Ground equipment oil leakage hidden danger identification method and system

Abstract

The invention discloses a ground equipment oil leakage hidden danger identification method and system, and the method comprises the steps: collecting the ground surface reflected light brightness data of a ground region of target equipment through a satellite multispectral channel, recognizing a suspected oil leakage region according to the ground surface reflected light brightness data, and extracting the position coordinates and boundary range of the suspected oil leakage region; the unmanned aerial vehicle detects the organic matter concentration of the suspected oil leakage area one by one according to the inspection flight route, determines concentration distribution data, identifies an abnormal concentration standard-exceeding area according to the concentration distribution data, and generates an abnormal concentration distribution area map corresponding to the abnormal concentration standard-exceeding area; and comparing and matching the suspected oil leakage region with the concentration anomaly region map, determining a comprehensive matching degree score about the position overlapping degree and the boundary consistency, comprehensively evaluating the suspected oil leakage region and the concentration anomaly exceeding region, and determining an actual oil leakage region, thereby improving the recognition efficiency and reliability of the oil leakage hidden danger.

Description

Technical Field

[0001] This invention relates to the field of substation equipment monitoring technology, and in particular to a method and system for identifying potential oil leaks in ground equipment. Background Technology

[0002] Oil leaks are a common problem in substation equipment due to prolonged operation, environmental factors, and equipment aging. If these leaks are not detected and addressed in time, they can not only waste resources but also cause safety accidents, posing a serious threat to the normal operation of the substation.

[0003] Traditional methods for identifying potential oil leaks mainly rely on manual inspections. This approach is not only inefficient, but also makes it difficult to comprehensively and promptly detect oil leak areas, easily overlooking potential oil leaks. Summary of the Invention

[0004] In view of this, in order to solve the above-mentioned technical problems, the present invention provides a method and system for identifying potential oil leaks in ground equipment.

[0005] The first aspect of this invention provides a method for identifying potential oil leaks in ground equipment, comprising:

[0006] The surface reflectance data of the target equipment's ground area is collected by satellite multispectral channels, and suspected oil leak areas are identified based on the surface reflectance data. The location coordinates and boundary range of the suspected oil leak areas are then extracted.

[0007] Based on the location coordinates and boundary range of the suspected oil leak area, an inspection flight route is planned, and the concentration of organic matter in each location of the suspected oil leak area is detected by drone according to the inspection flight route.

[0008] Based on the organic matter concentration detected at each location, concentration distribution data is determined, and areas with abnormally high concentrations are identified based on the concentration distribution data. A concentration abnormality distribution area map corresponding to the areas with abnormally high concentrations is then generated.

[0009] The suspected oil leak area is compared and matched with the concentration anomaly area map to determine a comprehensive matching score regarding location overlap and boundary consistency;

[0010] If the overall matching score is higher than the preset standard score threshold, the suspected oil leak area and the area with abnormally high concentration are comprehensively evaluated to determine the actual oil leak area.

[0011] Preferably, the step of acquiring surface reflectance data of the target equipment's ground area via satellite multispectral channels, identifying suspected oil leak areas based on the surface reflectance data, and extracting the location coordinates and boundary range of the suspected oil leak areas includes:

[0012] Satellite multispectral imagery data of the ground area of ​​the target equipment is acquired through satellite multispectral channels;

[0013] Based on the satellite multispectral image data, the surface reflectance brightness data for each band is determined; wherein, the bands include the red band and the near-infrared band;

[0014] The surface reflected light intensity data of each band are normalized to obtain the normalized surface reflected light intensity data of each band.

[0015] For each pixel in the satellite multispectral image data, the reflectance ratio is determined based on the normalized surface reflectance data of the near-infrared band and the normalized surface reflectance data of the red band.

[0016] If the reflected light brightness ratio is less than a preset brightness ratio threshold, then the pixel is marked as a potential oil leakage point;

[0017] The suspected oil leak area is obtained by spatial clustering of all the marked potential oil leak points using a connected component analysis algorithm.

[0018] Based on the surface reflectance data of each pixel in the suspected oil leak area, determine the brightness gradient value between each adjacent pixel;

[0019] Based on the brightness gradient values ​​between adjacent pixels, the pixel with the largest brightness gradient value in the suspected oil leak area is determined as the boundary point. All boundary points are connected to determine the boundary range of the suspected oil leak area. The suspected oil leak area is then transformed from pixel coordinate system to geographic coordinate system to obtain the location coordinates of the suspected oil leak area.

[0020] Preferably, the step of planning an inspection flight route based on the location coordinates and boundary range of the suspected oil leak area, and using a drone to detect the organic matter concentration at each location in the suspected oil leak area according to the inspection flight route, includes:

[0021] Based on the location coordinates and boundary range of the suspected oil leak area, the suspected oil leak area is divided into grids using a rasterization method to obtain the rasterized area of ​​the suspected oil leak area.

[0022] By determining the circumscribed rectangle of the rasterized region, a rectangular grid array that at least covers the suspected oil leak area is generated based on the circumscribed rectangle; wherein, each grid in the rectangular grid array records an index number, center coordinates, and an identifier indicating whether it is located inside the suspected oil leak area; wherein, the circumscribed rectangle is the smallest rectangular range defined by the bounding box of the rasterized region;

[0023] Based on the rectangular grid array, a serpentine path planning algorithm is used to generate an inspection flight route for the UAV; wherein, the inspection flight route includes each inspection point on the route and a preset flight altitude of the UAV at each inspection point; wherein, the preset flight altitude is determined based on the vertical distance between the flight altitude of the UAV and the center coordinates of the grid corresponding to the inspection point;

[0024] The drone uses the inspection flight route to detect the concentration of organic matter at each location in the suspected oil leak area.

[0025] Preferably, the step of determining concentration distribution data based on the organic matter concentration detected at each location, identifying areas with abnormally high concentrations based on the concentration distribution data, and generating a concentration anomaly distribution area map corresponding to the areas with abnormally high concentrations includes:

[0026] Based on the organic matter concentration detected at each location, determine the organic matter concentration difference between adjacent locations;

[0027] The concentration gradient change rate is determined based on the concentration difference of the organic matter. If the concentration gradient change rate exceeds a preset concentration gradient change threshold, the position is marked as a concentration mutation point.

[0028] The concentration distribution data is determined based on the organic matter concentration and the concentration abrupt change points;

[0029] Based on the concentration distribution data, a density clustering algorithm is used to spatially aggregate the concentration exceeding the concentration threshold of the organic matter to obtain multiple cluster spaces, and the average concentration value of each cluster space is determined.

[0030] The clustering spaces are iteratively expanded to generate continuous pollution regions as areas with abnormally high concentrations.

[0031] Based on the average concentration value of each cluster space, the concentration of organic matter in the concentration abnormality area is interpolated by the inverse distance weighted interpolation method, and a concentration contour line is constructed according to the preset concentration interval through the interpolated concentration abnormality area.

[0032] Based on the concentration contour lines and the concentration anomaly exceeding the standard areas after interpolation, the concentration anomaly distribution area map is generated.

[0033] Preferably, the step of comparing and matching the suspected oil leak area with the concentration anomaly area map to determine a comprehensive matching score regarding location overlap and boundary consistency includes:

[0034] By unifying the suspected oil leak area and the concentration anomaly area map to the same geographic reference system, an intermediate suspected oil leak area map and an intermediate concentration anomaly area map are obtained.

[0035] Extract the boundary point sequence of the suspected oil leak area in the middle, and use the boundary point sequence to determine the first polygonal geometric object;

[0036] Extract the contour coordinates of the concentration anomaly area map, and use the Douglas-Puk algorithm to simplify the contour coordinates to generate a second polygonal geometric object;

[0037] A spatial superposition operation is performed on the first polygonal geometric object and the second polygonal geometric object to obtain the ratio of the area of ​​the intersection region to the area of ​​the union region between the first polygonal geometric object and the second polygonal geometric object, which is used as the positional overlap degree.

[0038] The Hausdorff distance between the boundary points of the first polygonal geometric object and the second polygonal geometric object is used as a boundary consistency index.

[0039] The positional overlap and the boundary consistency index are normalized.

[0040] Obtain the preset weights corresponding to the normalized position overlap and the normalized boundary consistency index, and perform a weighted calculation on the normalized position overlap and the normalized boundary consistency index based on the preset weights to obtain the comprehensive matching degree score.

[0041] Preferably, when the overall matching score is higher than a preset standard score threshold, a comprehensive evaluation is performed on the suspected oil leak area and the area with abnormally high concentration to determine the actual oil leak area, including:

[0042] If the overall matching score is higher than the preset standard score threshold, the spectral attenuation degree is determined based on the surface reflected light brightness data, and the excess multiple of the concentration peak is determined based on the concentration distribution data.

[0043] The degree of spectral attenuation and the multiple of exceedance of the concentration peak are normalized, and the normalized degree of spectral attenuation and the multiple of exceedance of the normalized concentration peak are weighted to obtain the oil leakage confidence value.

[0044] The oil leak confidence value is compared with a preset confidence threshold. If the oil leak confidence value exceeds the preset confidence threshold, the suspected oil leak area is confirmed as the actual oil leak area.

[0045] Preferably, the method further includes:

[0046] The spatial offset vector is determined based on the spatial difference between the centroid coordinates of the actual oil leak area and the centroid coordinates of the area with abnormally high concentration.

[0047] Obtain the actual oil leakage area at at least two sampling times, and determine the oil leakage diffusion rate based on the area change of the actual oil leakage area at adjacent sampling times and the sampling time interval.

[0048] Based on the oil spill diffusion rate and the spatial offset vector, combined with the current environmental parameter data, the spatial concentration distribution of organic matter is determined using a Gaussian diffusion model.

[0049] Based on the spatial concentration distribution, the boundaries of areas where the organic matter concentration is greater than the preset environmental concentration threshold are selected to obtain the predicted oil spill impact range and the oil spill impact area corresponding to the predicted oil spill impact range.

[0050] The concentration exceedance multiple is determined based on the ratio of the organic matter concentration to the preset environmental concentration threshold.

[0051] The oil spill diffusion rate, the oil spill area, and the concentration exceeding the standard multiple are normalized, and the normalized oil spill diffusion rate, the normalized oil spill area, and the normalized concentration exceeding the standard multiple are weighted to obtain the oil spill impact score.

[0052] The impact level of the oil leak accident is determined based on the oil leak impact score.

[0053] Preferably, the step of determining the spatial concentration distribution of organic matter using a Gaussian diffusion model based on the oil spill diffusion rate and the spatial offset vector, combined with current environmental parameter data, includes:

[0054] The coordinates of the pollution source center are determined based on the spatial offset vector and the centroid coordinates of the actual oil leak area.

[0055] The source strength is determined based on the oil leakage diffusion rate.

[0056] Extract the boundary feature points of the actual oil leakage area at adjacent sampling times, determine the displacement change rate of the two boundary feature points in the direction of the spatial offset vector, and determine the horizontal diffusion coefficient based on the displacement change rate.

[0057] Based on the pollution source center coordinates, the source strength, the horizontal diffusion coefficient, and the current environmental parameter data, the spatial concentration distribution of organic matter is determined using the Gaussian diffusion model; wherein, the current environmental parameter data includes the average wind speed.

[0058] Preferably, the method further includes:

[0059] Based on the impact level of the oil leak accident, determine the equipment maintenance urgency index;

[0060] If the equipment maintenance urgency index is greater than the preset urgency index threshold, an emergency maintenance task is generated, and the center coordinates of the actual oil leak area are extracted.

[0061] Based on the geographical distribution of equipment within the substation, the equipment closest to the center coordinates of the actual oil leak area is identified as the potential hazard equipment, and the geographical coordinates of the potential hazard equipment are extracted.

[0062] Based on the emergency maintenance task, the geographical coordinates of the potentially hazardous equipment, and the impact level of the oil leak, an equipment maintenance urgency report is generated.

[0063] Secondly, the present invention also provides a ground equipment oil leak hazard identification system, comprising:

[0064] The suspected oil leak detection module is used to collect surface reflectance data of the target equipment's ground area through satellite multispectral channels, identify suspected oil leak areas based on the surface reflectance data, and extract the location coordinates and boundary range of the suspected oil leak areas.

[0065] The area inspection module is used to plan an inspection flight route based on the location coordinates and boundary range of the suspected oil leak area, and to use a drone to detect the organic matter concentration at each location of the suspected oil leak area according to the inspection flight route.

[0066] The concentration anomaly determination module is used to determine the concentration distribution data based on the organic matter concentration detected at each location, identify the concentration anomaly exceeding the standard area based on the concentration distribution data, and generate a concentration anomaly distribution area map corresponding to the concentration anomaly exceeding the standard area.

[0067] The comprehensive scoring module is used to compare and match the suspected oil leak area with the concentration anomaly area map to determine a comprehensive matching score regarding the degree of location overlap and boundary consistency.

[0068] The actual oil leak determination module is used to comprehensively evaluate the suspected oil leak area and the area with abnormally high concentration when the comprehensive matching score is higher than the preset standard score threshold, and determine the actual oil leak area.

[0069] As can be seen from the above technical solution, this invention collects surface reflectance data of the target equipment's ground area through satellite multispectral channels, identifies suspected oil leak areas based on the surface reflectance data, extracts the location coordinates and boundary range of the suspected oil leak areas to plan inspection flight routes, and uses a drone to detect the organic matter concentration at each location of the suspected oil leak area according to the inspection flight route. Based on the detected organic matter concentration at each location, concentration distribution data is determined, and areas with abnormally high concentrations are identified based on the concentration distribution data. A concentration anomaly distribution area map corresponding to the areas with abnormally high concentrations is generated. The suspected oil leak area is compared and matched with the concentration anomaly area map to determine a comprehensive matching degree score regarding location overlap and boundary consistency. If the comprehensive matching degree score is higher than a preset standard score threshold, the suspected oil leak area and the areas with abnormally high concentrations are comprehensively evaluated to determine the actual oil leak area. Thus, by cross-validating the satellite image recognition results with the drone gas detection data, the actual oil leak area is determined based on the degree of agreement between the two detection results, achieving accurate identification of oil leak hazards and improving the efficiency and reliability of oil leak hazard identification. Attached Figure Description

[0070] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0071] Figure 1 A flowchart illustrating a method for identifying potential oil leaks in ground equipment, provided in an embodiment of the present invention;

[0072] Figure 2 This is a schematic diagram of a ground equipment oil leak hazard identification system provided in an embodiment of the present invention. Detailed Implementation

[0073] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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.

[0074] like Figure 1 As shown in the figure, this application embodiment provides a method for identifying potential oil leaks in ground equipment, including the following steps S1 to S5. Wherein:

[0075] Step S1: Collect surface reflectance data of the target equipment's ground area through satellite multispectral channels, identify suspected oil leak areas based on the surface reflectance data, and extract the location coordinates and boundary range of the suspected oil leak areas.

[0076] Among them, the satellite multispectral channel is a remote sensing technology that can capture information on reflected light from different bands of the Earth's surface. Through this channel, the reflected light intensity data of the target equipment's ground area in key bands such as the near-infrared band and the red band can be obtained.

[0077] Oil leak detection in substation equipment areas is initially identified using satellite multispectral remote sensing technology. This method fully utilizes the differences in reflectance characteristics of the leak area across different spectral bands, particularly the changes in soil spectral reflectance caused by oil contamination. When transformer oil leaks and seeps into the soil, the soil's physicochemical properties change, leading to a significant decrease in reflectance in the near-infrared band, while the change in the visible light band is relatively small. Therefore, by capturing surface reflectance data of the equipment area, suspected oil leak areas can be identified, and their location coordinates and boundary ranges can be extracted.

[0078] Step S2: Based on the location coordinates and boundary range of the suspected oil leak area, plan the inspection flight route, and use drones to detect the organic matter concentration at each location of the suspected oil leak area according to the inspection flight route.

[0079] Among these methods, planning inspection flight routes and using drones to detect the concentration of organic matter in each suspected oil leak area can accurately pinpoint the specific situation of the oil leak area.

[0080] Step S3: Based on the organic matter concentration detected at each location, determine the concentration distribution data, identify areas with abnormal concentrations exceeding the standard based on the concentration distribution data, and generate a concentration abnormality distribution area map corresponding to the areas with abnormal concentrations exceeding the standard.

[0081] Step S4: Compare and match the suspected oil leak area with the concentration anomaly area map to determine the comprehensive matching score regarding location overlap and boundary consistency.

[0082] Among them, location overlap is an indicator that measures the degree of spatial overlap between the suspected oil spill area and the area with abnormal concentration, reflecting the similarity of the two areas in geographic space. Boundary consistency is used to assess the degree of agreement between the boundaries of the two areas. By comparing the position and shape of the boundary points, it can be determined whether the two areas have similar boundary features. The comprehensive matching score is a combined score calculated by combining the location overlap and boundary consistency indicators.

[0083] Step S5: If the overall matching score is higher than the preset standard score threshold, conduct a comprehensive evaluation of the suspected oil leak area and the area with abnormally high concentration to determine the actual oil leak area.

[0084] If the overall matching score is not higher than the preset standard scoring threshold, it indicates that the suspected oil leak area may not actually have an oil leak, or the leak may be insignificant. In this case, the suspected oil leak area can be marked and recorded for subsequent inspection and verification. The preset standard scoring threshold can be reasonably set based on the actual application scenario and historical data statistics, such as 0.7 or 0.8. Only when the overall matching score exceeds the preset standard scoring threshold will the subsequent comprehensive evaluation process be carried out.

[0085] It should be noted that this embodiment of the application collects surface reflectance data of the target equipment's ground area through satellite multispectral channels, identifies suspected oil leak areas based on the surface reflectance data, extracts the location coordinates and boundary range of the suspected oil leak areas to plan an inspection flight route, and uses a drone to detect the organic matter concentration at each location of the suspected oil leak area according to the inspection flight route. Based on the detected organic matter concentration at each location, concentration distribution data is determined, and areas with abnormally high concentrations are identified based on the concentration distribution data. An abnormal concentration distribution area map corresponding to the areas with abnormally high concentrations is generated. The suspected oil leak area is compared and matched with the abnormal concentration area map to determine a comprehensive matching score regarding location overlap and boundary consistency. If the comprehensive matching score is higher than a preset standard score threshold, the suspected oil leak area and the areas with abnormally high concentrations are comprehensively evaluated to determine the actual oil leak area. Thus, by cross-validating the satellite image recognition results with the drone gas detection data, the actual oil leak area is determined based on the degree of agreement between the two detection results, achieving accurate identification of potential oil leaks and improving the efficiency and reliability of potential oil leak identification.

[0086] In some embodiments, surface reflectance data of the target equipment's ground area are acquired via satellite multispectral channels, and suspected oil spill areas are identified based on the surface reflectance data. The location coordinates and boundary range of the suspected oil spill areas are then extracted, including:

[0087] Step S101: Collect satellite multispectral image data of the ground area of ​​the target equipment through satellite multispectral channels.

[0088] Satellite multispectral imagery data contains information across multiple spectral bands, which reflect the reflection and radiation of surface objects under different spectral characteristics. For the target equipment's ground area, different materials and surface conditions will exhibit unique responses in each spectral band.

[0089] Step S102: Determine the surface reflectance data for each band based on satellite multispectral image data; the bands include the red band and the near-infrared band.

[0090] Among them, the surface reflectance data in the red band and near-infrared band were determined from satellite multispectral image data.

[0091] Step S103: Normalize the surface reflected light intensity data of each band to obtain normalized surface reflected light intensity data of each band.

[0092] The surface reflectance data was converted to a range of 0 to 1 through normalization, eliminating the influence of radiation differences under different time phases and illumination conditions. The normalization process employed a linear transformation method, dividing the brightness value of each band by the maximum possible value for that band to obtain dimensionless surface reflectance data.

[0093] Step S104: For each pixel in the satellite multispectral image data, determine the reflectance ratio based on the normalized surface reflectance data of the near-infrared band and the normalized surface reflectance data of the red band.

[0094] For satellite multispectral image data, the ratio of the normalized surface reflectance data in the near-infrared band to the normalized surface reflectance data in the red band for each pixel in the image can be determined as the reflectance ratio.

[0095] Step S105: If the ratio of reflected light brightness is less than a preset brightness ratio threshold, then mark the pixel as a potential oil leakage point.

[0096] Generally, normal soil has a high reflectance-to-brightness ratio, while oil-contaminated soil experiences a sharp decrease in near-infrared reflectance due to the oil film covering it, resulting in a reflectance-to-brightness ratio significantly lower than a preset threshold. When the ratio of a pixel falls below the threshold, that pixel is marked as a potential oil leak point.

[0097] The preset brightness ratio threshold can be between 0.5 and 0.7, and the specific value can be flexibly adjusted and optimized according to factors such as the actual detection scenario, historical data statistics, and soil type.

[0098] Step S106: Spatial clustering is performed on all marked potential oil leak points using a connected component analysis algorithm to obtain suspected oil leak areas.

[0099] Connected component analysis (CBI) is a commonly used region identification method in image processing. It aggregates adjacent potential oil leak points based on the connectivity between pixels, forming regions of a certain shape and size – these are the suspected oil leak areas. CBI can effectively eliminate isolated noise points, improving the accuracy of suspected oil leak area identification.

[0100] Specifically, the connected component analysis algorithm performs spatial clustering on all labeled outliers. Based on the eight-neighborhood connectivity criterion, this algorithm merges spatially adjacent outliers into the same connected region. The specific process includes: first, scanning the binarized outlier image line by line and assigning an initial label to each outlier; then, merging region labels with the same adjacency relationship through multiple iterations; finally forming independent connected regions, each representing a suspected oil leak contour. The algorithm also sets a minimum area threshold to filter out scattered noise points and retain suspected oil leak areas of a certain size. Step S107: Based on the surface reflectance data of each pixel in the suspected oil leak area, determine the brightness gradient value between adjacent pixels.

[0101] The brightness gradient value can be obtained by calculating the difference in surface reflected light brightness data between adjacent pixels within a suspected oil spill area. This gradient value reflects the degree of brightness change within the area; due to changes in surface characteristics, the brightness gradient of an oil-contaminated area often exhibits a specific pattern. Specifically, the brightness difference between the inner and outer pixels in the normal direction can be calculated point by point along the contour boundary to form a gradient distribution map.

[0102] Step S108: Based on the brightness gradient values ​​between adjacent pixels, determine the pixel with the largest brightness gradient value in the suspected oil leak area as the boundary point, connect all the boundary points to determine the boundary range of the suspected oil leak area, and transform the suspected oil leak area from the pixel coordinate system to the geographic coordinate system to obtain the location coordinates of the suspected oil leak area.

[0103] The boundary between the oil spill area and normal soil typically exhibits a significant abrupt change in brightness, with a large gradient value. Therefore, based on the brightness gradient values ​​between adjacent pixels, the pixel with the largest brightness gradient value can be identified as the boundary point. By connecting all boundary points, a closed boundary contour can be formed, thus determining the boundary range of the suspected oil spill area. Simultaneously, by using coordinate transformation methods to convert the suspected oil spill area from the pixel coordinate system to a geographic coordinate system (such as the WGS84 geographic coordinate system), its accurate location coordinates can be obtained, providing spatial reference for subsequent inspection and assessment.

[0104] For example, the calibration transformation maps the pixel coordinate system to the WGS84 geographic coordinate system using a pre-calibrated transformation matrix. The transformation matrix contains the spatial resolution, projection parameters, and georeferenced information of the satellite imagery, accurately converting the row and column numbers of each pixel into latitude and longitude coordinates. Through this processing, the system can accurately extract the geographic location and coverage area of ​​suspected oil spill areas, providing reliable target guidance for subsequent precise UAV detection.

[0105] In some embodiments, based on the location coordinates and boundary range of the suspected oil spill area, an inspection flight route is planned, and an unmanned aerial vehicle (UAV) is used to detect the organic matter concentration at each location in the suspected oil spill area according to the inspection flight route, including:

[0106] Step S201: Based on the location coordinates and boundary range of the suspected oil leak area, the suspected oil leak area is divided into grids using a rasterization method to obtain the rasterized area of ​​the suspected oil leak area.

[0107] The gridding method transforms irregular target areas into regular grid structures based on the geometry and size of suspected oil leak areas. The side length of the grid cells is determined by combining the horizontal resolution of the gas sensor and the positioning accuracy of the UAV, and is typically set to 5 to 10 meters.

[0108] Step S202: Determine the circumscribed rectangle of the rasterized region, and generate a rectangular grid array that at least covers the suspected oil leak area based on the circumscribed rectangle; wherein, each grid in the rectangular grid array records an index number, center coordinates, and an identifier indicating whether it is located inside the suspected oil leak area; wherein, the circumscribed rectangle is the smallest rectangular range defined by the boundary box of the rasterized region.

[0109] Specifically, the outer rectangle of the rasterized region is determined, which is the smallest rectangular area defined by the bounding box of the rasterized region. This outer rectangle, as the bounding box, can completely contain all raster cells. Based on the outer rectangle, a rectangular grid array is generated that at least covers the suspected oil leak area.

[0110] Each grid cell in the rectangular grid array records the following information: first, a globally unique index number used for subsequent path planning and data processing; second, center coordinates to accurately pinpoint the grid's location in geographic space; and third, an identifier indicating whether it is located within a suspected oil leak area, determined by whether the grid's center point falls within the original rasterized area. This identifier effectively distinguishes between effective detection areas and redundant coverage areas, optimizing UAV flight path planning.

[0111] Step S203: Based on the rectangular grid array, generate the inspection flight route of the UAV using a serpentine path planning algorithm; wherein, the inspection flight route includes each inspection point on the route and the preset flight altitude of the UAV at each inspection point; wherein, the preset flight altitude is determined based on the vertical distance between the UAV's flight altitude and the center coordinates of the grid corresponding to the inspection point.

[0112] The serpentine path planning algorithm generates a continuous flight trajectory based on the topology of the rectangular grid array. Starting from the first grid cell in the southwest corner of the array, the algorithm flies east-west to the end of the row, then moves north one grid cell, and then flies in the opposite direction to the beginning of the row, repeating this process until all grid cells are covered. This path design reduces the number of turns and the flight distance of the UAV. At each turn, the algorithm inserts a Bézier curve transition section, allowing the UAV to turn smoothly and avoiding interference from sudden stops and turns in gas detection. The path planning also considers wind direction, prioritizing headwind or crosswind flight to reduce the impact of volatile organic compound diffusion on detection accuracy. The algorithm outputs a time-series three-dimensional waypoint sequence, with each waypoint including longitude, latitude, altitude, and estimated time of arrival.

[0113] In the inspection flight route planning, the flight altitude is also optimized. In a typical example, the UAV is equipped with a gas sensor to detect the concentration of organic matter on the ground. The preset flight altitude is determined based on the vertical distance between the UAV's flight altitude and the grid center coordinates corresponding to the inspection point. At the same time, the optimal detection altitude range of the gas sensor must be taken into account.

[0114] Typically, flight altitude optimization is based on adjustments to the vertical detection characteristics of the gas sensor. The effective detection range of the sensor is usually 10 to 30 meters, and the detection accuracy decreases with increasing distance. By calculating the ratio of the current flight altitude to the sensor's vertical detection distance, when the ratio approaches 1, it indicates that the flight altitude is at the detection limit. At this point, the flight altitude is lowered to within 0.6 to 0.8 times the detection distance to ensure the reliability of the detection data.

[0115] When setting volatile organic compound (VOC) concentration detection points at the center of each grid, the spatial density of detection points is determined based on the grid size and sensor response time. For a standard grid, the main detection point is set at the center, and auxiliary detection points are set at the four vertices, forming a five-point detection pattern. When an abnormal concentration is detected, the system automatically increases the detection density within that grid, inserting more detection points for refined detection.

[0116] Step S204: Using a drone, detect the concentration of organic matter at each location in the suspected oil leak area according to the inspection flight route.

[0117] For example, the sampling frequency of the gas sensor is synchronized with the flight speed of the drone through a timestamp alignment mechanism. The gas sensor samples gas concentration at a fixed frequency, typically 1 Hz to 10 Hz.

[0118] The drone's flight speed is dynamically adjusted based on the sampling frequency and grid size to ensure that it stays at each detection location for a sufficient amount of time to complete stable sampling.

[0119] In the implementation, when the drone approaches the detection location, it decelerates to a hovering state, and the gas sensor continuously samples 3 to 5 times, taking the median as the concentration value at that location. After sampling, the system records the GPS coordinates, barometric altitude, ambient temperature, humidity, and wind speed data at the time of detection; this auxiliary information is used for subsequent concentration data correction. Spatiotemporal continuity is ensured through an interpolation algorithm; for the region between adjacent detection points, the concentration distribution is estimated based on the Kriging interpolation method.

[0120] Meanwhile, the mapping table between detection locations and geographic coordinates is stored using a hash table data structure, with the index number of the detection location as the key and a data structure containing complete spatial and temporal information as the value. Each record in the mapping table includes the WGS84 coordinates of the detection location, relative altitude, estimated detection time, actual detection time, concentration measurement value, and data quality identifier.

[0121] The mapping table is updated in real time based on the actual flight trajectory to ensure the accuracy and completeness of the data. Through the implementation of the above spatial sampling scheme, systematic detection of volatile organic compounds in suspected oil spill areas was achieved. The generated detection data has clear spatial location and time stamps, providing a reliable data foundation for subsequent concentration distribution analysis and oil spill confirmation.

[0122] Meanwhile, taking into full account the matching relationship between the detection characteristics of the gas sensor and the flight parameters of the UAV, the UAV plans its flight path based on the spatial information of the suspected oil leak area. Through precise path design and detection location deployment, it achieves full coverage detection of volatile organic compounds in the target area.

[0123] In some embodiments, based on the organic matter concentration detected at each location, concentration distribution data is determined; based on the concentration distribution data, regions with abnormally high concentrations are identified; and a concentration anomaly distribution region map corresponding to the regions with abnormally high concentrations is generated, including:

[0124] Step S301: Determine the organic matter concentration difference between adjacent locations based on the organic matter concentration detected at each location.

[0125] The concentration gradient is calculated using the central difference method. For each detection location, the difference in organic matter concentration in the horizontal and vertical directions, i.e., the gradient component, is calculated using the concentration values ​​of its four adjacent detection points in the east, west, south, and north.

[0126] Step S302: Determine the concentration gradient change rate based on the concentration difference of organic matter. If the concentration gradient change rate exceeds the preset concentration gradient change threshold, mark the position as the concentration mutation point.

[0127] The concentration gradient change rate is characterized by the magnitude of the gradient vector. When the magnitude exceeds 30% of the environmental baseline concentration, the location is marked as a concentration abrupt change point. These abrupt change points usually appear at the edge of the oil spill contamination area.

[0128] Step S303: Determine the concentration distribution data based on the organic matter concentration and concentration abrupt change points.

[0129] The concentration distribution data is determined by using the concentration mutation point as the boundary of the concentration distribution data and combining the organic matter concentration values ​​and spatial location information at each detection location.

[0130] Step S304: Based on the concentration distribution data, use the density clustering algorithm to spatially aggregate the concentration exceeding the concentration threshold of organic matter to obtain multiple cluster spaces, and determine the average concentration value of each cluster space.

[0131] Among them, the density clustering algorithm is a density-based spatial clustering method when dealing with concentration exceeding the standard. The density-based spatial clustering method sets the search radius to the distance between two adjacent detection points, and sets the minimum number of cluster points to 3. For each concentration exceeding the standard point, the algorithm searches for other exceeding the standard points in its neighborhood. When the number of exceeding the standard points in the neighborhood meets the density requirement, these points are assigned to the same cluster, resulting in multiple cluster spaces.

[0132] Meanwhile, the average concentration value of each detection point within each cluster space is used as the characteristic concentration of that cluster space.

[0133] Step S305: Iteratively expand multiple cluster spaces to generate continuous pollution areas as areas with abnormally high concentrations.

[0134] The iterative expansion process employs a region growing algorithm, using the initial cluster space as the core and gradually merging adjacent regions with similar concentrations. The algorithm first calculates the boundary concentration gradient of each cluster space. When the concentration difference between adjacent regions is less than a preset threshold (usually set to 20% of the environmental background concentration) and the spatial distance is less than 1.5 times the side length of the grid cell, these two regions are merged into a new expanded region. During the expansion process, the average concentration value of the regions is continuously updated to ensure the consistency of concentration characteristics after merging. Through multiple iterations, until no new region meets the merging conditions, a continuous polluted region with clear boundaries is finally formed. This algorithm can effectively handle irregularly shaped pollution diffusion regions while avoiding misclassifying background noise as polluted areas. The generated continuous polluted region not only contains spatial extent information but also records the internal concentration distribution characteristics.

[0135] Step S306: Based on the average concentration value of each cluster space, the concentration of organic matter in the region with abnormal concentration is interpolated by the inverse distance weighted interpolation method, and the concentration contour lines are constructed according to the preset concentration interval through the region with abnormal concentration after interpolation.

[0136] The inverse distance weighted interpolation method is used to generate a continuous concentration field. For any undetected location within a region of abnormally high concentration, the concentration value is calculated by a weighted average of surrounding known detection points. The weight is inversely proportional to the square of the distance; closer detection points contribute more to the interpolation result. The interpolation process is performed on a regular grid with a grid resolution set to one-quarter of the detection point spacing to ensure sufficient spatial resolution in the generated concentration field.

[0137] Concentration contour lines are extracted from the interpolated grid data using the moving square algorithm. The moving square algorithm divides concentration values ​​into multiple levels according to preset intervals, with each level corresponding to a contour line. The contour lines are smoothed using cubic spline interpolation to eliminate jagged edges.

[0138] Specifically, the area with abnormally high concentrations is divided into multiple concentration level intervals, such as low, medium, and high concentration intervals, each corresponding to a specific concentration range. Based on the organic matter concentration value at each location obtained using the inverse distance weighted interpolation method, the corresponding concentration level interval is determined. Then, the moving square algorithm is used to search for all points within the area with abnormally high concentrations that meet the conditions of the same concentration level interval, and these points are connected to form the contour lines corresponding to that concentration level.

[0139] For key points on the contour lines, such as inflection points and intersections, cubic spline interpolation is used for smoothing to eliminate jagged edges, making the contour lines smoother and more continuous. This results in a clear and accurate concentration contour map, which can intuitively show the distribution of organic matter concentration in areas where concentrations exceed the standard.

[0140] Step S307: Generate a concentration anomaly distribution area map based on the concentration contour lines and the interpolated concentration anomaly exceeding the standard area.

[0141] The color rendering of the concentration anomaly distribution area map adopts a heat map color scheme, with concentration values ​​ranging from low to high corresponding to a gradient of five colors: blue, green, yellow, orange, and red, enabling technicians to intuitively identify the spatial distribution characteristics of concentration anomalies.

[0142] In some embodiments, the suspected oil leak area is compared and matched with an anomaly concentration area map to determine a comprehensive matching score regarding location overlap and boundary consistency, including:

[0143] Step S401: Unify the suspected oil leak area and the concentration anomaly area map to the same geographical reference system to obtain the intermediate suspected oil leak area and intermediate concentration anomaly area map.

[0144] The coordinate system transformation process employs a seven-parameter transformation model, including three translation parameters, three rotation parameters, and one scale parameter. Satellite imagery typically uses a projected coordinate system, while UAV detection results use a geographic coordinate system. The transformation parameters are calculated using pre-calibrated control points. During the transformation, the transformation error at each coordinate point is optimized using the least squares method to ensure that the accuracy of the transformed coordinates meets the requirements of subsequent spatial analysis.

[0145] Step S402: Extract the boundary point sequence of the suspected oil leak area in the middle, and use the boundary point sequence to determine the first polygonal geometric object.

[0146] The extraction of the boundary point sequence of the suspected oil leak area can be achieved using an eight-neighbor boundary tracking algorithm. Starting from any boundary point of the suspected oil leak area, all boundary pixels are traversed in a clockwise direction, and their coordinate values ​​are recorded to obtain the boundary point sequence of the suspected oil leak area.

[0147] The first polygonal geometric object is constructed based on an ordered sequence of boundary points. A closed polygonal outline is formed by sequentially connecting adjacent points in the boundary point sequence. The area and centroid position are obtained by calculating the geometric moments of the first polygonal geometric object. The area is calculated using Green's formula, and the centroid coordinates are determined by the ratio of the first moment to the zeroth moment. These spatial feature data provide the basic parameters for subsequent matching analysis.

[0148] Step S403: Extract the contour coordinates of the concentration anomaly area map, and use the Douglas-Puk algorithm to simplify the contour coordinates to generate a second polygonal geometric object.

[0149] Similarly, extracting the contour coordinates of the concentration anomaly region map involves first identifying all boundary pixels in the concentration anomaly region map and recording their coordinate values ​​to form an initial contour point set.

[0150] To reduce data redundancy, the Douglas-Pock algorithm is used to simplify the initial contour point set. The Douglas-Pock algorithm in contour simplification follows the principle of recursive divide-and-conquer. The algorithm first connects the start and end points of the contour to form a baseline, calculates the vertical distance from all points on the contour to the baseline, and finds the point with the largest distance. If this maximum distance exceeds a preset tolerance threshold, the point is used as a dividing point, splitting the contour into two segments, and the same process is recursively performed on each segment. If the maximum distance is less than the tolerance threshold, all intermediate points between the start and end points are deleted. The tolerance threshold needs to balance the degree of simplification and shape fidelity, and is typically set to one-thousandth of the original contour perimeter. Through this simplification, the number of contour points in the UAV detection area is reduced from thousands to hundreds. While maintaining the main shape features of the contour, the computational complexity of subsequent spatial operations is significantly reduced. The simplified contour can still accurately reflect the spatial distribution characteristics of the concentration anomaly area, and a second polygonal geometric object is generated based on the ordered contour coordinate points.

[0151] Step S404: Perform a spatial superposition operation on the first polygonal geometric object and the second polygonal geometric object to obtain the ratio of the area of ​​the intersection region to the area of ​​the union region between the first polygonal geometric object and the second polygonal geometric object, which is used as the positional overlap degree.

[0152] The spatial overlay operation uses a computational geometry library to implement Boolean operations on polygons. The intersection operation uses a scanline algorithm to determine the overlapping portion of two polygons, generating a new polygon object to represent the shared coverage area. The union operation merges all areas covered by the two polygons, including their individual parts and overlapping parts. The positional overlap is calculated by dividing the intersection area by the union area; this ratio ranges from 0 to 1, with a higher value indicating a greater degree of spatial overlap between the two detection results.

[0153] Step S405: Use the Hausdorff distance between the boundary points of the first polygonal geometric object and the second polygonal geometric object as the boundary consistency index.

[0154] The Hausdorff distance calculation evaluates the maximum deviation between the boundaries of two polygonal objects. For each point on the first polygonal geometry, the minimum distance from it to the boundary of the second polygonal geometry is calculated, and the maximum value among all minimum distances is taken as the one-way Hausdorff distance.

[0155] The one-way Hausdorff distance from the boundary points of the second polygonal geometry to the boundary points of the first polygonal geometry is calculated in reverse. The larger of the two one-way distances is the two-way Hausdorff distance. This distance metric reflects the worst-case deviation between the two boundaries and is sensitive to significant local differences at the boundaries. In practical calculations, a spatial index structure for the boundary points is constructed to accelerate the nearest neighbor search, reducing the computational complexity from quadratic to logarithmic. The Hausdorff distance is measured in meters and directly reflects the absolute error of the two detection methods in boundary localization.

[0156] Step S406: Normalize the position overlap and boundary consistency indices.

[0157] The normalization process converts the Hausdorff distance into a dimensionless index between 0 and 1. A negative exponential function is used for normalization; the smaller the distance, the closer the normalized value is to 1, indicating better boundary consistency. The normalization parameters are determined based on the typical dimensions of the substation equipment and the required testing accuracy.

[0158] Step S407: Obtain the preset weights corresponding to the normalized position overlap and the normalized boundary consistency index, and perform weighted calculation on the normalized position overlap and the normalized boundary consistency index based on the preset weights to obtain the comprehensive matching score.

[0159] The weighting mechanism considers the relative importance of different matching indicators. Location overlap reflects the consistency of overall coverage and is typically assigned a weight of 0.6. Boundary consistency reflects the degree of conformity in contour shape and is assigned a weight of 0.4. These weights can be adjusted according to specific application scenarios; location overlap is weighted more when area coverage is a greater concern, while boundary consistency is weighted more when boundary accuracy is a greater concern. The overall score is obtained through weighted summation, with a score ranging from 0 to 1.

[0160] In some embodiments, when the overall matching score is higher than a preset standard score threshold, a comprehensive assessment is performed on the suspected oil leak area and the area with abnormally high concentration to determine the actual oil leak area, including:

[0161] Step S501: If the overall matching score is higher than the preset standard score threshold, determine the spectral attenuation degree based on the surface reflected light brightness data, and determine the excess multiple of the concentration peak based on the concentration distribution data.

[0162] The preset standard scoring threshold can be 0.7. When the comprehensive matching score exceeds 0.7, it is considered that the two detection methods have obtained highly consistent results, providing double verification for the reliable identification of the oil leak area.

[0163] The degree of spectral attenuation is obtained by calculating the percentage decrease in surface reflectance data in the near-infrared band relative to the baseline surface reflectance data of normal soil.

[0164] The exceedance multiple of the concentration peak is the ratio of the highest detected organic matter concentration to the environmental background concentration.

[0165] Step S502: Normalize the degree of spectral attenuation and the multiple of the concentration peak exceeding the standard, and then weight the degree of spectral attenuation and the multiple of the concentration peak exceeding the standard after normalization to obtain the oil leakage confidence value.

[0166] In this process, the degree of spectral attenuation and the multiple of exceedance of concentration peaks are converted into standardized values ​​between 0 and 1 through normalization.

[0167] The weighting of spectral attenuation and the exceedance multiple of concentration peaks can be assigned using evidence theory, taking into account the reliability differences between the two detection methods. Satellite detection, with its macroscopic coverage advantage, is assigned a weight of 0.4; UAV detection, with its high-precision direct measurement advantage, is assigned a weight of 0.6.

[0168] Step S503: Compare the oil leak confidence value with the preset confidence threshold. If the oil leak confidence value exceeds the preset confidence threshold, then the suspected oil leak area is confirmed as the actual oil leak area.

[0169] The confidence threshold can be set to 0.75, which is determined based on statistical analysis of historical detection data. When the confidence value of the oil leak exceeds the confidence threshold, the suspected oil leak area is confirmed as the actual oil leak area, and the geographical boundary, center coordinates, coverage area, pollution level, and confidence value of the actual oil leak area are recorded. This information constitutes a complete record of the oil leak event, providing comprehensive data support for maintenance decisions.

[0170] In some embodiments, the method further includes:

[0171] Step S601: Determine the spatial offset vector based on the spatial difference between the centroid coordinates of the actual oil leak area and the centroid coordinates of the area with abnormally high concentration.

[0172] Among them, the spatial offset vector is calculated by the spatial difference between the centroid coordinates of the actual oil leak area and the centroid coordinates of the area with abnormally high concentration. This vector reflects the systematic deviation of the two detection methods in spatial positioning and serves as the dominant direction of diffusion.

[0173] Step S602: Obtain the actual oil leakage area at at least two sampling times, and determine the oil leakage diffusion rate based on the area change of the actual oil leakage area at adjacent sampling times and the sampling time interval.

[0174] The calculation of the oil spill diffusion rate is based on the actual oil spill area change over a continuous time series. First, the actual oil spill area contours at adjacent sampling times are extracted using multi-temporal detection data. Green's formula is then used to calculate the coverage area at each time point. The area change is obtained by subtracting the area at the previous time point from the area at the next time point; a positive change indicates an expansion of the oil spill area, while a negative change indicates a contraction of the contaminated area. The sampling time interval is determined by the UAV inspection cycle or the satellite revisit cycle, typically set to 30 minutes to 24 hours. The diffusion rate is obtained by dividing the area change by the sampling interval, expressed in square meters per hour. For non-uniform diffusion, the rate of change of area per unit perimeter is further calculated, which is the ratio of the area change to the perimeter of the area at the current time point. This indicator better reflects the speed of boundary expansion.

[0175] Step S603: Based on the oil spill diffusion rate and spatial offset vector, and combined with the current environmental parameter data, the spatial concentration distribution of organic matter is determined using a Gaussian diffusion model.

[0176] The Gaussian diffusion model is used to predict the diffusion process of pollutants in the atmosphere or soil. Based on the law of conservation of mass and Fick's law of diffusion, this model treats pollution sources as continuous point or area sources. The core parameters of the model include source strength, diffusion coefficient, and ambient wind speed.

[0177] Specifically, based on the oil spill diffusion rate and spatial offset vector, combined with current environmental parameter data, a Gaussian diffusion model is used to determine the spatial concentration distribution of organic matter, including:

[0178] Step S6031: Determine the center coordinates of the pollution source based on the spatial offset vector and the centroid coordinates of the actual oil leak area.

[0179] The spatial offset vector includes the horizontal offset vector Δx and the vertical offset vector Δy. The coordinates of the pollution source center are determined by superimposing the horizontal offset vector Δx and the vertical offset vector Δy on the centroid coordinates of the actual oil leak area. This coordinate serves as the origin of pollutant release in the Gaussian diffusion model, ensuring that the diffusion starting point simulated by the model is consistent with the actual oil leak center.

[0180] Step S6032: Determine the source strength based on the oil leakage diffusion rate.

[0181] Wherein, the source strength is equal to the product of the oil leak area and the evaporation rate, and the change of area over time is determined by the oil leak diffusion rate k, that is:

[0182] S(t) = S0 + k × t

[0183] Q(t) = S(t) × q

[0184] In the formula, S(t) represents the area of ​​the oil spill region, t represents time, Q(t) represents the source strength, and q represents the evaporation rate per unit area. The greater the source strength, the higher the peak concentration of pollutants simulated by the model, which aligns with the actual phenomenon that a larger oil spill area results in greater evaporation.

[0185] Step S6033: Extract the boundary feature points of the actual oil leakage area at adjacent sampling times, determine the displacement change rate of the two boundary feature points in the direction of the spatial offset vector, and determine the horizontal diffusion coefficient based on the displacement change rate.

[0186] The displacement change rate is calculated by tracking the displacement of feature points. Specifically, multiple feature points are uniformly selected on the leak boundary at the initial moment. A nearest neighbor matching algorithm is used to find corresponding points on the boundary at subsequent moments. The displacement distance of each feature point is calculated, the average value is obtained, and then divided by the time interval. That is:

[0187] v=ΔL / Δt

[0188] In the formula, v is the rate of change of displacement, and ΔL is the displacement change of the two boundary feature points.

[0189] Then establish the displacement change rate v and the horizontal diffusion coefficient The correlation is as follows: Based on empirical formulas from field experiments and numerical simulations, the change in the horizontal diffusion coefficient over time is directly proportional to the boundary advance velocity, i.e.:

[0190]

[0191] In the formula, The horizontal diffusion coefficient at the initial time (t=0) is taken as an empirical value based on atmospheric stability, such as 10-20 meters under neutral atmosphere. This is a proportionality coefficient (ranging from 0.3 to 0.5, depending on the complexity of the local terrain; 0.5 for flat terrain and 0.3 for complex terrain).

[0192] The faster the horizontal diffusion coefficient increases, the larger the horizontal diffusion range simulated by the model, which closely matches the dual characteristics of active expansion and atmospheric diffusion in actual oil leaks.

[0193] Step S6034: Determine the spatial concentration distribution of organic matter based on the pollution source center coordinates, source strength, horizontal diffusion coefficient, and current environmental parameter data, combined with the Gaussian diffusion model; wherein, the current environmental parameter data includes the average wind speed.

[0194] The Gaussian diffusion model is as follows:

[0195]

[0196] In the formula, To predict the concentration of organic matter at time t and coordinates (x, y), The average wind speed, This is the vertical diffusion coefficient (take an empirical value, 30-50). The ordinate of the pollution source center is the vertical coordinate. To predict the height of the point, The effective height of the pollution source.

[0197] Specifically, by substituting all spatial points (x,y) within the prediction area into the Gaussian diffusion model, the organic matter concentrations at all spatial points (x,y) are obtained, thus yielding the spatial concentration distribution of organic matter.

[0198] Step S604: Based on the spatial concentration distribution, select the boundary of the region where the organic matter concentration is greater than the preset environmental concentration threshold, and obtain the predicted oil spill impact range and the corresponding oil spill impact area.

[0199] The preset environmental concentration threshold, such as 0.1, is used to identify points where the organic matter concentration exceeds this threshold as affected areas. Connecting these points to form a closed curve determines the boundary of the predicted oil spill impact area. The area enclosed by this boundary is the predicted oil spill impact area. This area data visually reflects the potential size of the oil spill's reach, providing crucial reference for subsequent emergency response and resource allocation. Simultaneously, using the pollution source center coordinates as the origin, the distances required for the concentration to drop to the background value (environmental concentration threshold) in different directions are calculated, and the maximum distance in each direction is taken as the impact radius.

[0200] Step S605: Determine the concentration exceedance multiple based on the ratio of organic matter concentration to the preset environmental concentration threshold.

[0201] The concentration exceedance multiple is obtained by dividing the organic matter concentration at each spatial point by a preset environmental concentration threshold. This indicator directly reflects the severity of the oil spill pollution. When the ratio is greater than 1, it indicates that the organic matter concentration in the area has exceeded the safety standard; the larger the ratio, the more severe the pollution. By calculating the average concentration exceedance multiple of all areas exceeding the standard, the overall pollution level can be quantified, providing a scientific basis for assessing environmental risks and developing remediation plans.

[0202] Step S606: Normalize the oil spill diffusion rate, oil spill area, and concentration exceedance multiple, and then weight the normalized oil spill diffusion rate, normalized oil spill area, and normalized concentration exceedance multiple to obtain the oil spill impact score.

[0203] The pollution impact radius is weighted at 0.4, with full marks awarded when the radius exceeds 500 meters; the oil spill diffusion rate is weighted at 0.3, with full marks awarded when the diffusion rate exceeds 50 square meters per hour; and the concentration exceeding the standard multiple is weighted at 0.3, with full marks awarded when the highest concentration exceeds the standard by 10 times. The oil spill impact score is obtained by weighting and summing the normalized indicators, and this score ranges from 0 to 1.

[0204] Step S607: Determine the impact level of the oil leak accident based on the oil leak impact score.

[0205] The higher the oil spill impact score, the more severe the impact of the oil spill, and the more urgent and comprehensive the response measures required. For example, when the oil spill impact score exceeds 0.8, the highest level of emergency response mechanism should be activated immediately, mobilizing all available resources for oil spill control, pollution cleanup, and ecological restoration.

[0206] For example, an oil spill impact score of 0 to 0.3 is classified as a general accident, 0.3 to 0.7 as a major accident, and 0.7 to 1.0 as a serious accident. This classification mechanism comprehensively considers three dimensions: spatial scope, temporal evolution, and pollution intensity. Through the above spatiotemporal evolution analysis and accident level assessment, the system can accurately grasp the development trend of oil spill accidents, providing quantitative decision support for formulating targeted emergency response plans and achieving effective protection of substation equipment safety.

[0207] In some embodiments, the method further includes:

[0208] Step S701: Determine the equipment maintenance urgency index based on the impact level of the oil leak accident.

[0209] The equipment maintenance urgency index is positively correlated with the impact level of an oil leak. When an oil leak is assessed as a general accident, the equipment maintenance urgency index is set to a low value, such as between 0.2 and 0.4, indicating that it can be handled within the regular maintenance cycle. When assessed as a major accident, the urgency index increases to the range of 0.4 to 0.7, requiring a special inspection and repair plan to be initiated within 24 hours. When it reaches the level of a serious accident, the urgency index is directly assigned a value of 1.0, requiring the immediate shutdown of the relevant equipment and the simultaneous implementation of emergency response and root cause analysis.

[0210] The equipment maintenance urgency index quantifies the correlation between incident impact and maintenance response, providing a tiered decision-making basis for operation and maintenance resource allocation and ensuring that high-risk events receive priority resources. Specifically, linear interpolation can be used to map the oil leak impact score to an urgency index range of 0-1.

[0211] Step S702: If the equipment maintenance urgency index is greater than the preset urgency index threshold, an emergency maintenance task is generated, and the center coordinates of the actual oil leak area are extracted.

[0212] The preset urgency index threshold can be set to 0.6 based on actual maintenance needs. When the equipment maintenance urgency index exceeds this threshold, an emergency maintenance task generation process is automatically triggered. At this time, the centroid coordinates of the actual oil leak area are extracted using the geographic information system (GIS) positioning module as the maintenance work reference point. This coordinate data is simultaneously pushed to the maintenance management platform for planning the optimal repair route and allocating surrounding emergency resources. Alternatively, the center coordinates of the actual oil leak area can be obtained by calculating the arithmetic mean of all boundary point coordinates. The boundary range data includes the coordinates of the four vertices of the smallest bounding rectangle.

[0213] Step S703: Based on the geographical distribution of equipment within the substation, determine the equipment closest to the center coordinates of the actual oil leak area as the potential hazard equipment, and extract the geographical coordinates of the potential hazard equipment.

[0214] Specifically, a spatial database of substation equipment is constructed, and a spatial proximity analysis algorithm is used to calculate the Euclidean distance between the centroid coordinates of the actual oil leak area and the coordinates of each piece of equipment. The system automatically selects the nearest equipment as the potential hazard and records its equipment number, type, and precise geographical coordinates. This coordinate data will be spatially correlated with the parameters of the oil leak area, providing basic location information for subsequent pollution diffusion path analysis and the formulation of equipment protection measures.

[0215] Step S704: Generate an equipment maintenance urgency report based on the emergency maintenance task, the geographical coordinates of the potentially hazardous equipment, and the impact level of the oil leak accident.

[0216] The equipment maintenance urgency report adopts a structured format, which includes the emergency maintenance task, the geographical coordinates of the equipment with potential hazards, the impact level of the oil leak accident, the coordinates of the oil leak center, the radius of impact, the name and distance of the nearest equipment, and the recommended handling measures. The identification results of the equipment with potential hazards are presented in the form of geographic information layers.

[0217] Based on the same inventive concept, this application also provides a ground equipment oil leak hazard identification system for implementing the above-mentioned ground equipment oil leak hazard identification method.

[0218] The solution provided by this system is similar to the solution described in the above method. Therefore, the specific limitations of one or more ground equipment oil leak hazard identification system embodiments provided below can be found in the limitations of the ground equipment oil leak hazard identification method above, and will not be repeated here.

[0219] like Figure 2As shown in the figure, this application provides a ground equipment oil leak hazard identification system, including:

[0220] The suspected oil leak detection module 100 is used to collect surface reflectance data of the ground area of ​​the target equipment through satellite multispectral channels, identify suspected oil leak areas based on the surface reflectance data, and extract the location coordinates and boundary range of the suspected oil leak areas.

[0221] The area inspection module 200 is used to plan the inspection flight route based on the location coordinates and boundary range of the suspected oil leak area, and to use a drone to detect the organic matter concentration at each location of the suspected oil leak area according to the inspection flight route.

[0222] The concentration anomaly determination module 300 is used to determine the concentration distribution data based on the organic matter concentration detected at each location, identify the concentration anomaly exceeding the standard area based on the concentration distribution data, and generate a concentration anomaly distribution area map corresponding to the concentration anomaly exceeding the standard area.

[0223] The comprehensive scoring module 400 is used to compare and match suspected oil leak areas with areas of abnormal concentration to determine a comprehensive matching score regarding location overlap and boundary consistency.

[0224] The actual oil leak determination module 500 is used to comprehensively evaluate suspected oil leak areas and areas with abnormally high concentrations when the comprehensive matching score is higher than the preset standard score threshold, and to determine the actual oil leak area.

[0225] In some embodiments, the suspected oil leak detection module 100 is used for:

[0226] Satellite multispectral imagery data of the ground area of ​​the target equipment is acquired through satellite multispectral channels;

[0227] Based on satellite multispectral imagery data, the surface reflectance brightness data for each band were determined; the bands include the red band and the near-infrared band.

[0228] The surface reflected light intensity data of each band were normalized to obtain the normalized surface reflected light intensity data of each band.

[0229] For each pixel in the satellite multispectral image data, the reflectance ratio is determined based on the normalized surface reflectance data of the near-infrared band and the normalized surface reflectance data of the red band.

[0230] If the ratio of reflected light brightness is less than the preset brightness ratio threshold, then the pixel is marked as a potential oil leakage point;

[0231] The suspected oil leak areas are obtained by spatial clustering of all marked potential oil leak points using a connected component analysis algorithm.

[0232] Based on the surface reflectance data of each pixel in the suspected oil spill area, determine the brightness gradient value between adjacent pixels;

[0233] Based on the brightness gradient values ​​between adjacent pixels, the pixel with the largest brightness gradient value in the suspected oil leak area is determined as the boundary point. All boundary points are connected to determine the boundary range of the suspected oil leak area. The suspected oil leak area is then transformed from pixel coordinate system to geographic coordinate system to obtain the location coordinates of the suspected oil leak area.

[0234] In some embodiments, the area inspection module 200 is used for:

[0235] Based on the location coordinates and boundary range of the suspected oil leak area, the suspected oil leak area is divided into grids using a rasterization method to obtain the rasterized area of ​​the suspected oil leak area.

[0236] The circumscribed rectangle of the rasterized region is determined, and a rectangular grid array that at least covers the suspected oil leak area is generated based on the circumscribed rectangle. Each grid in the rectangular grid array records an index number, center coordinates, and an identifier indicating whether it is located inside the suspected oil leak area. The circumscribed rectangle is the smallest rectangular range defined by the bounding box of the rasterized region.

[0237] Based on the rectangular grid array, a serpentine path planning algorithm is used to generate the inspection flight route of the UAV. The inspection flight route includes each inspection point on the route and the preset flight altitude of the UAV at each inspection point. The preset flight altitude is determined based on the vertical distance between the UAV's flight altitude and the center coordinates of the grid corresponding to the inspection point.

[0238] The concentration of organic matter in each grid within the suspected oil leak area was detected by drones following the inspection flight route.

[0239] In some embodiments, the concentration anomaly determination module 300 is used for:

[0240] Based on the organic matter concentration detected at each location, determine the organic matter concentration difference between adjacent locations;

[0241] The concentration gradient change rate is determined based on the concentration difference of organic matter. If the concentration gradient change rate exceeds the preset concentration gradient change threshold, the position is marked as a concentration change point.

[0242] Based on the concentration of organic matter and the points of abrupt changes in concentration, the concentration distribution data is determined;

[0243] Based on the concentration distribution data, a density clustering algorithm was used to spatially aggregate the concentration points where the organic matter concentration was greater than the concentration exceeding the threshold, resulting in multiple cluster spaces, and the average concentration value of each cluster space was determined.

[0244] Multiple cluster spaces are iteratively expanded to generate continuous pollution areas as areas with abnormally high concentrations.

[0245] Based on the average concentration value of each cluster space, the concentration of organic matter in the region of abnormal concentration exceedance is interpolated by the inverse distance weight interpolation method, and the concentration contour lines are constructed according to the preset concentration interval through the interpolated region of abnormal concentration exceedance.

[0246] Based on the concentration contour lines and the interpolated areas of abnormal concentration, a map of the concentration anomaly distribution area is generated.

[0247] In some embodiments, the comprehensive scoring module 400 is used for:

[0248] By unifying the maps of suspected oil spill areas and areas with abnormal concentrations to the same geographic reference system, we can obtain intermediate maps of suspected oil spill areas and intermediate areas with abnormal concentrations.

[0249] Extract the boundary point sequence of the suspected oil leak area in the middle, and use the boundary point sequence to determine the first polygonal geometric object;

[0250] The contour coordinates of the concentration anomaly area map are extracted, and the Douglas-Puk algorithm is used to simplify the contour coordinates to generate a second polygonal geometric object.

[0251] A spatial superposition operation is performed on the first polygonal geometric object and the second polygonal geometric object to obtain the ratio of the area of ​​the intersection region to the area of ​​the union region between the first polygonal geometric object and the second polygonal geometric object, which is used as the positional overlap degree.

[0252] The Hausdorff distance between the boundary points of the first and second polygonal geometric objects is used as a boundary consistency index.

[0253] Normalize the positional overlap and boundary consistency indices;

[0254] Obtain the preset weights corresponding to the normalized position overlap and the normalized boundary consistency index, and calculate the weighted average of the normalized position overlap and the normalized boundary consistency index based on the preset weights to obtain the comprehensive matching score.

[0255] In some embodiments, the actual oil leak determination module 500 is used for:

[0256] If the overall matching score is higher than the preset standard score threshold, the degree of spectral attenuation is determined based on the surface reflected light brightness data, and the excess multiple of the concentration peak is determined based on the concentration distribution data.

[0257] The degree of spectral attenuation and the multiple of exceedance of the concentration peak were normalized, and the degree of spectral attenuation and the multiple of exceedance of the concentration peak after normalization were weighted to obtain the oil leakage confidence value.

[0258] The oil leak confidence value is compared with a preset confidence threshold. If the oil leak confidence value exceeds the preset confidence threshold, the suspected oil leak area is confirmed as the actual oil leak area.

[0259] In some embodiments, the system further includes:

[0260] The offset vector determination module is used to determine the spatial offset vector based on the spatial difference between the centroid coordinates of the actual oil leak area and the centroid coordinates of the area with abnormally high concentration.

[0261] The oil leakage diffusion rate determination module is used to obtain the actual oil leakage area at at least two sampling times, and determine the oil leakage diffusion rate based on the area change of the actual oil leakage area at adjacent sampling times and the sampling time interval.

[0262] The spatial concentration determination module is used to determine the spatial concentration distribution of organic matter based on the oil spill diffusion rate and spatial offset vector, combined with current environmental parameter data, using a Gaussian diffusion model.

[0263] The impact range determination module is used to filter out the boundaries of areas where the organic matter concentration is greater than the preset environmental concentration threshold based on the spatial concentration distribution, and obtain the predicted oil spill impact range and the corresponding oil spill impact area.

[0264] The concentration exceedance multiple calculation module is used to determine the concentration exceedance multiple based on the ratio of organic matter concentration to a preset environmental concentration threshold.

[0265] The influence score determination module is used to normalize the oil spill diffusion rate, oil spill area, and concentration exceedance multiple, and then weight the normalized oil spill diffusion rate, normalized oil spill area, and normalized concentration exceedance multiple to obtain the oil spill influence score.

[0266] The impact level determination module is used to determine the impact level of an oil spill accident based on the oil spill impact score.

[0267] In some embodiments, the spatial concentration determination module is used for:

[0268] The coordinates of the pollution source center are determined based on the spatial offset vector and the centroid coordinates of the actual oil leak area;

[0269] Determine the source strength based on the oil spill diffusion rate;

[0270] Extract the boundary feature points of the actual oil leakage area at adjacent sampling times, determine the displacement change rate of the two boundary feature points in the direction of the spatial offset vector, and determine the horizontal diffusion coefficient based on the displacement change rate.

[0271] Based on the pollution source center coordinates, source strength, horizontal diffusion coefficient, and current environmental parameter data, the spatial concentration distribution of organic matter is determined using a Gaussian diffusion model; the current environmental parameter data includes average wind speed.

[0272] In some embodiments, the system further includes: a report generation module, used for:

[0273] The urgency index of equipment maintenance is determined based on the impact level of the oil spill accident;

[0274] If the equipment maintenance urgency index is greater than the preset urgency index threshold, an emergency maintenance task will be generated, and the center coordinates of the actual oil leak area will be extracted.

[0275] Based on the geographical distribution of equipment within the substation, the equipment closest to the center coordinates of the actual oil leak area is identified as the potential hazard equipment, and the geographical coordinates of the potential hazard equipment are extracted.

[0276] Based on the emergency maintenance task, the geographical coordinates of the potentially hazardous equipment, and the impact level of the oil leak, an equipment maintenance urgency report is generated.

[0277] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0278] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0279] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0280] In the several embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0281] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. 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 of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying potential oil leaks in ground equipment, characterized in that, include: The surface reflectance data of the target equipment's ground area is collected by satellite multispectral channels, and suspected oil leak areas are identified based on the surface reflectance data. The location coordinates and boundary range of the suspected oil leak areas are then extracted. Based on the location coordinates and boundary range of the suspected oil leak area, an inspection flight route is planned, and the concentration of organic matter in each location of the suspected oil leak area is detected by drone according to the inspection flight route. Based on the organic matter concentration detected at each location, concentration distribution data is determined, and areas with abnormally high concentrations are identified based on the concentration distribution data. A concentration abnormality distribution area map corresponding to the areas with abnormally high concentrations is then generated. The suspected oil leak area is compared and matched with the concentration anomaly area map to determine a comprehensive matching score regarding location overlap and boundary consistency; If the overall matching score is higher than the preset standard score threshold, the suspected oil leak area and the area with abnormally high concentration are comprehensively evaluated to determine the actual oil leak area.

2. The method for identifying potential oil leaks in ground equipment according to claim 1, characterized in that, The process involves acquiring surface reflectance data of the target equipment's ground area via satellite multispectral channels, identifying suspected oil leak areas based on this data, and extracting the location coordinates and boundary range of the suspected oil leak areas, including: Satellite multispectral imagery data of the ground area of ​​the target equipment is acquired through satellite multispectral channels; Based on the satellite multispectral image data, the surface reflectance brightness data for each band is determined; wherein, the bands include the red band and the near-infrared band; The surface reflected light intensity data of each band are normalized to obtain the normalized surface reflected light intensity data of each band. For each pixel in the satellite multispectral image data, the reflectance ratio is determined based on the normalized surface reflectance data of the near-infrared band and the normalized surface reflectance data of the red band. If the reflected light brightness ratio is less than a preset brightness ratio threshold, then the pixel is marked as a potential oil leakage point; The suspected oil leak area is obtained by spatial clustering of all the marked potential oil leak points using a connected component analysis algorithm. Based on the surface reflectance data of each pixel in the suspected oil leak area, determine the brightness gradient value between each adjacent pixel; Based on the brightness gradient values ​​between adjacent pixels, the pixel with the largest brightness gradient value in the suspected oil leak area is determined as the boundary point. All boundary points are connected to determine the boundary range of the suspected oil leak area. The suspected oil leak area is then transformed from pixel coordinate system to geographic coordinate system to obtain the location coordinates of the suspected oil leak area.

3. The method for identifying potential oil leaks in ground equipment according to claim 1, characterized in that, The process involves planning an inspection flight route based on the location coordinates and boundary range of the suspected oil leak area, and then using a drone to detect the organic matter concentration at each location within the suspected oil leak area according to the inspection flight route, including: Based on the location coordinates and boundary range of the suspected oil leak area, the suspected oil leak area is divided into grids using a rasterization method to obtain the rasterized area of ​​the suspected oil leak area. By determining the circumscribed rectangle of the rasterized region, a rectangular grid array that at least covers the suspected oil leak area is generated based on the circumscribed rectangle; wherein, each grid in the rectangular grid array records an index number, center coordinates, and an identifier indicating whether it is located inside the suspected oil leak area; wherein, the circumscribed rectangle is the smallest rectangular range defined by the bounding box of the rasterized region; Based on the rectangular grid array, a serpentine path planning algorithm is used to generate an inspection flight route for the UAV; wherein, the inspection flight route includes each inspection point on the route and a preset flight altitude of the UAV at each inspection point; wherein, the preset flight altitude is determined based on the vertical distance between the flight altitude of the UAV and the center coordinates of the grid corresponding to the inspection point; The drone uses the inspection flight route to detect the concentration of organic matter at each location in the suspected oil leak area.

4. The method for identifying potential oil leaks in ground equipment according to claim 1, characterized in that, The step of determining concentration distribution data based on the detected organic matter concentrations at each location, identifying areas with abnormally high concentrations based on the concentration distribution data, and generating a concentration anomaly distribution area map corresponding to the areas with abnormally high concentrations includes: Based on the organic matter concentration detected at each location, determine the organic matter concentration difference between adjacent locations; The concentration gradient change rate is determined based on the concentration difference of the organic matter. If the concentration gradient change rate exceeds a preset concentration gradient change threshold, the position is marked as a concentration mutation point. The concentration distribution data is determined based on the organic matter concentration and the concentration abrupt change points; Based on the concentration distribution data, a density clustering algorithm is used to spatially aggregate the concentration exceeding the concentration threshold of the organic matter to obtain multiple cluster spaces, and the average concentration value of each cluster space is determined. The clustering spaces are iteratively expanded to generate continuous pollution regions as areas with abnormally high concentrations. Based on the average concentration value of each cluster space, the concentration of organic matter in the concentration abnormality area is interpolated by the inverse distance weighted interpolation method, and a concentration contour line is constructed according to the preset concentration interval through the interpolated concentration abnormality area. Based on the concentration contour lines and the concentration anomaly exceeding the standard areas after interpolation, the concentration anomaly distribution area map is generated.

5. The method for identifying potential oil leaks in ground equipment according to claim 1, characterized in that, The step of comparing and matching the suspected oil leak area with the concentration anomaly area map to determine a comprehensive matching score regarding location overlap and boundary consistency includes: By unifying the suspected oil leak area and the concentration anomaly area map to the same geographic reference system, an intermediate suspected oil leak area map and an intermediate concentration anomaly area map are obtained. Extract the boundary point sequence of the suspected oil leak area in the middle, and use the boundary point sequence to determine the first polygonal geometric object; Extract the contour coordinates of the concentration anomaly area map, and use the Douglas-Puk algorithm to simplify the contour coordinates to generate a second polygonal geometric object; A spatial superposition operation is performed on the first polygonal geometric object and the second polygonal geometric object to obtain the ratio of the area of ​​the intersection region to the area of ​​the union region between the first polygonal geometric object and the second polygonal geometric object, which is used as the positional overlap degree. The Hausdorff distance between the boundary points of the first polygonal geometric object and the second polygonal geometric object is used as a boundary consistency index. The positional overlap and the boundary consistency index are normalized. Obtain the preset weights corresponding to the normalized position overlap and the normalized boundary consistency index, and perform a weighted calculation on the normalized position overlap and the normalized boundary consistency index based on the preset weights to obtain the comprehensive matching degree score.

6. The method for identifying potential oil leaks in ground equipment according to claim 1, characterized in that, When the overall matching score is higher than a preset standard score threshold, a comprehensive evaluation is performed on the suspected oil leak area and the area with abnormally high concentration to determine the actual oil leak area, including: If the overall matching score is higher than the preset standard score threshold, the spectral attenuation degree is determined based on the surface reflected light brightness data, and the excess multiple of the concentration peak is determined based on the concentration distribution data. The degree of spectral attenuation and the multiple of exceedance of the concentration peak are normalized, and the normalized degree of spectral attenuation and the multiple of exceedance of the normalized concentration peak are weighted to obtain the oil leakage confidence value. The oil leak confidence value is compared with a preset confidence threshold. If the oil leak confidence value exceeds the preset confidence threshold, the suspected oil leak area is confirmed as the actual oil leak area.

7. The method for identifying potential oil leaks in ground equipment according to claim 1, characterized in that, Also includes: The spatial offset vector is determined based on the spatial difference between the centroid coordinates of the actual oil leak area and the centroid coordinates of the area with abnormally high concentration. Obtain the actual oil leakage area at at least two sampling times, and determine the oil leakage diffusion rate based on the area change of the actual oil leakage area at adjacent sampling times and the sampling time interval. Based on the oil spill diffusion rate and the spatial offset vector, combined with the current environmental parameter data, the spatial concentration distribution of organic matter is determined using a Gaussian diffusion model. Based on the spatial concentration distribution, the boundaries of areas where the organic matter concentration is greater than the preset environmental concentration threshold are selected to obtain the predicted oil spill impact range and the oil spill impact area corresponding to the predicted oil spill impact range. The concentration exceedance multiple is determined based on the ratio of the organic matter concentration to the preset environmental concentration threshold. The oil spill diffusion rate, the oil spill area, and the concentration exceeding the standard multiple are normalized, and the normalized oil spill diffusion rate, the normalized oil spill area, and the normalized concentration exceeding the standard multiple are weighted to obtain the oil spill impact score. The impact level of the oil leak accident is determined based on the oil leak impact score.

8. The method for identifying potential oil leaks in ground equipment according to claim 7, characterized in that, The step of determining the spatial concentration distribution of organic matter based on the oil spill diffusion rate and the spatial offset vector, combined with current environmental parameter data, using a Gaussian diffusion model includes: The coordinates of the pollution source center are determined based on the spatial offset vector and the centroid coordinates of the actual oil leak area. The source strength is determined based on the oil leakage diffusion rate. Extract the boundary feature points of the actual oil leakage area at adjacent sampling times, determine the displacement change rate of the two boundary feature points in the direction of the spatial offset vector, and determine the horizontal diffusion coefficient based on the displacement change rate. Based on the pollution source center coordinates, the source strength, the horizontal diffusion coefficient, and the current environmental parameter data, the spatial concentration distribution of organic matter is determined using the Gaussian diffusion model; wherein, the current environmental parameter data includes the average wind speed.

9. The method for identifying potential oil leaks in ground equipment according to claim 7, characterized in that, Also includes: Based on the impact level of the oil leak accident, determine the equipment maintenance urgency index; If the equipment maintenance urgency index is greater than the preset urgency index threshold, an emergency maintenance task is generated, and the center coordinates of the actual oil leak area are extracted. Based on the geographical distribution of equipment within the substation, the equipment closest to the center coordinates of the actual oil leak area is identified as the potential hazard equipment, and the geographical coordinates of the potential hazard equipment are extracted. Based on the emergency maintenance task, the geographical coordinates of the potentially hazardous equipment, and the impact level of the oil leak, an equipment maintenance urgency report is generated.

10. A ground equipment oil leak hazard identification system, characterized in that, include: The suspected oil leak detection module is used to collect surface reflectance data of the target equipment's ground area through satellite multispectral channels, identify suspected oil leak areas based on the surface reflectance data, and extract the location coordinates and boundary range of the suspected oil leak areas. The area inspection module is used to plan an inspection flight route based on the location coordinates and boundary range of the suspected oil leak area, and to use a drone to detect the organic matter concentration at each location of the suspected oil leak area according to the inspection flight route. The concentration anomaly determination module is used to determine the concentration distribution data based on the organic matter concentration detected at each location, identify the concentration anomaly exceeding the standard area based on the concentration distribution data, and generate a concentration anomaly distribution area map corresponding to the concentration anomaly exceeding the standard area. The comprehensive scoring module is used to compare and match the suspected oil leak area with the concentration anomaly area map to determine a comprehensive matching score regarding the degree of location overlap and boundary consistency. The actual oil leak determination module is used to comprehensively evaluate the suspected oil leak area and the area with abnormally high concentration when the comprehensive matching score is higher than the preset standard score threshold, and determine the actual oil leak area.

Images