Direct power grid risk assessment method

Abstract

The application provides a direct power transmission network risk assessment method, according to the line configuration information of a monitoring end, each monitoring line in a monitoring area is segmented to obtain a plurality of sub-line sections, and monitoring data collected by corresponding environmental monitoring equipment of each sub-line section is obtained; a sub-line section satisfying a completion condition is determined as a completion line section, completion data of the completion line section is obtained by calling a completion strategy, and environmental data of the corresponding sub-line section is determined according to the monitoring data or the completion data; icing analysis is performed on the environmental data based on an icing prediction model to obtain predicted icing data, and abnormal prediction data is obtained according to actual icing data collected by an inspection end and the predicted icing data; and the standard prediction data of the icing prediction model is adjusted and fed back according to the abnormal prediction data to obtain a trained prediction model.

Description

Technical Field

[0001] This invention relates to data processing technology, and more particularly to a method for risk assessment of direct power transmission networks. Background Technology

[0002] When transmission lines are covered by ice and snow, the impedance of the lines increases, which can lead to obstruction or interruption of power transmission. In addition, ice and snow above transmission lines may also cause short circuits, resulting in electrical fires or equipment failures. Therefore, monitoring ice accumulation on transmission lines is extremely important.

[0003] In the existing technology, when monitoring icing on transmission lines, it is usually done by manual inspection of the icing conditions on-site. However, since the length of the icing lines is very large, this method of monitoring icing is very labor-intensive and has low accuracy.

[0004] Therefore, improving the efficiency and accuracy of icing monitoring of transmission lines has become an urgent problem to be solved. Summary of the Invention

[0005] This invention provides a method for risk assessment of direct power transmission networks, which can improve the efficiency and accuracy of icing monitoring of transmission lines.

[0006] A first aspect of the present invention provides a method for risk assessment of direct transmission networks, comprising:

[0007] Based on the line configuration information of the monitoring terminal, each monitoring line in the monitoring area is segmented to obtain multiple sub-line segments, and the monitoring data collected by the environmental monitoring equipment corresponding to each sub-line segment is obtained.

[0008] Sub-segments that meet the completion conditions are identified as completion segments. The completion strategy is retrieved to obtain the completion data of the completion segments. The environmental data of the corresponding sub-segments are determined based on the monitoring data or completion data.

[0009] Based on the icing prediction model, the environmental data is analyzed to obtain predicted icing data. Based on the actual icing data and predicted icing data collected by the inspection terminal, anomaly prediction data is obtained.

[0010] Based on the abnormal prediction data, the standard prediction data of the icing prediction model is adjusted by feedback to obtain a trained prediction model.

[0011] Optionally, in one possible implementation of the first aspect, the sub-line segment that meets the completion conditions is determined as the completion segment, the completion strategy is invoked to obtain the completion data of the completion segment, and the environmental data of the corresponding sub-line segment is determined based on the monitoring data or the completion data, including:

[0012] When the monitoring data is missing in the sub-line segment, it is determined that the sub-line segment meets the completion condition. The monitoring data includes at least temperature data, humidity data, and wind speed data.

[0013] The sub-line segments that meet the completion conditions are identified as the completed line segments, and the monitoring map corresponding to the monitoring area is retrieved.

[0014] The monitoring line where the supplementary line segment is located in the monitoring map is identified as the target line, and a parallel line in the monitoring map that is pre-set to run parallel to the target line is determined.

[0015] The target line is subjected to same-line segment screening to obtain a first reference line segment, and the parallel lines are subjected to parallel line segment screening to obtain a second reference line segment;

[0016] The supplementary data of the supplementary line segment is determined based on the monitoring data of the first reference line segment and the second reference line segment, and the monitoring data or supplementary data is determined to be the environmental data of the corresponding sub-line segment.

[0017] Optionally, in one possible implementation of the first aspect, the target line is subjected to same-line segment filtering to obtain a first reference line segment, and the parallel lines are subjected to parallel-line segment filtering to obtain a second reference line segment, including:

[0018] The position points located at both ends of the supplemented line segment are obtained as reference points, and the position point in the target line that is at a preset reference distance from the corresponding reference point is determined as the termination point;

[0019] The termination point is located in a sub-segment of the target line other than the completed line segment;

[0020] The sub-line segment from the corresponding reference point to the termination point is determined as the first reference line segment;

[0021] Connect the two reference points to obtain the range delineation line, and determine the midpoint of the range delineation line as the reference point;

[0022] Generate a baseline perpendicular to the defined range line and passing through the reference point, and obtain the line segment distance between the midpoint and any of the reference points;

[0023] Two points on the baseline that are at a distance from the baseline point are respectively identified as connection points. The connection points and the two reference points are connected to obtain the parallel filtering range.

[0024] The sub-line segment within the parallel filtering range of the parallel lines is identified as the second reference line segment.

[0025] Optionally, in one possible implementation of the first aspect, after determining that the sub-line segment encompassed from the corresponding reference point to the termination point is the first reference line segment, the method further includes:

[0026] The center position point of each first reference line segment is obtained as the first positioning point, and the first reference line segment whose first positioning point is not in the line area corresponding to the reference point to the termination point is determined as the first line segment to be screened.

[0027] Determine the phase difference between the first positioning point and the termination point of the first line segment to be screened, and delete the first line segment to be screened whose phase difference is greater than the reference phase difference.

[0028] After determining that the sub-line segment within the parallel filtering range of the parallel lines is the second reference line segment, the method further includes:

[0029] The center point of each second reference line segment is obtained as the second positioning point, and the second reference line segment whose second positioning point is located outside the parallel screening range is determined as the second line segment to be screened.

[0030] Determine the shortest distance between the second positioning point of the second screening line segment and the parallel screening range, and delete the second screening line segment whose shortest distance is greater than the reference distance.

[0031] Optionally, in one possible implementation of the first aspect, determining the completion data of the completion segment based on the monitoring data of the first reference line segment and the second reference line segment includes:

[0032] The missing data attributes of the completed line segment are obtained as completion attributes, and the data attributes include at least temperature attributes, humidity attributes, and wind speed attributes;

[0033] The monitoring data corresponding to the first reference line segment and the second reference line segment and the completion attribute are determined as reference data;

[0034] The center point of the completed line segment is obtained as the target point, and the center points of the first reference line segment and the second reference line segment are obtained as the positioning points;

[0035] Based on the point distance between the positioning point and the target point, the line distance between each of the first reference line segment and the second reference line segment and the supplementary line segment is obtained;

[0036] Based on the ratio of the reference distance to the line distance, the influence coefficient corresponding to the first reference line segment or the second reference line segment is obtained;

[0037] The weighted sum is obtained by multiplying the corresponding influence coefficients and the reference data. The supplementary data of the supplementary line segment is determined based on the ratio of the weighted sum to the sum of the coefficients corresponding to each influence coefficient.

[0038] Optionally, in one possible implementation of the first aspect, icing analysis is performed on the environmental data based on an icing prediction model to obtain predicted icing data, and anomaly prediction data is obtained based on the actual icing data collected by the inspection terminal and the predicted icing data, including:

[0039] Based on the icing prediction model, standard icing intervals corresponding to each of the data attributes are obtained. The monitoring values ​​corresponding to environmental data with the same data attributes are compared with the standard icing intervals. The standard prediction data includes the standard icing intervals.

[0040] The sub-line segments whose monitored values ​​are all within the standard icing range are identified as predicted icing line segments, and the predicted icing line segments in the monitoring map are highlighted to obtain predicted icing data.

[0041] The actual inspection images collected by the inspection terminal for each of the sub-line segments are obtained, and the actual ice-covered line segments with ice are determined based on the actual inspection images.

[0042] Based on the actual icing line segment, actual icing data is obtained. If the sub-segment corresponding to the actual icing line segment and the predicted icing line segment are not the same, the corresponding sub-segment is determined to be an abnormal line segment, and the abnormal line segment is used as abnormal prediction data.

[0043] Optionally, in one possible implementation of the first aspect, the standard prediction data of the icing prediction model is adjusted based on the abnormal prediction data to obtain a trained prediction model, including:

[0044] Obtain the geographic information corresponding to the center location point of the abnormal line segment, and the geographic information includes at least the topographic features;

[0045] Based on the aforementioned landform features, the standard icing intervals corresponding to each of the aforementioned data attributes are adjusted to obtain the adjusted icing intervals. The aforementioned landform features include at least mountains.

[0046] The data attributes and the landform features are bound together, and the standard icing range of the data attributes under the landform features is updated to the adjusted icing range.

[0047] The icing prediction model is adjusted based on the adjusted icing interval to obtain a trained prediction model.

[0048] Optionally, in one possible implementation of the first aspect, the standard icing intervals corresponding to each of the data attributes are adjusted based on the geomorphic features to obtain adjusted icing intervals, wherein the geomorphic features at least include mountains, including:

[0049] The sub-segment corresponding to the actual icing line segment in the abnormal prediction data is determined to be a false negative line segment, and the sub-segment corresponding to the predicted icing line segment is determined to be a false positive line segment.

[0050] Obtain the icing trend corresponding to each of the data attributes of the false negative line segment, and determine the standard icing interval corresponding to the data attribute with the icing trend being a downward trend as the adjustment interval;

[0051] The standard icing interval corresponding to the data attribute indicating an upward icing trend is determined as the upward adjustment interval; or,

[0052] Obtain the non-icing trend corresponding to each of the data attributes of the false positive line segment, and determine the standard icing interval corresponding to the data attribute whose non-icing trend is a downward trend as the adjustment interval;

[0053] The standard icing interval corresponding to the data attribute where the non-icing trend is upward is determined as the upward adjustment interval;

[0054] Obtain historical monitoring data corresponding to the landform features, and adjust the upward or downward adjustment range based on the historical monitoring data to obtain the adjusted icing range corresponding to the data attributes.

[0055] Optionally, in one possible implementation of the first aspect, historical monitoring data corresponding to the landform feature is obtained, and the upward or downward adjustment interval is adjusted based on the historical monitoring data to obtain the adjusted icing interval corresponding to the data attribute, including:

[0056] When the data attribute corresponds to the upward adjustment interval, obtain the historical ice-free value corresponding to the data attribute under the landform feature, and determine the first mean of each historical ice-free value;

[0057] Obtain the first difference between the monitored value corresponding to the data attribute and the first mean, and obtain the upward adjustment coefficient based on the ratio of the first difference to the benchmark difference;

[0058] Obtain the first interval extreme value corresponding to the upward adjustment interval, obtain the upward adjustment extreme value based on the sum of the upward adjustment coefficient and the first interval extreme value, and obtain the adjustment icing interval corresponding to the data attribute based on the upward adjustment extreme value;

[0059] When the data attribute corresponds to the downward adjustment interval, the historical icing value corresponding to the data attribute under the landform feature is obtained, and the second mean of each historical icing value is determined. The historical monitoring data includes historical no-icing value and historical icing value.

[0060] Obtain the second difference between the monitoring value corresponding to the data attribute and the second mean, and adjust the coefficient downward according to the ratio of the second difference to the benchmark difference;

[0061] Obtain the second interval extreme value corresponding to the downward adjustment interval, obtain the downward adjustment extreme value based on the difference between the downward adjustment coefficient and the second interval extreme value, and obtain the adjustment icing interval corresponding to the data attribute based on the downward adjustment extreme value.

[0062] A second aspect of the present invention provides a direct transmission network risk assessment and processing system, comprising:

[0063] The segmentation module is used to segment each monitoring line in the monitoring area according to the line configuration information of the monitoring terminal to obtain multiple sub-line segments, and to obtain the monitoring data collected by the environmental monitoring equipment corresponding to each sub-line segment.

[0064] The retrieval module is used to determine the sub-line segments that meet the completion conditions as completion segments, retrieve the completion strategy to obtain the completion data of the completion segments, and determine the environmental data of the corresponding sub-line segments based on the monitoring data or completion data.

[0065] The analysis module is used to perform icing analysis on the environmental data based on the icing prediction model to obtain predicted icing data, and to obtain anomaly prediction data based on the actual icing data and predicted icing data collected by the inspection terminal.

[0066] The feedback module is used to adjust the standard prediction data of the icing prediction model based on the abnormal prediction data to obtain a trained prediction model.

[0067] A third aspect of the present invention provides a readable storage medium storing a computer program, which, when executed by a processor, is used to implement the first aspect of the present invention and various methods possibly involved in the first aspect.

[0068] The beneficial effects of this invention are as follows:

[0069] 1. This invention can accurately predict whether a transmission line will be iced based on relevant data, thereby remotely identifying iced lines and improving de-icing efficiency. First, this invention can divide each monitored line into multiple sub-segments, making the acquired monitoring data more consistent with the actual environment of the corresponding segment. This allows for more accurate predictions when predicting icing on sub-segments based on the monitoring data. Furthermore, this invention can use a completion strategy to supplement missing data in the completed segments with corresponding attribute data, followed by icing analysis, which improves the accuracy of icing prediction results and thus increases the efficiency of line de-icing. Finally, this invention allows the icing prediction model to be continuously trained on data, with feedback adjustments to obtain a trained prediction model, enabling more accurate predictions of line icing outcomes.

[0070] 2. This invention can complete the attribute data of missing data in the supplementary route segments to improve the accuracy of icing prediction results for sub-route segments. Specifically, this invention can filter out a first reference route segment and a second reference route segment based on the location information of the supplementary route segment, thereby obtaining the corresponding environmental data. Furthermore, based on the influence of the first and second reference route segments on the supplementary route segment, the corresponding environmental data is calculated with different weights, thereby improving the accuracy of the supplementary data and enhancing the accuracy of the icing prediction results.

[0071] 3. This invention can adjust the standard icing interval of corresponding attribute data according to different geographical information, so that the obtained adjusted icing interval can more accurately predict the icing result of the corresponding sub-line segment. At the same time, it can obtain training prediction models corresponding to different landform features, so as to facilitate the rapid transfer of training prediction models, thereby improving the efficiency of icing prediction and accelerating the speed of line de-icing operation. Attached Figure Description

[0072] Figure 1 A flowchart of a direct power transmission network risk assessment method provided by the present invention;

[0073] Figure 2 This is a schematic diagram of determining a first reference line segment provided by the present invention;

[0074] Figure 3 This is a schematic diagram of determining a second reference line segment provided by the present invention;

[0075] Figure 4 This is a schematic diagram of the structure of a direct power transmission network risk assessment and processing system provided by the present invention. Detailed Implementation

[0076] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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.

[0077] See Figure 1 This is a flowchart illustrating a direct power transmission network risk assessment method provided in an embodiment of the present invention. The power transmission network risk assessment method includes steps S1 to S4, as detailed below:

[0078] S1. Based on the line configuration information of the monitoring terminal, the monitoring lines in the monitoring area are segmented to obtain multiple sub-line segments, and the monitoring data collected by the environmental monitoring equipment corresponding to each sub-line segment is obtained.

[0079] It should be noted that during power transmission, there may be multiple transmission lines between two locations, rather than just one. Furthermore, the distance between the starting and ending points of power transmission may be considerable. Therefore, some sections of the same line may be covered in ice, while others may remain un-iced. The transmission line can be divided into several shorter sub-segments based on distance or the locations it passes through. This allows for subsequent data analysis of these sub-segments to determine whether they are covered in ice, quickly pinpointing the exact location of icy sections and facilitating de-icing operations.

[0080] It is easy to understand that since line icing is generally related to the environment in which the line is located, environmental monitoring equipment can be installed to collect environmental data on the line in order to ensure the normal operation of the line. For example, environmental monitoring equipment can collect environmental data such as temperature, humidity and wind speed of the line section, so that it can determine whether the line is icing based on the collected data, and thus understand the operation of the line.

[0081] Among them, the monitoring terminal is the information terminal of the personnel who conduct line monitoring, the line configuration information is the line segment information configured by the monitoring personnel for each line, the monitoring area is the area where icing monitoring is carried out, the monitoring line is the line where icing monitoring is carried out, the sub-line segment is the line segment obtained after dividing the monitoring line into segments, the environmental monitoring equipment is the equipment that collects environmental data, and the monitoring data is the data collected by the environmental monitoring equipment, which may include temperature data, humidity data and wind speed data, for example, the temperature is 26°, the humidity is 55% and the wind speed is 5m / s.

[0082] Through the above embodiments, the present invention can obtain real-time monitoring data of the line collected by environmental monitoring equipment, so as to predict whether the corresponding sub-line segment will be covered with ice based on the monitoring data.

[0083] S2, determine the sub-line segment that meets the completion conditions as the completion line segment, retrieve the completion strategy to obtain the completion data of the completion line segment, and determine the environmental data of the corresponding sub-line segment based on the monitoring data or completion data.

[0084] It should be noted that in practical applications, data may be missing from sub-line segments. For example, environmental monitoring equipment for a sub-line segment may be damaged due to harsh environments, resulting in missing data collected from that sub-line segment. For instance, to distinguish each sub-line segment, a corresponding number can be assigned to each sub-line segment. When the temperature monitoring sensor corresponding to sub-line segment number 8 is damaged, it can be determined that sub-line segment number 8 has missing data. Therefore, to improve the accuracy of subsequent icing prediction, data from relevant lines can be supplemented according to a supplementation strategy to ensure the completeness of environmental data for each sub-line segment.

[0085] Understandably, if any data in the monitoring data of a sub-line segment is missing, it can be determined that the sub-line segment meets the supplementation conditions, and the corresponding sub-line segment is used as the supplementation line segment. For example, if the temperature display data of sub-line segment No. 8 is missing, it can be determined that sub-line segment No. 8 meets the supplementation conditions, and sub-line segment No. 8 is used as the supplementation line segment.

[0086] Among them, the supplementary data refers to the data after supplementing the missing data for the corresponding line segment, and the environmental data refers to the relevant environmental data corresponding to each sub-line segment. When there is no missing data for a sub-line segment, its corresponding environmental data is the monitoring data; when there is missing data for a sub-line segment, its corresponding environmental data is the supplementary data.

[0087] Based on the above embodiments, step S2 (determining the sub-line segment that meets the completion conditions as the completion line segment, retrieving the completion strategy to obtain the completion data of the completion line segment, and determining the environmental data of the corresponding sub-line segment based on the monitoring data or completion data) can be implemented in the following ways:

[0088] S21, when the monitoring data of the sub-line segment is missing, it is determined that the sub-line segment meets the completion condition, and the monitoring data includes at least temperature data, humidity data and wind speed data.

[0089] Understandably, when any monitoring data for a sub-line segment is missing, i.e., the monitoring data is incomplete, it can be determined that the corresponding sub-line segment meets the conditions for data completion, i.e., the conditions for data completion can be met. This allows the data for the relevant sub-line segment to be completed, so that the predicted icing results can be more accurate.

[0090] S22, determine the sub-line segment that meets the completion conditions as the completion line segment, and retrieve the monitoring map corresponding to the monitoring area.

[0091] Understandably, the supplementary line segment refers to the line segment that meets the supplementary conditions, and the monitoring map is a map that shows the arrangement of the lines. For example, when there are 5 lines in the monitoring area, the monitoring map can show the arrangement of the corresponding 5 lines. Thus, the target line that needs to be supplemented with data and its corresponding parallel line can be determined through the monitoring map, so that the missing monitoring data of the supplementary line segment can be supplemented through the target line and parallel line.

[0092] S23, obtain the monitoring line where the supplementary line segment is located in the monitoring map as the target line, and determine the parallel line in the monitoring map that is pre-set to run parallel to the target line.

[0093] It is understandable that the target route is the monitoring route on which the supplementary route segment is located in the monitoring map. For example, when there are 5 routes in the monitoring map, and the supplementary route segment is on monitoring route 03 in the monitoring map, then the target route is monitoring route 03. The other routes in the monitoring map that are parallel to the target route are parallel routes. For example, monitoring routes 01, 02, 04, and 05 in the monitoring map are all parallel to monitoring route 03, so they can be considered as parallel routes.

[0094] In practical applications, parallel lines can generally be other lines located on the same high-voltage tower as the target line, or other lines located on adjacent high-voltage towers. These lines are generally close to the target line, and their corresponding environmental data are generally similar. Therefore, the missing data of the target line can be supplemented by the environmental data corresponding to the parallel lines. The parallel lines can be pre-configured, and after the target line is determined, the corresponding parallel lines can be directly obtained.

[0095] S24, the target line is screened for same-line segments to obtain a first reference line segment, and the parallel lines are screened for parallel-line segments to obtain a second reference line segment.

[0096] It's understandable that, since the supplementary line segment is located on the target line, and the environmental data of line segments on the same line may be quite similar, other sub-line segments on the same target line as the supplementary line segment can be filtered to obtain monitoring data corresponding to sub-line segments with similar environmental data. This provides data reference and improves the accuracy of subsequent data supplementation. It's also understandable that there may be multiple sub-line segments on parallel lines, and the distances between different sub-line segments and the supplementary line segment may vary. To filter out sub-line segments with data more closely similar to the supplementary line segment, the parallel lines can also be filtered accordingly.

[0097] The first reference line segment is a sub-line segment selected on the same target line as the supplementary line segment, and the second reference line segment is a second reference line segment obtained after selecting parallel line segments.

[0098] Through the above implementation methods, the present invention can obtain corresponding first reference line segment and second reference line segment, so that the subsequent obtained corresponding monitoring data is more consistent with the missing monitoring data of the line segment, making the preset results more accurate.

[0099] Based on the above embodiments, the specific implementation of step S24 (selecting the target line segment to obtain the first reference line segment and selecting the parallel line segment to obtain the second reference line segment) can be as follows:

[0100] S241, obtain the position points located at both ends of the supplemented line segment as reference points, and determine the position point in the target line that is at a preset reference distance from the corresponding reference point as the termination point.

[0101] The termination point is located in a sub-segment of the target line other than the line completion segment.

[0102] Understandably, the preset reference distance is used to determine the distance to nearby line segments, and is used to identify other line segments in the target line that are close to the supplementary line segment. This can be preset by the user based on the actual situation, and can be as low as 20m.

[0103] It's easy to understand that, in order to identify other line segments that are close to the completed line segment, the termination point is not within the completed line segment. Specifically, the termination point's location can be determined based on the reference point's position, for example... Figure 2 As shown, this is a schematic diagram of determining a first reference line segment provided by the present invention. By supplementing the reference point 1 on the left side of the line segment, the preset reference distance to the left can be determined, thereby determining the position of the left end point 1. By supplementing the reference point 2 on the right side of the line segment, the preset reference distance to the right can be determined, thereby determining the position of the right end point 2.

[0104] Through the above implementation methods, the present invention can obtain the corresponding termination point position, so as to subsequently determine the first reference line segment that meets the data requirements.

[0105] S242, determine the sub-line segment contained from the corresponding reference point to the termination point as the first reference line segment.

[0106] Through the above implementation methods, the present invention can identify the first reference road segment on the target route that is close to the supplementary road segment, thereby facilitating the subsequent acquisition of corresponding monitoring data and enabling accurate prediction of the missing monitoring data for the supplementary road segment.

[0107] Based on the above embodiment, after step S242 (determining the sub-line segment included from the corresponding reference point to the termination point as the first reference line segment), the method further includes:

[0108] S2421, obtain the center position point of each of the first reference line segments as the first positioning point, and determine the first reference line segment whose first positioning point is not in the line area corresponding to the reference point to the termination point as the first line segment to be screened.

[0109] It is understandable that the first positioning point is the center point of the first reference line segment, and the first line segment to be screened is the first reference line segment whose first positioning point is not located within the line area corresponding to the reference point to the end point.

[0110] For example: Figure 2 As shown, the center positions of the first reference line segment 1 and the first reference line segment 2 can be obtained as the first positioning point 1 and the first positioning point 2, respectively. If the first positioning point 1 and the first positioning point 2 are not in the line area between the reference point and the end point, the corresponding first reference line segment 1 and the first reference line segment 2 can be used as the first line segment to be screened, that is, the line segment waiting to be screened.

[0111] It is easy to understand that when the first positioning point is not within the route area corresponding to the reference point to the termination point, it means that only a small part of the corresponding first reference route segment may be within the route area included from the reference point to the termination point, while most of the route segment may be outside the route area corresponding to the reference point to the termination point. The collected environmental data corresponds to the entire route segment. In this case, there may be a discrepancy between the data of the corresponding route segment and the supplementary route segment. Therefore, the first reference route segment can be screened a second time to find the route segment with a higher degree of fit with the data of the supplementary route segment for subsequent icing prediction, thereby improving the accuracy of icing prediction.

[0112] S2422, determine the phase difference distance between the first positioning point and the termination point of the first screening line segment, and delete the first screening line segment whose phase difference distance is greater than the reference phase difference distance.

[0113] Understandably, when the distance between the first positioning point and the end point of the first line segment to be screened is relatively close and within the allowable baseline distance, it indicates that most of the line area of ​​the first line segment to be screened is close to the supplementary line segment, and the corresponding environmental data is similar to that of the supplementary line segment. When the distance is greater than the baseline distance, it indicates that most of the line area in the first line segment to be screened is far from the supplementary line segment, and the corresponding environmental data may also be deviated. Therefore, in order to ensure the accuracy of the supplementary data, the first line segment to be screened with a large distance can be deleted.

[0114] The phase difference distance is the distance between the first positioning point and the end point of the first screening line segment, and the benchmark phase difference distance is the threshold corresponding to the deviation distance between the first positioning point and the end point, which can be preset by humans.

[0115] S243, connect the two reference points to obtain the range delineation line, and determine the midpoint of the range delineation line as the reference point.

[0116] It is understandable that the range demarcation line is the line connecting two reference points, thus obtaining the benchmark point. The benchmark point and the range demarcation line are used to determine the screening range for the second reference line segment, so as to determine the specific second reference line segment.

[0117] The reference point is the midpoint of the boundary demarcation line, for example, such as... Figure 3 The diagram shown is a schematic diagram of a method for determining a second reference line segment provided by the present invention. The position of the reference point can be determined by the positions of two reference points.

[0118] S244, Generate a baseline perpendicular to the defined range line and passing through the reference point, and obtain the line segment distance between the midpoint and any of the reference points.

[0119] Understandably, in order to select parallel lines that meet the conditions and make the completion data corresponding to the line segments more accurate, the corresponding selection area can be determined by the target line, thereby obtaining the corresponding second reference line segment.

[0120] The baseline is a straight line that passes through the reference point and is perpendicular to the range demarcation line, and the line segment distance is the distance between the midpoint and any reference point.

[0121] Through the above implementation methods, the present invention can obtain the corresponding baseline and line segment distance, so as to subsequently determine the range for screening the second reference line segment.

[0122] S245, determine two position points on the baseline that are at a distance from the reference point as connection points, and connect the connection points and the two reference points to obtain the parallel filtering range.

[0123] It is understandable that the connection point is the location point that determines the parallel filtering range for the connection reference point, such as... Figure 3 As shown, connection points are determined at intervals of line segments from the reference point upwards and downwards. Each connection point is then connected to two reference points to obtain the enclosed area as the parallel filtering range.

[0124] Through the above implementation methods, the present invention can determine the parallel screening range, so as to select a suitable second reference line segment in the subsequent screening, thereby making the prediction results more accurate.

[0125] S246, determine the sub-line segment within the parallel filtering range of the parallel line as the second reference line segment.

[0126] Through the above implementation methods, the present invention can determine a second reference line segment, so as to obtain the corresponding environmental data in the future, thereby making the data for completing the line segment more accurate.

[0127] Based on the above embodiment, after step S246 (determining the sub-line segment within the parallel filtering range of the parallel line as the second reference line segment), the method further includes:

[0128] S2461, obtain the center position point of each of the second reference line segments as the second positioning point, and determine the second reference line segment whose second positioning point is located outside the parallel screening range as the second line segment to be screened.

[0129] It is understandable that the parallel screening range may include multiple second reference line segments. However, some second reference line segments may only be a small part of the parallel screening range, while the collected environmental data corresponds to the entire line segment. In this case, there may be a discrepancy between the data of the corresponding sub-line segment and the supplementary line segment. Therefore, the monitoring data of the corresponding sub-line segment may not match the environmental data of the actual supplementary line segment. Therefore, the second reference line segments can be screened a second time to obtain second reference line segments with more accurate environmental data, so as to improve the accuracy of subsequent icing prediction results.

[0130] It's easy to understand that the monitoring map shows multiple routes within the parallel filtering range for the completion route. The sub-segments of each route can be divided differently. For example, some routes may be short but traverse complex terrain, resulting in significantly different monitoring data. For instance, a route in mountainous terrain might pass both a mountaintop and a foothill, where temperature, humidity, and other environmental factors differ greatly. Therefore, the resulting sub-segments might be short. However, there might be parallel routes with longer distances within the filtering range. Similarly, if parallel routes traverse areas with similar environments, the monitoring data within the corresponding sub-segments might be identical, leading to longer sub-segments. Therefore, to more accurately select environmental data that matches the actual completion route segment, the second reference route segment is further filtered to ensure the accuracy of subsequent predictions.

[0131] The second positioning point is the center point of the second reference line segment, and the second line segment to be screened is the second reference line segment where the second positioning point is located outside the parallel screening range.

[0132] S2462, determine the shortest distance between the second positioning point of the second screening line segment and the parallel screening range, and delete the second screening line segment whose shortest distance is greater than the reference distance.

[0133] It is understandable that if the shortest distance between the second location point of the second route to be screened and the parallel screening range is greater than the baseline distance, it indicates that the distance between the corresponding second route segment to be screened and the supplementary route segment is relatively large. In this case, the sub-route segment within the parallel screening range is relatively short. Since the monitoring data corresponds to the environmental data of the longer route segment, there is a large deviation between the corresponding monitoring data and the environmental data of the supplementary route segment. Therefore, the corresponding second route segment to be screened can be deleted so that the error can be reduced to a greater extent when calculating the obtained data in the future, thereby improving the accuracy of the supplementary data.

[0134] It is easy to understand that the shortest difference distance is the shortest distance from the second positioning point to the parallel screening range.

[0135] S25, determine the supplementary data of the supplementary line segment based on the monitoring data of the first reference line segment and the second reference line segment, and determine the monitoring data or supplementary data as the environmental data of the corresponding sub-line segment.

[0136] Through the above-described embodiments, the present invention can obtain environmental data of the corresponding sub-line segment, so that it can predict whether the sub-line segment will be covered with ice based on the environmental data, thereby improving the protection against icing of high-voltage power lines and timely and accurate de-icing of iced lines.

[0137] Based on the above embodiments, step S25 (determining the supplementary data of the supplementary line segment based on the monitoring data of the first reference line segment and the second reference line segment) can be implemented in the following ways:

[0138] S251, obtain the missing data attributes of the line segment to be filled as the filling attributes, and the data attributes include at least temperature attributes, humidity attributes and wind speed attributes.

[0139] It is understandable that the data attributes are the attributes of the environmental data. For example, the attribute corresponding to temperature data is the temperature attribute, the attribute corresponding to humidity data is the humidity attribute, and the attribute corresponding to wind speed data is the wind speed attribute.

[0140] It is easy to understand that when the missing data in the line segment is a temperature value, the temperature attribute corresponding to the temperature data can be used as the missing attribute so that it can be matched with the data with the same attribute obtained later for the first and second reference line segments.

[0141] S252, determine the monitoring data corresponding to the first reference line segment and the second reference line segment and the completion attribute as reference data.

[0142] It is understandable that the reference data refers to the monitoring data of the first and second reference line segments with the same completion attributes.

[0143] For example, when the completion attribute is temperature, the temperature data of the first reference line segment and the second reference line segment can be used as reference data so that the temperature value corresponding to the temperature value of the completion line segment can be obtained later.

[0144] S253, obtain the center position point of the completed line segment as the target point, and the center position points of the first reference line segment and the second reference line segment as the positioning points.

[0145] It is understandable that the target point is the midpoint of the completed line segment, and the positioning point is the center point of the first and second reference line segments.

[0146] Through the above implementation methods, the present invention can determine the target points corresponding to the supplementary line segments and the positioning points corresponding to the first and second reference line segments, so as to obtain the distance between the points. This facilitates the determination of the proportion weight of the monitoring data corresponding to the first and second reference line segments based on the distance between the points, making the calculated data more accurate.

[0147] S254, based on the point distance between the positioning point and the target point, obtain the line distance between each of the first reference line segment and the second reference line segment and the supplementary line segment.

[0148] It is understandable that the point distance is the distance between the positioning point and the target point. Furthermore, there can be multiple first reference line segments and multiple second reference line segments. Therefore, the line distance is the distance between each first reference line segment, second reference line segment, and the supplementary line segment.

[0149] For example, when the distance between the positioning point on the first reference line segment and the target point is 20m, and the distance between the positioning point on the second reference line segment and the target point is 30m, then the corresponding line distance between the first reference line segment and the supplementary line segment is 20m, and the line distance between the second reference line segment and the supplementary line segment is 30m.

[0150] S255, based on the ratio of the reference distance to the line distance, obtain the influence coefficient corresponding to the first reference line segment or the second reference line segment.

[0151] The weighted sum is obtained by multiplying the corresponding influence coefficients and the reference data. The supplementary data of the supplementary line segment is determined based on the ratio of the weighted sum to the sum of the coefficients corresponding to each influence coefficient.

[0152] Specifically, the above-mentioned supplementary data can be obtained using the following formula:

[0153]

[0154] Where B represents the supplementary data, k1 represents the influence coefficient of the first or second reference line segment, C1 represents the reference data of the first or second reference line segment, k2 represents the influence coefficient of the second or second reference line segment, C2 represents the reference data of the second or second reference line segment, and k n C is the influence coefficient of the nth first reference line segment or second reference line segment. n This is the reference data for the nth first or second reference line segment, where d0 is the base distance, and d n This represents the line distance of the nth first or second reference line segment.

[0155] It is understandable that the baseline distance is a pre-set distance threshold, and the influence coefficient is a weighted value for calculating the degree of influence of the reference data corresponding to different line segments on the supplementary data. Since the degree of influence of the reference data corresponding to the reference line segments at different distances on the supplementary data is different, the supplementary data can be calculated by weighting the different influence coefficients corresponding to the reference line segments, so that the calculated supplementary data is more accurate.

[0156] For example, when there are three reference line segments, with distances of 5m, 10m, and 50m from the supplementary line segment respectively, and the terrain of the monitoring area is mountainous, the greater the distance between the reference line segments and the actual environmental data of the supplementary line segment, the greater the deviation between the reference data and the actual environmental data of the supplementary line segment. Consequently, the reference value of the reference data will decrease. Furthermore, as can be seen from the above formula, the greater the distance between the lines, the smaller the corresponding influence coefficient. Therefore, by multiplying the different influence coefficients with the corresponding reference data and averaging them, the obtained supplementary data becomes more accurate.

[0157] S3, based on the icing prediction model, perform icing analysis on the environmental data to obtain predicted icing data, and obtain anomaly prediction data based on the actual icing data collected by the inspection terminal and the predicted icing data.

[0158] Understandably, once the environmental data for each sub-line segment is obtained, it is possible to predict whether the corresponding sub-line segment will experience icing. For example, if the environmental data of the line's location meets the standard conditions for icing, it can be predicted that the corresponding line is already icy. For instance, the standard conditions for icing are a line surface temperature below 0°C, an ambient humidity above 85%, and a wind speed greater than 1 m / s. Therefore, if the line's environmental data meets the conditions of a temperature below 0°C, an ambient humidity above 85%, and a wind speed greater than 1 m / s, it can be predicted that the line may experience icing. This allows the drone to perform de-icing operations in a timely manner, thereby ensuring the normal operation of the line.

[0159] The icing prediction model is a model for predicting icing on the line. The predicted icing data is the prediction result of whether a sub-line segment is iced or not. The inspection terminal is the equipment for inspection and detection, which can be a drone. The actual icing data is the result of whether the sub-line segment is actually iced or not. For example, the actual shooting by the inspection terminal shows that the sub-line segment is actually iced, or the actual shooting by the inspection terminal shows that the sub-line segment is actually not iced. Abnormal prediction data is data where the actual icing data is inconsistent with the predicted icing data. That is, when the predicted result is inconsistent with the actual detected result, it can be said that the predicted result data is abnormal.

[0160] Based on the above embodiments, the specific implementation of step S3 (obtaining predicted icing data by performing icing analysis on the environmental data based on the icing prediction model, and obtaining anomaly prediction data based on the actual icing data and predicted icing data collected by the inspection terminal) can be as follows:

[0161] S31, Based on the icing prediction model, obtain the standard icing intervals corresponding to each of the data attributes, and compare the monitoring values ​​corresponding to the environmental data with the same data attributes with the standard icing intervals. The standard prediction data includes the standard icing intervals.

[0162] It is understandable that the standard icing range is the standard data range within which the line can be iced. For example, the standard icing range for temperature is (0°, -5°), the standard icing range for humidity is (85%, 100%), and the standard icing range for wind speed is (1m / s, 15m / s). The monitored values ​​are the values ​​monitored for the corresponding data attributes in the environmental data.

[0163] It is easy to understand that by comparing the monitored values ​​with the standard icing range with the same data attributes, if the monitored values ​​are within the standard icing range, it can be said that the line section has the conditions for icing.

[0164] S32, determine the sub-line segments where all the monitored values ​​are within the standard icing range as predicted icing line segments, and highlight the predicted icing line segments in the monitoring map to obtain predicted icing data.

[0165] It is understandable that when the monitored values ​​of all attributes are within the corresponding standard icing range, it can be predicted that the corresponding sub-segment may ic. Therefore, the corresponding sub-segment can be regarded as the predicted icing segment, and the predicted icing segment can be highlighted in the monitoring map for monitoring personnel to view, so as to call up the drone to carry out de-icing operation on the predicted icing segment.

[0166] Among them, the predicted icing data is the image data of the monitoring map after highlighting the predicted icing line sections.

[0167] S33, acquire the actual inspection images collected by the inspection terminal for each of the sub-line segments, and determine the actual icing line segments with icing based on the actual inspection images.

[0168] Understandably, the inspection terminal can collect actual inspection images of each sub-line segment, thus allowing a direct view of whether the corresponding sub-line segment is icy under actual conditions. When the corresponding sub-line segment is identified as icy in the actual inspection image, the actual icy line segment can be identified.

[0169] It is easy to understand that once the actual icy section of the road is identified, drones can be operated to perform de-icing operations.

[0170] S34, obtain actual icing data based on the actual icing line segment. If the sub-line segment corresponding to the actual icing line segment and the predicted icing line segment are not the same, determine the corresponding sub-line segment as an abnormal line segment, and use the abnormal line segment as abnormal prediction data.

[0171] It is understandable that actual icing data refers to information on the actual icing line segments, abnormal line segments refer to predicted icing line segments where the sub-segments corresponding to the actual icing line segments and the predicted icing line segments are different, and abnormal prediction data refers to data that displays abnormal line segments.

[0172] Through the above implementation methods, the present invention can obtain corresponding abnormal prediction data, so that the icing prediction model can be adjusted based on the abnormal prediction data after subsequent judgment, making the prediction results more accurate.

[0173] S4. Based on the abnormal prediction data, the standard prediction data of the icing prediction model is adjusted to obtain a training prediction model.

[0174] It is easy to understand that, in order to make the prediction results more accurate, the icing prediction model can be continuously trained using data actually collected by drones to improve the accuracy of the prediction results. This allows for a faster and more accurate determination of whether the corresponding line is icy and the implementation of appropriate measures.

[0175] It is understandable that when abnormal prediction data appears, it indicates that there is a certain difference between the predicted icing results and the actual icing results. Therefore, it is necessary to train the icing prediction model with data to improve the accuracy of the prediction results.

[0176] The training prediction model is a model that has been adjusted by feedback from the icing prediction model. Using this model to predict icing on the railway line can make the prediction results more accurate.

[0177] Based on the above embodiments, step S4 (adjusting the standard prediction data of the icing prediction model according to the abnormal prediction data to obtain the trained prediction model) can be implemented as follows:

[0178] S41, Obtain the geographic information corresponding to the center location point of the abnormal line segment, wherein the geographic information includes at least the topographic features.

[0179] It is understandable that geographic information refers to the geographical location of the route, including topographic features, such as mountains.

[0180] It is easy to understand that icing conditions may vary depending on the terrain. For example, the temperature, humidity, and wind speed for icing on a mountainous area may be different from those for icing on a plain. Therefore, the icing prediction model can be adjusted based on geographical information.

[0181] S42, based on the landform features, the standard icing intervals corresponding to each of the data attributes are adjusted to obtain the adjusted icing intervals, wherein the landform features include at least mountains.

[0182] It is understandable that by adjusting the standard icing intervals of each data attribute through different landform features, accurate adjusted icing intervals corresponding to different landforms can be obtained. This allows for rapid prediction of the icing situation of the line at the corresponding location based on the landform features, and makes the predicted icing results more accurate.

[0183] Among them, the adjusted icing range is the icing data range adjusted according to different landform characteristics.

[0184] It is worth mentioning that when the geographical attributes are the same, the trained model can be directly used to predict icing on railway lines in areas with the same geographical information, thereby improving the efficiency of icing prediction.

[0185] Based on the above embodiments, step S42 (adjusting the standard icing intervals corresponding to each data attribute according to the landform features to obtain the adjusted icing intervals, wherein the landform features at least include mountains) can be implemented as follows:

[0186] S421, determine that the sub-segment corresponding to the actual icing line segment in the abnormal prediction data is a false negative line segment, and the sub-segment corresponding to the predicted icing line segment is a false positive line segment.

[0187] It is understandable that a false negative line segment is a sub-line segment that was predicted not to be icy, but was actually found to be icy, while a false positive line segment is a sub-line segment that was predicted to be icy, but was actually not found to be icy.

[0188] Through the above-described embodiments, the present invention can identify false negative and false positive line segments, so as to facilitate subsequent adjustments based on the data corresponding to different types of line segments.

[0189] S422, obtain the icing trend corresponding to each of the data attributes of the false negative line segment, and determine the standard icing interval corresponding to the data attribute with the icing trend being a downward trend as the adjustment interval.

[0190] It is understandable that, regarding humidity in icing conditions, the higher the humidity, the easier it is for the line to become icy. Therefore, when a sub-line segment is a false negative segment, that is, the predicted line is not icy, but icing is actually detected on the line segment, it can be said that the humidity icing prediction range corresponding to this geographical location is high. Therefore, the icing prediction range corresponding to the predicted humidity attribute can be adjusted downward to make the icing prediction corresponding to this terrain more accurate and consistent with the actual icing situation.

[0191] The icing trend refers to the adjustment trend of the corresponding data attributes in the route. This includes both downward and upward trends. For example, when the humidity of the environmental data of the geographical location of the route is 80%, the standard icing interval should be above 85% to predict that the corresponding sub-route is ic-free. However, when the drone is used for detection, it is determined that the sub-route is already icy. Therefore, the standard icing interval (85%, 100%) corresponding to the humidity can be adjusted downward to include the humidity data value of the actual icing. The downward trend is the downward decreasing trend, and the downward adjustment interval is the standard icing interval of the data that needs to be adjusted downward.

[0192] It is easy to understand that, for wind speed in icing conditions, the higher the wind speed, the more likely the line is to become icy. Therefore, when a sub-line segment is a false negative segment, that is, the predicted line is not icy, but icing is actually detected on the line segment, it can be said that the wind speed icing prediction range corresponding to this geographical location is high. Therefore, the icing prediction range corresponding to the predicted wind speed attribute can be adjusted downward to make the icing prediction corresponding to this terrain more accurate and consistent with the actual icing situation.

[0193] S423, determine the standard icing interval corresponding to the data attribute whose icing trend is upward as the upward adjustment interval.

[0194] It is understandable that, for the temperature in the icing conditions, the lower the temperature, the easier it is for the line to become icy. Therefore, when a sub-line segment is a false negative segment, that is, the predicted line is not icy, but the line segment is actually detected to be icy, it can be said that the temperature value corresponding to the geographical location for judging icing is high. Therefore, the icing prediction range corresponding to the predicted temperature attribute can be adjusted upward, so that the icing prediction corresponding to the terrain is more accurate and consistent with the actual icing situation.

[0195] Among them, the upward trend refers to the upward increase, and the upward adjustment range refers to the standard icing range corresponding to the data attribute that needs to be adjusted upward.

[0196] S424, obtain the non-icing trend corresponding to each of the data attributes of the false positive line segment, and determine the standard icing interval corresponding to the data attribute whose non-icing trend is a downward trend as the adjustment interval.

[0197] It is understandable that false positive line segments are predicted to be icy, but the actual lines are not icy. Therefore, the trend corresponding to each data attribute in the false positive line segments is the no-icing trend, which includes both downward and upward trends.

[0198] It is easy to understand that, regarding temperature in icing conditions, the lower the temperature, the easier it is for the line to become icy. Therefore, when a sub-line segment is a false positive segment—that is, the predicted line is icy, but the actual detected line segment is not icy—it indicates that the temperature value corresponding to the geographical location for judging icing is relatively high. To match the actual situation of no icing, the corresponding no-icing trend is a decreasing trend. Therefore, the icing prediction interval corresponding to the predicted temperature attribute can be adjusted downwards to make the icing prediction corresponding to this terrain more accurate and consistent with the actual icing situation. S425, the standard icing interval corresponding to the data attribute whose no-icing trend is an increasing trend is determined as the upward adjustment interval.

[0199] Understandably, when the humidity of the environmental data at the location of the line is 85%, the standard icing interval requires a humidity of 85% or higher to predict that the corresponding sub-line segment will be icy. However, when the drone is used for detection, it is determined that there is no icing on the sub-line segment. Therefore, the standard icing interval (85%, 100%) corresponding to the humidity can be adjusted upward. That is, the data interval for judging the line icing at the corresponding geographical location can be increased so that the actual humidity data value of icing matches the corresponding icing situation, making the predicted icing result more accurate.

[0200] It is worth mentioning that when the humidity is 100%, when the standard icing interval (85%, 100%) is adjusted upwards, the maximum value of the corresponding interval can still be 100%, for example, the interval after upward adjustment is (86%, 100%). S426, obtain the historical monitoring data corresponding to the landform feature, and adjust the upward or downward interval according to the historical monitoring data to obtain the adjusted icing interval corresponding to the data attribute.

[0201] It is understandable that historical monitoring data refers to data monitored for the corresponding landform features during historical periods, such as environmental data monitored during icing of sub-road sections in the landform over the past 3 years.

[0202] Through the above implementation methods, the present invention can obtain the adjusted icing intervals corresponding to each data attribute, so as to improve the accuracy of the prediction results when making predictions by adjusting the icing intervals.

[0203] Based on the above embodiments, step S426 (obtaining historical monitoring data corresponding to the landform features, adjusting the upward or downward adjustment interval according to the historical monitoring data, and obtaining the adjusted icing interval corresponding to the data attribute) can be implemented as follows:

[0204] S4261, when the data attribute corresponds to the upward adjustment interval, obtain the historical ice-free value corresponding to the data attribute under the landform feature, and determine the first mean of each historical ice-free value.

[0205] Understandably, the historical no-icing value refers to the data value of the sub-line segment that did not experience icing within a historical period. For example, if the humidity was 85%, 87%, and 86% during the historical period, the line did not experience icing, thus obtaining historical no-icing values ​​with humidity of 85%, 87%, and 86%. The first mean is the average of the historical no-icing values. For example, if the historical no-icing values ​​have humidity of 85%, 87%, and 86%, then the first mean can be obtained as (85% + 87% + 86%) / 3 = 86%.

[0206] Through the above implementation methods, the present invention can obtain the corresponding first mean value, so as to obtain the corresponding adjustment coefficient.

[0207] S4262, obtain the first difference between the monitoring value corresponding to the data attribute and the first mean, and obtain the adjustment coefficient based on the ratio of the first difference to the benchmark difference.

[0208] Understandably, the first difference is the difference between the monitored value corresponding to the data attribute of the upward adjustment interval and the first mean, the benchmark difference is the allowable benchmark deviation between the monitored value and the first mean, and the upward adjustment coefficient is the offset value used to adjust the value of the standard interval upward.

[0209] It is easy to understand that, in order to improve the accuracy of icing prediction results, the larger the difference between the monitored value and the first average value under historical icing-free conditions, the larger the corresponding upward adjustment coefficient can be. This allows the subsequent adjustment range to be larger, making the adjusted range closer to the judgment range under icing-free conditions in the corresponding area.

[0210] Through the above implementation methods, the present invention can obtain the upward adjustment coefficient of the corresponding attribute data, so as to subsequently determine the corresponding adjustment icing range.

[0211] S4263, obtain the first interval extreme value corresponding to the upward adjustment interval, obtain the upward adjustment extreme value based on the sum of the upward adjustment coefficient and the first interval extreme value, and obtain the adjustment icing interval corresponding to the data attribute based on the upward adjustment extreme value.

[0212] It is understandable that the upward adjustment interval has a corresponding first interval extreme value, which corresponds to the maximum value and minimum value of the interval, respectively. Therefore, the upward adjustment coefficient and the value in the first interval extreme value can be added together to obtain the upward adjustment extreme value, and then the corresponding adjustment icing interval can be obtained based on the upward adjustment extreme value.

[0213] Among them, the extreme values ​​of the first interval are the maximum and minimum values ​​of the interval in the upward adjustment interval. For example, when the upward adjustment interval is the temperature corresponding to (-5°, 0°), the extreme values ​​of the first interval are -5° and 0°, and the extreme value of the upward adjustment is the sum of the upward adjustment coefficient and the extreme value of the first interval.

[0214] S4264, when the data attribute corresponds to the downward adjustment interval, obtain the historical icing value corresponding to the data attribute under the landform feature, and determine the second mean value of each historical icing value. The historical monitoring data includes historical no-icing value and historical icing value.

[0215] Understandably, the second mean is the average of the historical icing values ​​corresponding to the data attributes in the downward adjustment interval of the data attributes, and the historical icing value is the data value corresponding to the icing of the sub-line segment within the historical time period.

[0216] S4265, obtain the second difference between the monitoring value corresponding to the data attribute and the second mean, and obtain the adjustment coefficient based on the ratio of the second difference to the benchmark difference.

[0217] Understandably, the second difference is the difference between the monitored value of the data attribute corresponding to the downgrade interval and the second mean, and the downgrade coefficient is the offset value used to downgrade the standard interval.

[0218] Through the above implementation methods, the present invention can obtain the down-adjustment coefficient of the down-adjustment interval, so as to adjust the data of the down-adjustment interval and obtain the corresponding adjusted icing interval, thereby improving the accuracy of the icing prediction results.

[0219] S4266, obtain the second interval extreme value corresponding to the downward adjustment interval, obtain the downward adjustment extreme value based on the difference between the downward adjustment coefficient and the second interval extreme value, and obtain the adjustment icing interval corresponding to the data attribute based on the downward adjustment extreme value.

[0220] It is understandable that the extreme values ​​of the second interval are the maximum and minimum values ​​of the interval corresponding to the downward adjustment interval, and the extreme value of the downward adjustment is the difference between the downward adjustment coefficient and the extreme value of the second interval. Therefore, the numerical range of the icing adjustment interval can be determined based on the extreme value of the downward adjustment.

[0221] Through the above implementation methods, the present invention can obtain the adjusted icing range after the adjustment range is lowered, so as to improve the accuracy of subsequent icing prediction for sub-line segments.

[0222] S43, bind the data attribute and the landform feature, and update the standard icing range of the data attribute under the landform feature to the adjusted icing range.

[0223] It is understandable that binding data attributes with geomorphic features makes it easier to retrieve the corresponding icing adjustment range based on the geomorphic features for subsequent icing prediction, thereby improving the accuracy of icing prediction for sub-segments of that geomorphic terrain.

[0224] S44, the icing prediction model is adjusted based on the adjusted icing interval to obtain a trained prediction model.

[0225] Understandably, by continuously adjusting the icing range and providing feedback adjustments to the icing prediction model, a trained prediction model can be obtained, thereby improving the accuracy of the prediction results.

[0226] See Figure 4 This is a schematic diagram of a power transmission network risk assessment and processing system provided in an embodiment of the present invention. The power transmission network risk assessment and processing system includes:

[0227] The segmentation module is used to segment each monitoring line in the monitoring area into multiple sub-line segments based on the line configuration information of the monitoring terminal, and to obtain the monitoring data collected by the environmental monitoring equipment corresponding to each sub-line segment.

[0228] The retrieval module is used to determine the sub-line segments that meet the completion conditions as completion segments, retrieve the completion strategy to obtain the completion data of the completion segments, and determine the environmental data of the corresponding sub-line segments based on the monitoring data or completion data.

[0229] The analysis module is used to perform icing analysis on the environmental data based on the icing prediction model to obtain predicted icing data, and to obtain anomaly prediction data based on the actual icing data and predicted icing data collected by the inspection terminal.

[0230] The feedback module is used to adjust the standard prediction data of the icing prediction model based on the abnormal prediction data to obtain a trained prediction model.

[0231] The present invention also provides a readable storage medium storing a computer program, which, when executed by a processor, is used to implement the methods provided in the various embodiments described above.

[0232] The readable storage medium can be a computer storage medium or a communication medium. A communication medium includes any medium that facilitates the transfer of computer programs from one location to another. A computer storage medium can be any available medium accessible to a general-purpose or special-purpose computer. For example, a readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application-Specific Integrated Circuit (ASIC). Alternatively, the ASIC can be located in a user equipment. Of course, the processor and the readable storage medium can also exist as discrete components in a communication device. The readable storage medium can be a read-only memory (ROM), random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0233] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A risk assessment method for direct transmission networks, characterized in that, include: Based on the line configuration information of the monitoring terminal, the monitoring lines in the monitoring area are segmented to obtain multiple sub-line segments, and the monitoring data collected by the environmental monitoring equipment corresponding to each sub-line segment is obtained. Sub-segments that meet the completion criteria are identified as completion segments. Completion data for these segments is obtained by retrieving the completion strategy. Environmental data for the corresponding sub-segments is determined based on the monitoring data or completion data, including: When the monitoring data is missing in the sub-line segment, it is determined that the sub-line segment meets the completion condition. The monitoring data includes at least temperature data, humidity data, and wind speed data. The sub-line segments that meet the completion conditions are identified as the completed line segments, and the monitoring map corresponding to the monitoring area is retrieved. The monitoring line where the supplementary line segment is located in the monitoring map is identified as the target line, and a parallel line in the monitoring map that is pre-set to run parallel to the target line is determined. The target line is subjected to segment filtering to obtain a first reference line segment, and the parallel lines are subjected to parallel segment filtering to obtain a second reference line segment, including: The position points located at both ends of the supplemented line segment are obtained as reference points, and the position point in the target line that is at a preset reference distance from the corresponding reference point is determined as the termination point; The termination point is located in a sub-segment of the target line other than the completed line segment; The sub-line segment from the corresponding reference point to the termination point is determined as the first reference line segment; Connect the two reference points to obtain the range delineation line, and determine the midpoint of the range delineation line as the reference point; Generate a baseline perpendicular to the defined range line and passing through the reference point, and obtain the line segment distance between the midpoint and any of the reference points; Two points on the baseline that are at a distance from the baseline point are respectively identified as connection points. The connection points and the two reference points are connected to obtain the parallel filtering range. The sub-line segment within the parallel filtering range of the parallel lines is identified as the second reference line segment; The supplementary data of the supplementary line segment is determined based on the monitoring data of the first reference line segment and the second reference line segment, and the monitoring data or supplementary data is determined to be the environmental data of the corresponding sub-line segment. Based on the icing prediction model, the environmental data is analyzed to obtain predicted icing data. Based on the actual icing data and predicted icing data collected by the inspection terminal, anomaly prediction data is obtained. Based on the abnormal prediction data, the standard prediction data of the icing prediction model is adjusted by feedback to obtain a trained prediction model.

2. The method according to claim 1, characterized in that, After determining that the sub-line segment encompassed from the corresponding reference point to the termination point is the first reference line segment, the method further includes: The center position point of each first reference line segment is obtained as the first positioning point, and the first reference line segment whose first positioning point is not in the line area corresponding to the reference point to the termination point is determined as the first line segment to be screened. Determine the phase difference between the first positioning point and the termination point of the first line segment to be screened, and delete the first line segment to be screened whose phase difference is greater than the reference phase difference. After determining that the sub-line segment within the parallel filtering range of the parallel lines is the second reference line segment, the method further includes: The center point of each second reference line segment is obtained as the second positioning point, and the second reference line segment whose second positioning point is located outside the parallel screening range is determined as the second line segment to be screened. Determine the shortest distance between the second positioning point of the second screening line segment and the parallel screening range, and delete the second screening line segment whose shortest distance is greater than the reference distance.

3. The method according to claim 1, characterized in that, The completion data for the completion segment is determined based on the monitoring data of the first reference line segment and the second reference line segment, including: The missing data attributes of the completed line segment are obtained as completion attributes, and the data attributes include at least temperature attributes, humidity attributes, and wind speed attributes; The monitoring data corresponding to the first reference line segment and the second reference line segment and the completion attribute are determined as reference data; The center point of the completed line segment is obtained as the target point, and the center points of the first reference line segment and the second reference line segment are obtained as the positioning points; Based on the point distance between the positioning point and the target point, the line distance between each of the first reference line segment and the second reference line segment and the supplementary line segment is obtained; Based on the ratio of the reference distance to the line distance, the influence coefficient corresponding to the first reference line segment or the second reference line segment is obtained; The weighted sum is obtained by multiplying the corresponding influence coefficients and the reference data. The supplementary data of the supplementary line segment is determined based on the ratio of the weighted sum to the sum of the coefficients corresponding to each influence coefficient.

4. The method according to claim 3, characterized in that, Based on the icing prediction model, icing analysis is performed on the environmental data to obtain predicted icing data. Anomaly prediction data is obtained based on the actual icing data collected by the inspection terminal and the predicted icing data, including: Based on the icing prediction model, standard icing intervals corresponding to each of the data attributes are obtained. The monitoring values ​​corresponding to environmental data with the same data attributes are compared with the standard icing intervals. The standard prediction data includes the standard icing intervals. The sub-line segments whose monitored values ​​are all within the standard icing range are identified as predicted icing line segments, and the predicted icing line segments in the monitoring map are highlighted to obtain predicted icing data. The actual inspection images collected by the inspection terminal for each of the sub-line segments are obtained, and the actual ice-covered line segments with ice are determined based on the actual inspection images. Based on the actual icing line segment, actual icing data is obtained. If the sub-segment corresponding to the actual icing line segment and the predicted icing line segment are not the same, the corresponding sub-segment is determined to be an abnormal line segment, and the abnormal line segment is used as abnormal prediction data.

5. The method according to claim 4, characterized in that, Based on the abnormal prediction data, the standard prediction data of the icing prediction model is adjusted using feedback to obtain a trained prediction model, including: Obtain the geographic information corresponding to the center location point of the abnormal line segment, and the geographic information includes at least the topographic features; Based on the aforementioned landform features, the standard icing intervals corresponding to each of the aforementioned data attributes are adjusted to obtain the adjusted icing intervals. The aforementioned landform features include at least mountains. The data attributes and the landform features are bound together, and the standard icing range of the data attributes under the landform features is updated to the adjusted icing range. The icing prediction model is adjusted based on the adjusted icing interval to obtain a trained prediction model.

6. The method according to claim 5, characterized in that, Based on the aforementioned geomorphic features, the standard icing intervals corresponding to each of the aforementioned data attributes are adjusted to obtain adjusted icing intervals. The geomorphic features at least include mountains, including: The sub-segment corresponding to the actual icing line segment in the abnormal prediction data is determined to be a false negative line segment, and the sub-segment corresponding to the predicted icing line segment is determined to be a false positive line segment. Obtain the icing trend corresponding to each of the data attributes of the false negative line segment, and determine the standard icing interval corresponding to the data attribute with the icing trend being a downward trend as the adjustment interval; The standard icing interval corresponding to the data attribute indicating an upward icing trend is determined as the upward adjustment interval; or, Obtain the non-icing trend corresponding to each of the data attributes of the false positive line segment, and determine the standard icing interval corresponding to the data attribute whose non-icing trend is a downward trend as the adjustment interval; The standard icing interval corresponding to the data attribute where the non-icing trend is upward is determined as the upward adjustment interval; Obtain historical monitoring data corresponding to the landform features, and adjust the upward or downward adjustment range based on the historical monitoring data to obtain the adjusted icing range corresponding to the data attributes.

7. The method according to claim 6, characterized in that, Obtain historical monitoring data corresponding to the landform features, and adjust the upward or downward adjustment interval based on the historical monitoring data to obtain the adjusted icing interval corresponding to the data attribute, including: When the data attribute corresponds to the upward adjustment interval, obtain the historical ice-free value corresponding to the data attribute under the landform feature, and determine the first mean of each historical ice-free value; Obtain the first difference between the monitored value corresponding to the data attribute and the first mean, and obtain the upward adjustment coefficient based on the ratio of the first difference to the benchmark difference; Obtain the first interval extreme value corresponding to the upward adjustment interval, obtain the upward adjustment extreme value based on the sum of the upward adjustment coefficient and the first interval extreme value, and obtain the adjustment icing interval corresponding to the data attribute based on the upward adjustment extreme value; When the data attribute corresponds to the downward adjustment interval, the historical icing value corresponding to the data attribute under the landform feature is obtained, and the second mean of each historical icing value is determined. The historical monitoring data includes historical no-icing value and historical icing value. Obtain the second difference between the monitoring value corresponding to the data attribute and the second mean, and obtain the down adjustment coefficient based on the ratio of the second difference to the benchmark difference; Obtain the second interval extreme value corresponding to the downward adjustment interval, obtain the downward adjustment extreme value based on the difference between the downward adjustment coefficient and the second interval extreme value, and obtain the adjustment icing interval corresponding to the data attribute based on the downward adjustment extreme value.

8. A readable storage medium, characterized in that, The readable storage medium stores a computer program, which, when executed by a processor, is used to implement the method described in any one of claims 1 to 7.

Images