Search

CN-121981527-A - Remote sensing image interpretation and power service-based transmission line alarm prediction method

CN121981527ACN 121981527 ACN121981527 ACN 121981527ACN-121981527-A

Abstract

The invention discloses a transmission line warning prediction method based on remote sensing image interpretation and power service, which comprises the steps of firstly carrying out non-supervision classification on various ground objects in a remote sensing image by combining sample information through a remote sensing image interpretation algorithm, extracting edge information and standardizing the edge information into JSON files; and finally, judging the threat level of the ground object and generating early warning pushing by combining scene perception data and a distance early warning guide rule. The remote sensing image classification adopts a K-means algorithm optimized by an elbow algorithm, improves the distance measurement and calculation precision through accurate coordinate calculation and drop point matching, and supports integration with a monitoring platform to realize multi-scene thematic early warning. The invention breaks through the limitation of traditional manual inspection, realizes large-scale, dynamic and non-contact intelligent monitoring of the environment of the transmission line and the channel, and remarkably improves threat identification accuracy and emergency response efficiency.

Inventors

  • YU CHUNJIAN
  • LIN GUOFANG
  • CHEN LIANG
  • LI WEIFENG
  • ZHU YANG
  • CHEN JINXIANG

Assignees

  • 福建和盛高科技产业有限公司

Dates

Publication Date
20260505
Application Date
20251230

Claims (10)

  1. 1. The power transmission line warning prediction method based on remote sensing image interpretation and power service is characterized by comprising the following steps of: step 1, a remote sensing image interpretation algorithm is adopted to interpret the remote sensing image, various ground objects in the remote sensing image are classified by combining the acquired sample information, the edge information of the various ground objects is extracted and counted, and the edge information and the attribute information of the corresponding ground objects are standardized into a classification interpretation result JSON file; step 2, based on a drop point statistical algorithm, integrating the line pole tower coordinate information in the power transmission business ledger and the ground object classification statistical information in the JSON file, and calculating the actual distance between the classified ground object area and the power transmission pole tower; And 3, combining data acquired by the scene sensing equipment with a preset transmission line distance early warning guide rule, judging threat degrees of the specific category of ground features according to the actual distance between the resolved classified ground feature areas and the transmission towers, generating early warning and warning information and pushing the early warning and warning information.
  2. 2. The method for predicting the power transmission line alarm based on remote sensing image interpretation and power business according to claim 1, wherein the step 1 comprises the following steps: step 1-1, collecting labeling sample data on a remote sensing image to generate a sample image according to the interpretation and classification requirements of ground objects, and collecting attribute information of the sample image; step 1-2, performing feature classification on the remote sensing image by adopting an unsupervised classification algorithm to generate an interpretation result image containing feature class information; and step 1-3, extracting edge information of an interpretation result by using an edge extraction algorithm, acquiring coordinate data of all classified features based on the edge information, and generating a standard JSON file by combining attribute information of a sample image.
  3. 3. The method for predicting the warning of the power transmission line based on remote sensing image interpretation and power service according to claim 2, wherein the unsupervised classification algorithm in the step 1-2 is a K-means algorithm, the optimal type number K of the remote sensing image ground object clusters is determined through an elbow algorithm, and unsupervised classification of the remote sensing image is completed by combining the optimal type number K with sample information through the K-means algorithm, so that a classification result A is generated.
  4. 4. The power transmission line warning prediction method based on remote sensing image interpretation and power service according to claim 2, wherein the generation process of the JSON file in step 1-3 comprises the following steps: step 1-3-1, obtaining attribute information of an original remote sensing image, wherein the attribute information comprises longitude and latitude ranges (lonMax, latMax, lonMin, latMin) and row and column numbers (X, Y); Step 1-3-2, calculating image step information, wherein, longitude step positionRateLon = (lonMax-lonMin)/Y, latitude step positionRateLat = (latMax-latMin)/X; Step 1-3-3, extracting an edge coordinate p (tx, ty) of each classification area in the ground object classification result, and calculating longitude and latitude (Plon, plat) corresponding to the edge coordinate by a formula Plon = lonMin + positionRateLon x ty and plat= latMin + positionRateLat x to form an edge coordinate set B of the classification result; And step 1-3-4, carrying out standardized processing on the boundary index set and the corresponding ground object attribute information by using a standard JSON library to generate a JSON file.
  5. 5. The method for predicting the power transmission line alarm based on remote sensing image interpretation and power business according to claim 1, wherein the step 2 comprises the following steps: step 2-1, acquiring the standing book information of the power transmission line, and extracting longitude and latitude coordinates (lon, lat) of each tower of the line; step 2-2, back calculating coordinates (x, y) of a falling point of the tower on the remote sensing image by using the image stepping information; step 2-3, acquiring a body unit edge coordinate set corresponding to each classified ground object from the JSON file; step 2-4, judging the target body unit where the tower is or the periphery corresponds to remote sensing interpretation according to the falling point coordinates (x, y) of the tower; And 2-5, calculating the actual distance between the central point coordinate of the target body unit and the tower coordinate.
  6. 6. The method for predicting power transmission line alarm based on remote sensing image interpretation and power service as claimed in claim 5, wherein the solution formula of the coordinates (x, y) of the falling point in step 2-2 is x= (lat-latMin)/positionRateLat, y= (lon-lonMin)/positionRateLon, wherein latMin, lonMin is the minimum latitude and the minimum longitude of the original remote sensing image, and positionRateLat, positionRateLon is the latitude step and the longitude step, respectively.
  7. 7. The method for predicting the power transmission line alarm based on remote sensing image interpretation and power service according to claim 5, wherein the steps 2-4 comprise the following steps: step 2-4-1, converting a body unit edge longitude and latitude coordinate set in the JSON file into a body unit edge image coordinate set M1 through a falling point coordinate calculation formula; Step 2-4-2, performing range matching on the tower falling point coordinates (x, y) and the body unit edge image coordinate set M1, screening out a subset M2 closest to the tower falling point coordinates in the body unit edge image coordinate set M1, obtaining center point coordinates M (mlon, mlat) of the subset M2, and solving image coordinates Mimage (ix, iy) corresponding to the center point M on an image according to a falling point coordinate (x, y) solving formula, wherein ix= (mlat-latMin)/positionRateLat, iy= (mlon-lonMin)/positionRateLon, and latMin, lonMin is the minimum latitude and the minimum longitude of an original remote sensing image respectively, and positionRateLat, positionRateLon is the latitude step and the longitude respectively; step 2-4-3, resolving the image distance d between the coordinates (x, y) of the tower landing point and the central coordinates Mimage (ix, iy) of the subset M2, And calculating the actual distance dis between the coordinates of the central point of the body unit corresponding to the subset M2 and the coordinates of the falling point of the tower by adopting a distance dissociation calculation formula dis=d.s, wherein s represents the ground resolution represented by each pixel of the image.
  8. 8. The method for predicting the warning of the power transmission line based on remote sensing image interpretation and power service according to claim 1, wherein in step 3, the warning level is determined according to the actual distance between the power transmission line tower and each ground object category through a distance warning guide rule according to the calculated power transmission line tower and remote sensing interpretation, the warning level in the distance warning guide comprises high threat, medium threat and low threat, and the distance threshold corresponding to each level is set according to the safety operation specification of the power transmission line.
  9. 9. The method for predicting the power transmission line alarm based on remote sensing image interpretation and power service according to claim 8, wherein the method is characterized in that the method is judged to be a high threat when the predicted actual distance is less than or equal to 15 meters, is judged to be a medium threat when the predicted actual distance is greater than 15 meters and less than or equal to 20 meters, and is judged to be a low threat when the predicted actual distance is greater than 20 meters.
  10. 10. The method for predicting the power transmission line alarm based on remote sensing image interpretation and power service according to claim 1, wherein step 3 further comprises the steps of combining a power transmission line type threat prediction model with a type distance early warning diagnosis model, constructing a line type threat application scene, and integrating the application scene into a power transmission equipment state monitoring platform to realize integrated application.

Description

Remote sensing image interpretation and power service-based transmission line alarm prediction method Technical Field The invention relates to the field of intelligent monitoring of transmission lines, in particular to a transmission line alarm prediction method based on remote sensing image interpretation and power business. Background The power industry is the basic pulse of national economy, and the safe, stable and efficient operation of the power industry is of great importance. Traditional power facility inspection, planning and management modes mainly depend on manual field operation, and have the inherent defects of low efficiency, high cost, large risk, limited coverage and the like. The remote sensing technology, in particular satellite remote sensing and aerial remote sensing (including man-machine and unmanned aerial vehicle), has the capability of acquiring surface information in a macroscopic, rapid, objective and regular way. The method combines the two, can realize large-scale, dynamic and non-contact monitoring of wide-area distributed electric power facilities and surrounding environments thereof, and is a key technical support for converting the electric power industry into digital and intelligent conversion. Disclosure of Invention The invention aims to provide a power transmission line warning prediction method based on remote sensing image interpretation and power service, which is used for solving the threat degree of a power transmission line body and the surrounding environment of a line channel to a pole tower by combining remote sensing images with power transmission service information so as to realize hidden danger threat and early warning of the power transmission line and improve emergency treatment efficiency. The technical scheme adopted by the invention is as follows: The transmission line warning prediction method based on remote sensing image interpretation and power service comprises the following steps: step 1, a remote sensing image interpretation algorithm is adopted to interpret the remote sensing image, various ground objects in the remote sensing image are classified by combining the acquired sample information, the edge information of the various ground objects is extracted and counted, and the edge information and the attribute information of the corresponding ground objects are standardized into a classification interpretation result JSON file; step 2, based on a drop point statistical algorithm, integrating the line pole tower coordinate information in the power transmission business ledger and the ground object classification statistical information in the JSON file, and calculating the actual distance between the classified ground object area and the power transmission pole tower; And 3, combining data acquired by the scene sensing equipment with a preset transmission line distance early warning guide rule, judging threat degrees of the specific category of ground features according to the actual distance between the resolved classified ground feature areas and the transmission towers, generating early warning and warning information and pushing the early warning and warning information. Further, step 1 includes the steps of: step 1-1, collecting labeling sample data on a remote sensing image to generate a sample image according to the interpretation and classification requirements of ground objects, and collecting attribute information of the sample image; step 1-2, performing feature classification on the remote sensing image by adopting an unsupervised classification algorithm to generate an interpretation result image containing feature class information; and step 1-3, extracting edge information of an interpretation result by using an edge extraction algorithm, acquiring coordinate data of all classified features based on the edge information, and generating a standard JSON file by combining attribute information of a sample image. Further, the unsupervised classification algorithm in the step 1-2 is a K-means algorithm, the optimal type number K of the remote sensing image ground feature clusters is determined through an elbow algorithm, the unsupervised classification of the remote sensing images is completed by combining the K-means algorithm with the optimal type number K and sample information, and a classification result A is generated. Further, the generation process of the JSON file in the step 1-3 comprises the following steps: step 1-3-1, obtaining attribute information of an original remote sensing image, wherein the attribute information comprises longitude and latitude ranges (lonMax, latMax, lonMin, latMin) and row and column numbers (X, Y); Step 1-3-2, calculating image step information, wherein, longitude step positionRateLon = (lonMax-lonMin)/Y, latitude step positionRateLat = (latMax-latMin)/X; Step 1-3-3, extracting an edge coordinate p (tx, ty) of each classification area in the ground object classification result, and calculating lon