conductor-to-ground distance, and conductor-to-tree distance all meet the preset safety conditions, the positive correction magnitude of the corridor propagation threat value is reduced, or the corridor propagation threat value is kept unchanged. Generate effective threat values ​​for line faults based on the revised corridor propagation threat values.

8. The transmission line fault prediction method based on prior knowledge according to claim 1, characterized in that, Generate disposal priority values, including: Generate fault risk components based on the effective threat value of line faults; Generate line influence components based on the line's importance; A processing timeliness component is generated based on the processing accessibility tag; A handling priority value is generated based on the fault risk component, the line impact component, and the handling timeliness component. Specifically, when the estimated arrival time of the predicted segment represented by the accessibility tag is longer than the preset arrival time limit, or when the distance of the available emergency resources is greater than the preset resource distance, the disposal priority value is increased.

9. The transmission line fault prediction method based on prior knowledge according to claim 1, characterized in that, Following the output line fault risk level and handling priority level, the following are also included: Section-level handling strategies are generated based on the line fault risk level and handling priority level; When the line fault risk level is low and the handling priority level is low, a continuous monitoring or remote review strategy is generated. When the risk level of a line fault is medium, or the priority level of handling is increased, a strategy of video re-verification, drone patrol or on-site inspection is generated. When the line fault risk level is high and the handling priority level is high, generate on-site duty, passage clearing, local emergency linkage or operation and maintenance emergency work order strategy. When the line fault risk level is extremely high, or the predicted section is an important transmission channel and the accessibility of handling is lower than the preset accessibility conditions, a dispatch-side risk warning, key personnel arrival, or continuous status tracking strategy is generated.

10. A transmission line fault prediction system based on prior knowledge, characterized in that, include: The segment acquisition module is used to acquire multiple predicted segments of the transmission line and the basic segment information of each predicted segment; The prior label generation module is used to generate channel fuel load labels, line insulation exposure labels, and accessibility labels for predicted sections based on the basic information of the sections. The fire threat generation module is used to generate basic fire threat values ​​for the predicted area based on wildfire fire point information and meteorological information. The corridor propagation threat generation module is used to correct the basic fire point threat value using the channel fuel load label to obtain the corridor propagation threat value; The effective threat generation module is used to correct corridor propagation threat values ​​using line insulation exposure tags to obtain effective threat values ​​for line faults. The priority generation module is used to generate priority values ​​based on the effective threat value of the line fault, the importance of the line, and the accessibility tag for handling. The results output module is used to output the line fault risk level based on the effective threat value of the line fault, and to output the handling priority level based on the handling priority value. The accessibility tag is used to adjust the priority value of the handling and does not participate in increasing the effective threat value of line faults.

Images