Satellite remote sensing-based gas pipeline network leakage source quantitative inversion method and system

Abstract

The application discloses a kind of based on satellite remote sensing's gas pipe network leak source quantitative inversion method and system, it is related to gas remote sensing monitoring technical field, including, acquisition satellite remote sensing spectrum data and gas pipe network vector topological data, generate gas column concentration distribution data;Gas concentration in satellite remote sensing observation area is carried out abnormal identification, and gas abnormal area chart is output;Gas abnormal area chart and gas pipe network vector topological data are overlaid in space, and leak source probability distribution chart is output;Gas abnormal area chart is carried out geometric feature extraction by morphological analysis method, and gas diffusion dominant parameter is output;Gas diffusion dominant parameter and leak source probability distribution chart are coupled analysis;Gas column concentration distribution data and gas diffusion dominant parameter are carried out leakage flux inversion, and gas pipe network leakage is output.The application is cooperated by double mechanism of space superposition and ground flow field inversion, realizes the accurate positioning of gas pipe network leak source and the reliable inversion of leakage flux.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of gas remote sensing monitoring, and in particular to a gas pipeline network leakage source quantitative inversion method and system based on satellite remote sensing. BACKGROUND

[0002] In recent years, satellite remote sensing gas monitoring methods and gas pipeline network safety management have accelerated the integration, promoting the innovation of leakage detection methods. Current methods mainly rely on hyperspectral satellite data, use differential absorption spectrum algorithm to invert gas column concentration distribution, combine threshold segmentation and spatial clustering to identify concentration anomaly area, use pipeline vector data to constrain the possible range of leakage source, use morphological analysis to extract gas plume geometric features to assist in judging diffusion conditions, and based on Gaussian diffusion model or mass conservation principle to estimate leakage flux, forming a basic analysis framework from concentration identification to source intensity estimation.

[0003] However, the existing methods have limitations in positioning accuracy and inversion reliability. The pipeline constraint based on simple spatial superposition fails to effectively integrate the coupling relationship between real-time wind field and pipeline topology, the gas plume morphology analysis and background wind field inversion are disconnected, the leakage source probability evaluation lacks multi-source observation data collaborative verification mechanism, and the flux calculation does not consider the disturbance of complex underlying surface to the diffusion process, resulting in large deviation of leakage source positioning and high uncertainty of quantitative inversion results. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a gas pipeline network leakage source quantitative inversion method based on satellite remote sensing to solve the problem of limitations in positioning accuracy and inversion reliability.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides a satellite remote sensing-based gas pipeline network leakage source quantitative inversion method, which comprises the following steps: collecting satellite remote sensing spectral data and gas pipeline network vector topological data; performing radiation calibration and atmospheric correction processing on the satellite remote sensing spectral data by using a differential absorption spectrum algorithm to generate gas column concentration distribution data; identifying abnormal gas concentration in a satellite remote sensing observation area based on the gas column concentration distribution data, and outputting a gas abnormal area graph; superimposing the gas abnormal area graph and the gas pipeline network vector topological data in space to output gas superimposed data, and performing weighted statistics on the gas superimposed data to output a leakage source probability distribution graph; extracting geometric features of the gas abnormal area graph by using a morphological analysis method to obtain gas diffusion characteristic parameters, and performing ground flow field inversion on the gas diffusion characteristic parameters to output gas diffusion dominant parameters; coupling the gas diffusion dominant parameters and the leakage source probability distribution graph to form a leakage source probability graph, and performing extreme value positioning on the leakage source probability graph to determine a gas pipeline network leakage source position; and based on the gas pipeline network leakage source position, performing leakage flux inversion on the gas column concentration distribution data and the gas diffusion dominant parameters to output a gas pipeline network leakage amount.

[0008] As a preferred scheme of the satellite remote sensing-based gas pipeline network leakage source quantitative inversion method, the radiation calibration and atmospheric correction processing on the satellite remote sensing spectral data by using the differential absorption spectrum algorithm to generate the gas column concentration distribution data comprises the following steps:

[0009] Performing radiation calibration processing and geographical geometry calibration on the satellite remote sensing spectral data to generate calibrated remote sensing spectral data;

[0010] Performing atmospheric correction on the calibrated remote sensing spectral data by using a reflectivity method to output corrected spectral data;

[0011] Based on the corrected spectral data, extracting the absorption intensity of the gas at a characteristic wave band by using a differential absorption spectrum algorithm, and performing vertical column concentration conversion on the absorption intensity to generate the gas column concentration distribution data.

[0012] As a preferred scheme of the satellite remote sensing-based gas pipeline network leakage source quantitative inversion method, the abnormal gas concentration identification based on the gas column concentration distribution data to output the gas abnormal area graph comprises the following steps:

[0013] Comparing the gas column concentration distribution data with a preset gas concentration threshold value to determine abnormal gas points;

[0014] Aggregating and splicing the spatially continuous abnormal gas points by using a spatial clustering algorithm to output the gas abnormal area graph.

[0015] As a preferred scheme of the quantitative inversion method of the gas pipeline network leakage source based on satellite remote sensing, wherein: the gas anomaly area graph is spatially overlaid with the gas pipeline network vector topological data, gas overlay data is output, and the gas overlay data is weighted and counted to output a leakage source probability distribution graph, and the specific steps are,

[0016] The gas anomaly area graph is spatially aligned with the gas pipeline network vector topological data, and is integrated to form a gas pipeline network leakage data pair;

[0017] Attribute association and geometric matching are performed on the gas pipeline network leakage data pair to output gas overlay data;

[0018] The pipeline attribute information and abnormal region information of the gas pipeline network are extracted from the gas overlay data, and the pipeline distance weight value of the satellite remote sensing observation area is calculated through a distance decay function;

[0019] The gas column concentration in the gas anomaly area graph is multiplied by the pipeline distance weight value to output a leakage source probability, and the leakage source probability is gridded by a spatial interpolation method to output a leakage source probability distribution graph.

[0020] As a preferred scheme of the quantitative inversion method of the gas pipeline network leakage source based on satellite remote sensing, wherein: the gas anomaly area graph is geometrically characterized by a morphological analysis method to obtain gas diffusion characteristic parameters, and the specific steps are,

[0021] The gas anomaly area graph is subjected to morphological closing operation processing to generate a morphological anomaly graph;

[0022] The morphological anomaly graph is subjected to contour extraction and ellipse fitting by a boundary tracking algorithm and a least square ellipse fitting algorithm to form a gas anomaly geometric feature graph;

[0023] The gas anomaly geometric feature graph is subjected to plume parameter calculation by principal component analysis to obtain gas diffusion characteristic parameters.

[0024] As a preferred scheme of the quantitative inversion method of the gas pipeline network leakage source based on satellite remote sensing, wherein: the gas diffusion characteristic parameters are subjected to ground flow field inversion to output gas diffusion dominant parameters, and the specific steps are,

[0025] The gas diffusion characteristic parameters are subjected to diffusion physical relationship conversion to generate a wind direction and speed parameter set;

[0026] The wind direction and speed parameter set is subjected to two-dimensional flow field reconstruction by a reverse analysis method to form a flow field vector point set, and the flow field vector point set is subjected to spatial interpolation to generate gas diffusion flow field data;

[0027] The dominant wind direction and the average wind speed characteristic value are extracted from the gas diffusion flow field data, and the gas diffusion dominant parameters are generated by integration.

[0028] As a preferred scheme of the satellite remote sensing-based gas pipeline network leakage source quantitative inversion method, the gas diffusion dominant parameters are coupled with the leakage source probability distribution map to form a leakage source probability map, and the specific steps are as follows:

[0029] The dominant wind direction information in the gas diffusion dominant parameters is coupled with the leakage source probability distribution map in space to generate a wind direction coupling probability distribution map.

[0030] Based on the wind direction coupling probability distribution map, the wind direction influence weight value of each grid in the wind direction coupling probability distribution map is calculated by using a wind direction weight function, and the leakage source probability in the wind direction coupling probability distribution map is fused to form a leakage source probability map.

[0031] As a preferred scheme of the satellite remote sensing-based gas pipeline network leakage source quantitative inversion method, the gas diffusion dominant parameters are coupled with the leakage source probability distribution map to form a leakage source probability map, and the specific steps are as follows:

[0032] The region maximum value of the leakage source probability map is searched to identify the probability extreme point in the leakage source probability map, and a leakage source probability extreme point set is generated by integration.

[0033] Based on the leakage source probability extreme point set, the probability extreme point is converted into a geographic coordinate point by coordinate mapping to form a leakage source candidate position coordinate set.

[0034] The leakage source candidate position coordinate set is matched with the gas pipeline network vector topological data in space to determine the geographic coordinate point closest to the gas pipeline network as the gas pipeline network leakage source position.

[0035] As a preferred scheme of the satellite remote sensing-based gas pipeline network leakage source quantitative inversion method, the gas column concentration distribution data and the gas diffusion dominant parameters are inversed based on the gas pipeline network leakage source position to output the gas pipeline network leakage amount, and the specific steps are as follows:

[0036] Based on the gas pipeline network leakage source position, the gas column concentration value of the corresponding position in the gas column concentration distribution data is extracted, and the data is integrated with the gas diffusion dominant parameters to generate a leakage flux inversion data set.

[0037] The leakage flux inversion data set is calculated by the Gauss integral method to generate a preliminary leakage flux value.

[0038] The preliminary leakage flux value is corrected in time scale with the wind speed data in the gas diffusion dominant parameters to output the gas pipeline network leakage amount.

[0039] In a second aspect, the application provides a satellite remote sensing-based gas pipeline network leakage source quantitative inversion system, comprising,

[0040] A concentration distribution module is configured to collect satellite remote sensing spectral data and gas pipeline network vector topological data, perform radiation calibration and atmospheric correction processing on the satellite remote sensing spectral data through a differential absorption spectrum algorithm, and generate gas column concentration distribution data.

[0041] An anomaly identification module is configured to perform anomaly identification on the gas concentration of a satellite remote sensing observation area based on the gas column concentration distribution data, and output a gas anomaly area map.

[0042] A probability and statistics module is configured to perform spatial overlaying of the gas anomaly area map and the gas pipeline network vector topological data, output gas overlaying data, and perform weighted statistics on the gas overlaying data to output a leakage source probability distribution map.

[0043] A dominant parameter module is configured to perform geometric feature extraction on the gas anomaly area map through a morphological analysis method, obtain gas diffusion characteristic parameters, perform ground flow field inversion on the gas diffusion characteristic parameters, and output gas diffusion dominant parameters.

[0044] A leakage position module is configured to perform coupling analysis of the gas diffusion dominant parameters and the leakage source probability distribution map to form a leakage source probability map, and perform extreme value positioning on the leakage source probability map to determine the leakage source position of the gas pipeline network.

[0045] A flux inversion module is configured to perform leakage flux inversion on the gas column concentration distribution data and the gas diffusion dominant parameters based on the leakage source position of the gas pipeline network, and output the leakage amount of the gas pipeline network.

[0046] The application has the following beneficial effects: through the cooperation of the spatial overlaying and ground flow field inversion double mechanisms, the precise positioning of the gas pipeline network leakage source and the reliable inversion of the leakage flux are realized. Based on the spatial overlaying of the gas anomaly area map and the gas pipeline network vector topological data, the leakage source probability distribution map is generated, the leakage source search range is physically constrained in the pipeline network space, the geometric feature extraction is performed on the gas anomaly area map to obtain the gas diffusion characteristic parameters, the ground flow field inversion is performed to obtain the gas diffusion dominant parameters, the coupling analysis of the gas diffusion dominant parameters and the leakage source probability distribution map forms the leakage source probability map, and the leakage source positioning based on the double constraint optimization and the flux quantitative evaluation based on the actual diffusion conditions are realized. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0048] Fig. 1 Flow chart of a satellite remote sensing based gas pipeline network leakage source quantitative inversion method.

[0049] Fig. 2 Schematic diagram of a satellite remote sensing based gas pipeline network leakage source quantitative inversion system.

[0050] Fig. 3 Flow chart of outputting a leakage source probability distribution diagram.

[0051] Fig. 4 Flow chart of generating a gas diffusion dominant parameter. DETAILED DESCRIPTION

[0052] In order to make the above objectives, features and advantages of the present application more apparent and comprehensible, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0053] In the following description, a large number of specific details are set forth in order to facilitate a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the connotation of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0054] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0055] Reference Figs. 1-4 For one embodiment of the present application, the embodiment provides a satellite remote sensing based gas pipeline network leakage source quantitative inversion method, comprising the following steps:

[0056] S1, satellite remote sensing spectral data and gas pipeline network vector topological data are collected, satellite remote sensing spectral data is processed by differential absorption spectrum algorithm for radiation calibration and atmospheric correction, and gas column concentration distribution data is generated.

[0057] The satellite remote sensing spectral data is processed by radiation calibration and geographic geometry calibration to generate calibrated remote sensing spectral data;

[0058] Specifically, satellite remote sensing spectral data is collected by a hyperspectral remote sensing sensor mounted on a satellite, and a linear transformation is performed on the digital quantization values of the satellite remote sensing spectral data to form physical radiance values according to the sensor calibration parameters provided in the satellite remote sensing spectral data, so as to eliminate the influence of response differences of the hyperspectral remote sensing sensor; the radiometric calibration processing refers to a process of converting the digital quantization values in the satellite remote sensing spectral data into physical radiance values with physical meaning; in combination with orbit parameters and attitude information attached to the satellite remote sensing spectral data, the satellite remote sensing spectral data after the radiometric calibration is matched with geographical coordinates, ground control point correction and projection transformation are performed, image pixels in the satellite remote sensing spectral data are accurately mapped to latitude and longitude coordinates on the earth's surface, and calibrated remote sensing spectral data with accurate spatial position information are generated.

[0059] The calibrated remote sensing spectral data are subjected to atmospheric correction by a reflectance method, and corrected spectral data are output.

[0060] Specifically, the reflectance method is an atmospheric correction method for eliminating the influence of the atmosphere by estimating the ground reflectance, and the core is to derive the ground true reflectance by using known radiance and observation geometry information, that is, the solar zenith angle, the observation zenith angle and the relative azimuth angle corresponding to the image pixels in the calibrated remote sensing spectral data are used in combination with the physical radiance value of the ground uniform area to establish the physical relationship between the ground reflectance and the physical radiance value at the top of the atmosphere, to obtain the reflectance component affected by the atmosphere and eliminate it, to obtain the ground true reflectance, to restore the ground true reflectance to the physical radiance value without atmospheric interference according to the radiometric calibration parameters of the calibrated remote sensing spectral data, and to output the corrected spectral data.

[0061] Based on the corrected spectral data, the differential absorption spectrum algorithm is used to extract the absorption intensity of the gas at the characteristic waveband, and the absorption intensity is converted into vertical column concentration to generate gas column concentration distribution data.

[0062] Specifically, the differential absorption spectrum algorithm refers to analyzing the absorption characteristics of the gas in a specific wave band, using spectral differentiation method to separate and extract narrow-band absorption signals, that is, selecting the narrow wavelength interval with absorption characteristics of the gas in the infrared wave band from the corrected spectral data, using the differential absorption spectrum algorithm to perform spectral differentiation processing on the physical radiation brightness values of adjacent wave bands, separating the narrow-band absorption signals caused by gas absorption, identifying and extracting the absorption intensity of the gas in the characteristic wave band by comparing the absorption peak position with the characteristic absorption spectrum line in the standard spectrum library (the standard spectrum library is a database containing the absorption characteristics of known gas molecules at different wavelengths, which records the accurate absorption spectral line position and intensity of various gas components (such as methane) in the infrared wave band; the standard spectrum library is obtained through precise spectral experiments in a controlled environment, and has high accuracy and repeatability), according to the Bill-Lambert law relationship between absorption intensity and gas concentration, combining the solar zenith angle, observation zenith angle and atmospheric transmittance information corresponding to the corrected spectral data, correcting the absorption path and converting the absorption intensity into vertical column concentration value per unit area, to generate the spatial distribution form of the gas column concentration distribution data.

[0063] S2, based on the gas column concentration distribution data, identifying the abnormal gas concentration in the satellite remote sensing observation area, and outputting the gas abnormal area map.

[0064] Comparing the gas column concentration distribution data with the preset gas concentration threshold value to determine the abnormal gas point;

[0065] Specifically, the normal concentration level of the gas in the background atmosphere in the satellite remote sensing observation area is obtained, that is, by analyzing the long-term observation values of the unaffected area in the historical gas column concentration distribution data, the average value and fluctuation range of the vertical column concentration are calculated to determine the normal concentration level of the gas in the background atmosphere, and a gas concentration threshold value for identifying leakage is set according to the historical statistical analysis results and the detection sensitivity requirement (the value range of the gas concentration threshold value is set according to the concentration level and fluctuation range of the gas in the background atmosphere, and the average value of the background concentration is 2 to 3 times the standard deviation, so as to ensure that the leakage signal significantly higher than the background level can be effectively identified); the vertical column concentration value in the gas column concentration distribution data is compared with the gas concentration threshold value for each grid in the gas column concentration distribution data, and when the vertical column concentration value of the grid is greater than the gas concentration threshold value, the grid is marked as an abnormal gas point, and an abnormal gas point set composed of all the abnormal gas points is formed;

[0066] The spatially continuous abnormal gas points are aggregated and spliced by a spatial clustering algorithm, and the gas abnormal area map is output;

[0067] Specifically, for all grid points marked as abnormal gas points in the gas column concentration distribution data, spatial relationship analysis is performed according to the geographic coordinate positions, and the Euclidean distance between adjacent grid points is used as the connection judgment condition. The abnormal gas points with a Euclidean distance less than the set neighborhood range (the neighborhood range is determined according to the spatial resolution of the satellite remote sensing spectral data to determine the ground coverage size of a single grid, and combined with the spatial continuity characteristics presented by the gas diffusion in the atmosphere, the shape scale of the historical gas abnormal area is analyzed, and the typical distance between adjacent abnormal gas points is counted, and the maximum distance between the abnormal gas points that can be reasonably connected in the geographic coordinates is determined based on the typical distance, and the maximum distance is used as the neighborhood range for judging whether the points belong to the same abnormal area in the spatial clustering process) are gradually merged into the same set to form a point group with spatial connectivity; continue to iterate and aggregate until all adjacent abnormal gas points are merged into independent continuous regions; each connected region composed of multiple abnormal gas points is integrated into a spatial closed surface feature to generate a gas abnormal area map containing the entire abnormal distribution range.

[0068] S3, superimpose the gas abnormal area map and the gas pipe network vector topology data in space, output the gas superposition data, and perform weighted statistics on the gas superposition data to output the leakage source probability distribution map.

[0069] The gas abnormal area map and the gas pipe network vector topology data are spatially aligned and integrated to form a gas pipe network leakage data pair.

[0070] Specifically, the geographic coordinates of the gas abnormal area map are uniformly converted to the coordinate system used by the gas pipe network vector topology data through coordinate transformation method to realize spatial position matching alignment; after completing the spatial coordinate alignment, the intersection analysis of the closed surface features in the gas abnormal area map and the line features in the gas pipe network vector topology data is performed using the spatial superposition method to obtain the regions with overlapping or adjacent relationship in space; based on the spatial range of the gas abnormal area and the geographic distribution of the gas pipe network, each pair of abnormal areas and pipe network segments with spatial correlation is combined into a gas pipe network leakage data pair containing the corresponding relationship.

[0071] Attribute association and geometric matching are performed on the gas pipe network leakage data pair to output the gas superposition data.

[0072] Specifically, the gas pipeline network leakage data pair contains the closed surface element of the gas abnormal area graph and the line element of the gas pipeline network vector topological data. Based on the vertical column concentration maximum value, average concentration and area attribute of the gas abnormal area and the pipe diameter, material and buried year attribute of the gas pipeline network segment, the corresponding attribute information is added to the corresponding gas pipeline network leakage data pair to realize attribute association. The shortest distance between the centroid coordinates of the gas abnormal area and the gas pipeline network segment is counted, and it is judged whether the centroid is located within the buffer range of the pipeline network segment (the buffer range is a fixed width area around the pipeline network segment determined comprehensively according to the geographical accuracy of the gas pipeline network segment, the spatial resolution of the satellite remote sensing spectral data and the actual influence distance of gas diffusion). If it is located within the buffer range of the pipeline network segment, it is confirmed that the gas pipeline network leakage data pair has a spatial matching relationship. The attribute association result and the geometric matching result are integrated to integrate the gas pipeline network leakage data pair with complete attribute information and verified by geometric matching into gas overlay data.

[0073] The pipeline network attribute information and abnormal area information of the gas pipeline network are extracted from the gas overlay data, and the pipeline distance weight value of the satellite remote sensing observation area is calculated through the distance attenuation function.

[0074] Specifically, the gas overlay data contains the gas pipeline network segment and the corresponding gas abnormal area which have passed attribute association and geometric matching. The pipe diameter, material and buried year of each gas pipeline network segment are extracted as pipeline network attribute information through data query from the gas overlay data, and the vertical column concentration maximum value, average concentration and area of the corresponding gas abnormal area are extracted as abnormal area information. Based on the spatial position of the gas pipeline network segment, the Euclidean distance to the nearest gas pipeline network segment is calculated for each grid in the satellite remote sensing observation area. Using the distance attenuation function (the distance attenuation function refers to a mathematical function that decreases the influence degree according to the increase of spatial distance, which is used to quantify the relationship that the influence of the gas pipeline network on the surrounding area decreases with the increase of distance), according to the Euclidean distance between the grid and the gas pipeline network segment, the pipeline distance weight value of each grid is counted. The closer the Euclidean distance, the higher the weight, and the farther the Euclidean distance, the lower the weight, so as to generate the pipeline distance weight value distribution covering the entire observation area.

[0075] The formula for calculating the pipeline distance weight value is

[0076] ;

[0077] Among them, represents the actual distance from the center point of the grid to the nearest gas pipeline network segment, represents the pipeline distance weight value of the current grid, represents the distance attenuation coefficient (the distance attenuation coefficient is determined by experimental measurement, which reflects the degree of weakening of signal or physical quantity with the increase of distance), represents the power parameter, which determines the speed at which the pipe network distance weight decreases with the increase of distance.

[0078] The gas column concentration in the gas anomaly area map is multiplied by the pipe network distance weight value, and the leakage source probability is output. The leakage source probability is processed by a spatial interpolation method to output a leakage source probability distribution map.

[0079] Specifically, the grid of the gas anomaly area map is aligned in space with the pipe network distance weight value distribution of the satellite remote sensing observation area, ensuring that the gas column concentration in the gas anomaly area map and the pipe network distance weight value have the same geographic coordinate range and spatial resolution. For each overlapping grid, the corresponding gas column concentration value is multiplied by the pipe network distance weight value at the corresponding position to obtain a leakage source probability value reflecting the combined effect of concentration intensity and pipe network proximity. All leakage source probability values that have completed the multiplication operation are used as discrete point data, and a reverse distance weighted interpolation method is used to perform numerical interpolation based on the distance relationship of adjacent points to generate a continuous leakage source probability distribution map covering the entire satellite remote sensing observation area.

[0080] S4, geometric feature extraction is performed on the gas anomaly area map by a morphological analysis method to obtain gas diffusion characteristic parameters, and ground flow field inversion is performed on the gas diffusion characteristic parameters to output gas diffusion dominant parameters.

[0081] The gas anomaly area map is processed by morphological closing operation to generate a morphological anomaly map.

[0082] Specifically, the gas anomaly area map is processed by morphological closing operation through a morphological analysis method (the morphological analysis method refers to using mathematical morphological operations such as dilation and erosion to process the gas anomaly area map to change the shape features, thereby realizing the repair and analysis of the gas anomaly area map, and the gas anomaly area map is processed by dilation and then erosion to fill small holes, smooth edges and ensure the continuity and integrity of the abnormal area). The morphological closing operation processing refers to applying a dilation operation to each local minimum value in the gas anomaly area map. The dilation operation is to move a structural element in the gas anomaly area map at each position and compare it with the currently covered pixel value, select the maximum pixel value in the coverage range of the structural element to replace the center pixel value, expand the foreground object and fill small holes or gaps; a corrosion operation is performed on the dilated image. The corrosion operation is the inverse process of the dilation operation, which selects the minimum pixel value in the coverage range of the structural element to replace the center pixel value, thereby reducing the boundary of the foreground object to approach the original position. After the morphological closing operation processing, a morphological anomaly map without small holes and with smoother edges is generated, ensuring the continuity and integrity of the abnormal area.

[0083] The morphological abnormality graph is subjected to contour extraction and ellipse fitting through a boundary tracking algorithm and a least square ellipse fitting algorithm to form a gas abnormality geometric feature graph;

[0084] Specifically, a boundary tracking algorithm (the boundary tracking algorithm is to search and connect adjacent edge points pixel by pixel to obtain a closed contour of a target region in an image) is applied to the morphological abnormality graph to identify all the closed contours of the target regions, edges are detected by traversing each pixel in the morphological abnormality graph, and positions of all edge pixels are recorded to form a series of contour point sets; a least square ellipse fitting algorithm (the least square ellipse fitting algorithm is to determine parameters of a best fitting ellipse by minimizing a sum of squares of geometric distances from the contour point sets to the ellipse boundary) is applied to each contour point set to statistically obtain ellipse parameters of the contour point set, including a center position, a major axis, a minor axis and a rotation angle, to ensure that a sum of squares of errors between the contour and the ellipse is minimized; the contour extraction and the ellipse fitting are performed to generate a graph showing geometric features of the gas abnormality region, i.e., the gas abnormality geometric feature graph.

[0085] The gas diffusion characteristic parameters are calculated through principal component analysis on the gas abnormality geometric feature graph to obtain gas diffusion characteristic parameters;

[0086] Specifically, coordinate data of all the contour point sets in the gas abnormality geometric feature graph are statistically counted to form a coordinate data set, covariance statistics are performed on the coordinate data set, a covariance matrix is output, eigenvalues and eigenvectors of the covariance matrix are solved, and main variation directions of the gas abnormality geometric feature graph are determined according to the eigenvalues and the eigenvectors; the first two eigenvectors that explain the maximum variability are selected to represent main and secondary directions of gas diffusion; based on eigenvalues corresponding to the main and secondary directions (indicating diffusion degrees in the respective directions), a major axis length, a minor axis length and a direction angle of gas diffusion are statistically counted to form the plume parameters; and the plume parameters are integrated to form the gas diffusion characteristic parameters.

[0087] The gas diffusion characteristic parameters are subjected to diffusion physical relationship conversion to generate a wind direction and wind speed parameter set;

[0088] Specifically, the long axis direction angle in the gas diffusion characteristic parameter is taken as the main extension direction of the gas plume in space, which is mapped to the azimuth angle in the geographic coordinate system to determine the dominant direction of the gas diffusion driven by the airflow, i.e., to be inverted to the wind direction; the ratio of the long axis length to the short axis length is used to reflect the stretching degree in the diffusion process, and the geometric proportion of the transverse diffusion width to the downflow diffusion length is converted to the velocity index of the airflow motion in combination with the similarity relationship in the atmospheric turbulence diffusion theory (the atmospheric turbulence diffusion theory refers to the theory of studying the mixing and diffusion law of gas or matter in the atmosphere due to turbulent motion, the diffusion is caused by irregular motion in the atmosphere, involving the changes of wind speed, wind direction, and the influences of environmental factors such as temperature and humidity, and is used to explain and predict the propagation path and concentration distribution of pollutants or gas in the air); according to the empirical relationship between the diffusion shape and the wind speed, the average wind speed is estimated by the ratio of the long axis length to the diffusion duration, and the conversion results of the wind direction and the wind speed are combined into a complete set of wind direction and wind speed parameters.

[0089] The two-dimensional flow field is reconstructed by the reverse analysis method for the wind direction and wind speed parameter set to form a flow field vector point set, and the flow field vector point set is spatially interpolated to generate gas diffusion flow field data;

[0090] Specifically, the reverse analysis method is to convert each set of wind direction and wind speed values in the wind direction and wind speed parameter set into a planar vector with direction and size, take the geographic center position of the gas anomaly area as the initial reference point, construct the flow field vector points in the local two-dimensional plane, according to the spatial distribution range of gas diffusion, in combination with the physical delay relationship between satellite observation time and gas propagation, the wind direction and wind speed are reversely distributed to the grid positions at different distances and directions according to the diffusion path, a plurality of flow field vector points covering the abnormal area and the surrounding range are formed to constitute the flow field vector point set; all the vector points in the flow field vector point set are spatially interpolated, the bilinear interpolation or Kriging interpolation method is used to estimate the wind direction and wind speed values of the missing positions according to the direction and size relationship of the adjacent vector points, and the continuous, gridded gas diffusion flow field data covering the entire satellite remote sensing observation area is generated.

[0091] The dominant wind direction and average wind speed characteristic values are extracted from the gas diffusion flow field data to integrate and generate the gas diffusion dominant parameters;

[0092] Specifically, the wind direction and wind speed values of all the grids in the gas diffusion flow field data are read one by one, the direction interval with the highest frequency of the wind direction in the range of 0° to 360° is determined as the dominant wind direction, the arithmetic mean value of all the effective grid wind speed values in the gas diffusion flow field data is calculated to obtain the average wind speed characteristic value; the dominant wind direction and the average wind speed characteristic value are combined according to a unified data structure to form a set of comprehensive parameters representing the overall diffusion trend of the region, i.e., the gas diffusion dominant parameters.

[0093] S5, coupling and analyzing the gas diffusion dominant parameter and the leakage source probability distribution map to form a leakage source probability map, and positioning the extreme value of the leakage source probability map to determine the leakage source position of the gas pipe network.

[0094] The dominant wind direction information in the gas diffusion dominant parameter is spatially aligned and coupled with the leakage source probability distribution map to generate a wind direction coupling probability distribution map.

[0095] Specifically, the dominant wind direction information and the leakage source probability distribution map are subjected to unified geographic coordinate conversion to ensure that the dominant wind direction information and the leakage source probability distribution map are accurately aligned in the same geographic coordinate system. At each grid point, the probability value at the corresponding position in the leakage source probability distribution map is adjusted according to the dominant wind direction information, i.e., the probability cumulative influence of each grid point in the dominant wind direction is counted to reflect the influence degree of the wind direction on the leakage source diffusion. For each grid point, the leakage source probability of the grid point is accumulated according to the distance weight, wherein the closer the distance, the greater the weight, to realize the redistribution of the probability value. The adjusted probability values of all grid points are integrated to generate a wind direction coupling probability distribution map that comprehensively reflects the influence of the wind direction.

[0096] Based on the wind direction coupling probability distribution map, a wind direction weight function is used to calculate the wind direction influence weight value of each grid in the wind direction coupling probability distribution map, and a fusion operation is performed with the leakage source probability in the wind direction coupling probability distribution map to form a leakage source probability map.

[0097] Specifically, the wind direction weight function is a mathematical function constructed according to the dominant wind direction and the distance decay relationship, which is used to quantify the influence degree of different positions on the current grid point. The wind direction weight function considers the distance weight within the influence range of the dominant wind direction to ensure that the closer the distance, the greater the influence of the grid point on the current grid point. For each grid point, all grid points within a specified distance along the wind direction are searched according to the dominant wind direction information, and the contribution degree of the grid points to the selected grid point is counted, which is adjusted in proportion according to the distance. The wind direction influence weight value obtained by each grid point is fused with the leakage source probability in the wind direction coupling probability distribution map to output the leakage source probability. By integrating the adjusted leakage source probabilities of all grid points, a leakage source probability map is formed, which comprehensively reflects the combined influence of the wind direction and the initial probability distribution of the leakage source.

[0098] The leakage source probability map is subjected to regional maximum value search to identify the probability extreme points in the leakage source probability map, and a set of leakage source probability extreme points is integrated and generated.

[0099] Specifically, a fixed-size search window is defined based on the resolution of the leakage source probability map. The search window refers to a square or rectangular area with a specific size (such as 5x5, 10x10, etc.). By sliding the search window, the maximum value of the region in the leakage source probability map is searched to cover every possible location. For each location of the search window, the average value of the leakage source probability of all grid points within the search window is calculated, and the average value is compared with the leakage source probability value of the grid point at the center of the window. If the leakage source probability value of the grid point at the center of the window is not lower than the average value, the grid point is marked as a potential probability extreme point. The search process is repeated until the search window traverses the entire leakage source probability map. All marked potential probability extreme points are integrated to generate a set of leakage source probability extreme points. The set of leakage source probability extreme points comprehensively reflects the local area with the highest probability in the leakage source probability map.

[0100] Based on the set of probabilistic extreme points of the leakage source, the probabilistic extreme points are converted into geographic coordinate points through coordinate mapping to form a set of coordinates for candidate locations of the leakage source.

[0101] Specifically, based on the set of probabilistic extreme points of the leakage source, the row and column positions of each probabilistic extreme point in the leakage source probability map are read. Combined with the geographic coordinate range, spatial resolution, and projection information of the leakage source probability map, the position of each probabilistic extreme point is converted from the row and column index on the image to the latitude and longitude or Cartesian coordinates on the actual Earth surface using the correspondence between image coordinates and geographic coordinates. The conversion process is based on the geometric positioning parameters of the satellite remote sensing observation area to ensure the accuracy of the position mapping. All converted geographic coordinate points are summarized to form a set of multiple geographic coordinate points, namely the coordinate set of candidate leakage source locations.

[0102] Spatial matching is performed between the candidate location coordinate set of the leak source and the vector topology data of the gas pipeline network to determine the geographical coordinate point closest to the gas pipeline network as the location of the gas pipeline network leak source.

[0103] Specifically, for each geographic coordinate point in the candidate location coordinate set of the leak source, the gas pipeline network segments within the surrounding range are searched in the gas pipeline network vector topology data; for each candidate location coordinate point of the leak source, the shortest Euclidean distance from the candidate location coordinate point to each adjacent gas pipeline network segment is calculated, and the minimum distance value and the corresponding gas pipeline network segment are recorded; the minimum distance between all candidate location coordinate points of the leak source and the gas pipeline network segments are compared, and the geographic coordinate point closest to the gas pipeline network vector topology data is selected, and the geographic coordinate point is determined as the location of the gas pipeline network leak source.

[0104] S6. Based on the location of the gas pipeline leakage source, perform leakage flux inversion on the gas column concentration distribution data and the dominant gas diffusion parameters, and output the gas pipeline leakage amount.

[0105] Based on the gas pipe network leakage source position, the gas column concentration value of the corresponding position in the gas column concentration distribution data is extracted, and the data is integrated with the gas diffusion dominant parameters to generate a leakage flux inversion data set;

[0106] Specifically, based on the gas pipe network leakage source position, the gas column concentration value corresponding to each gas pipe network leakage source position is extracted from the gas column concentration distribution data, that is, through accurate spatial position matching, the geographic coordinates of the gas pipe network leakage source are matched with the coordinates in the gas column concentration distribution data to obtain the gas column concentration value at the corresponding geographic coordinate point, ensuring that the extraction process is based on accurate spatial position matching to obtain accurate concentration information of each leakage source position; the gas column concentration value is integrated with the dominant wind direction and average wind speed characteristic value in the gas diffusion dominant parameters, the correlation between the gas column concentration value and the gas diffusion dominant parameters (such as the dominant wind direction and the average wind speed) is calculated, the relationship between the gas column concentration value and the diffusion parameters is established, and a comprehensive data set containing position, gas column concentration and gas diffusion characteristics, i.e. a leakage flux inversion data set, is formed.

[0107] The leakage flux inversion data set is calculated by the Gauss integral method to generate a preliminary leakage flux value;

[0108] Specifically, according to the data integration results of the gas column concentration distribution data and the gas diffusion dominant parameters (including the dominant wind direction and the average wind speed), the gas column concentration value and the diffusion characteristics corresponding to each leakage source position are determined; the Gauss integral method is used to integrate and operate the concentration distribution of each leakage point. The Gauss integral method is a numerical integration method, which selects specific sampling points (called Gauss points) in the integral interval and gives corresponding weights in the form of weighted sum to approximate the integral value, which can realize high-precision integral result with less calculation amount; that is, the concentration distribution of each leakage point is integrated and operated, the integral operation considers the concentration contribution of different positions in space, and combines the wind direction and wind speed and other influencing factors to adjust the integral range and weight, and outputs the preliminary leakage flux value. The results of all integrations are summarized to form a comprehensive leakage flux value list, which reflects the relative leakage intensity and distribution characteristics of each leakage point.

[0109] The preliminary leakage flux value is time-scaled with the wind speed data in the gas diffusion dominant parameters to output the gas pipe network leakage amount;

[0110] Specifically, according to the preliminary leakage flux value of each leakage point and the corresponding measurement time point, a time sequence of the leakage flux changing with time is established; for the leakage flux value at each time point, the leakage flux value is adjusted using the corresponding wind speed data, and the leakage flux value is adjusted by statistically analyzing the distance and concentration distribution of gas diffusion under different wind speeds to reflect the actual diffusion of gas under different wind speed conditions; all the leakage flux values after wind speed correction are summarized to form a comprehensive gas pipeline network leakage list, which reflects the relative leakage intensity and distribution characteristics of each leakage point after considering the influence of wind speed, that is, the output gas pipeline network leakage.

[0111] The embodiment also provides a satellite remote sensing-based gas pipeline network leakage source quantitative inversion system, comprising:

[0112] A concentration distribution module is configured to collect satellite remote sensing spectral data and gas pipeline network vector topological data, perform radiation calibration and atmospheric correction processing on the satellite remote sensing spectral data through a differential absorption spectrum algorithm, and generate gas column concentration distribution data.

[0113] An anomaly recognition module is configured to perform anomaly recognition on the gas concentration of a satellite remote sensing observation area based on the gas column concentration distribution data, and output a gas anomaly area map.

[0114] A probability and statistics module is configured to perform spatial overlaying on the gas anomaly area map and the gas pipeline network vector topological data, output gas overlaying data, and perform weighted statistics on the gas overlaying data, and output a leakage source probability distribution map.

[0115] A dominant parameter module is configured to perform geometric feature extraction on the gas anomaly area map through a morphological analysis method, obtain gas diffusion characteristic parameters, perform ground flow field inversion on the gas diffusion characteristic parameters, and output gas diffusion dominant parameters.

[0116] A leakage position module is configured to perform coupling analysis on the gas diffusion dominant parameters and the leakage source probability distribution map, form a leakage source probability map, perform extreme value positioning on the leakage source probability map, and determine a gas pipeline network leakage source position.

[0117] A flux inversion module is configured to perform leakage flux inversion on the gas column concentration distribution data and the gas diffusion dominant parameters based on the gas pipeline network leakage source position, and output a gas pipeline network leakage.

[0118] In summary, the present application realizes the accurate positioning of the gas pipeline network leakage source and the reliable inversion of the leakage flux through the double mechanism of spatial superposition and ground flow field inversion. Based on the spatial superposition of the gas abnormal area graph and the gas pipeline network vector topological data, a leakage source probability distribution graph is generated, the leakage source search range is physically constrained in the pipeline network space, the geometric feature extraction is performed on the gas abnormal area graph to obtain the gas diffusion characteristic parameters, the gas diffusion dominant parameters are obtained through the ground flow field inversion, the coupling analysis of the gas diffusion dominant parameters and the leakage source probability distribution graph forms a leakage source probability graph, and the leakage source positioning based on the double constraint optimization and the flux quantitative evaluation under the actual diffusion condition are realized.

[0119] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the present application, which should be covered in the scope of the claims of the present application.

Claims

1. A satellite remote sensing-based gas pipeline network leakage source quantitative inversion method, characterized in that: The application relates to a method for detecting a gas pipeline leakage source based on satellite remote sensing. Satellite remote sensing spectral data and gas pipeline network vector topological data are collected, satellite remote sensing spectral data is processed through a differential absorption spectrum algorithm to realize radiation calibration and atmospheric correction, and gas column concentration distribution data are generated; Based on the gas column concentration distribution data, the gas concentration of a satellite remote sensing observation area is abnormally identified, and a gas abnormal area graph is output; The gas abnormal area graph and the gas pipeline network vector topological data are spatially overlaid, gas overlay data are output, and the gas overlay data are statistically weighted to output a leakage source probability distribution graph; The gas abnormal area graph is geometrically characterized through a morphological analysis method, gas diffusion characteristic parameters are obtained, the gas diffusion characteristic parameters are inversed on a ground flow field, and gas diffusion dominant parameters are output; The gas diffusion dominant parameters and the leakage source probability distribution graph are coupled and analyzed to form a leakage source probability graph, and the leakage source probability graph is positioned at an extreme value to determine a gas pipeline network leakage source position; Based on the gas pipeline network leakage source position, the gas column concentration distribution data and the gas diffusion dominant parameters are inversed on a leakage flux to output a gas pipeline network leakage amount. The method comprises the following steps: The gas abnormal area graph and the gas pipeline network vector topological data are spatially aligned and integrated to form a gas pipeline network leakage data pair; The gas pipeline network leakage data pair is subjected to attribute association and geometric matching to output gas overlay data; The pipeline network attribute information and the abnormal area information of the gas pipeline network are extracted from the gas overlay data, and a pipeline distance weight value of a satellite remote sensing observation area is calculated through a distance attenuation function; The gas column concentration in the gas abnormal area graph is multiplied by the pipeline distance weight value to output a leakage source probability, and the leakage source probability is subjected to grid processing through a spatial interpolation method to output a leakage source probability distribution graph. The method comprises the following steps: The gas diffusion characteristic parameters are subjected to diffusion physical relationship conversion to generate a wind direction and speed parameter set; The wind direction and speed parameter set is subjected to two-dimensional flow field reconstruction through a reverse analysis method to form a flow field vector point set, and the flow field vector point set is subjected to spatial interpolation to generate gas diffusion flow field data; Dominant wind direction and average wind speed characteristic values are extracted from the gas diffusion flow field data, and the gas diffusion dominant parameters are integrated and generated.

2. The satellite remote sensing based gas pipeline network leakage source quantitative inversion method according to claim 1, characterized in that: The method comprises the following steps: The satellite remote sensing spectral data are subjected to radiation calibration processing and geographical geometric calibration to generate calibrated remote sensing spectral data; The calibrated remote sensing spectral data are subjected to atmospheric correction through a reflectivity method to output corrected spectral data; Based on the corrected spectral data, the absorption intensity of gas at a characteristic waveband is extracted through a differential absorption spectrum algorithm, and the absorption intensity is subjected to vertical column concentration conversion to generate gas column concentration distribution data.

3. The satellite remote sensing based gas pipeline network leak source quantification inversion method according to claim 2, characterized in that: The abnormal gas point is determined by comparing the gas column concentration distribution data with the preset gas concentration threshold value. The abnormal gas points are aggregated and spliced by a spatial clustering algorithm to output the gas anomaly area graph. The geometric feature of the gas anomaly area graph is extracted by a morphological analysis method to obtain the gas diffusion characteristic parameter.

4. The satellite remote sensing based gas pipeline network leakage source quantitative inversion method according to claim 3, characterized in that: The gas anomaly area graph is processed by a morphological closing operation to generate a morphological anomaly graph. The contour of the morphological anomaly graph is extracted and fitted by an ellipse fitting algorithm to form a gas anomaly geometric feature graph. The plume parameter is calculated by principal component analysis to obtain the gas diffusion characteristic parameter. The gas diffusion dominant parameter and the leakage source probability distribution graph are coupled and analyzed to form a leakage source probability graph.

5. The satellite remote sensing based gas pipeline network leak source quantification inversion method according to claim 4, characterized in that: The dominant wind direction information in the gas diffusion dominant parameter is spatially aligned and coupled with the leakage source probability distribution graph to generate a wind direction coupling probability distribution graph. Based on the wind direction coupling probability distribution graph, the wind direction influence weight value of each grid in the wind direction coupling probability distribution graph is calculated by using a wind direction weight function, and the wind direction coupling probability distribution graph is fused and operated with the leakage source probability to form a leakage source probability graph. The maximum value of the leakage source probability graph is located to determine the gas pipeline network leakage source position.

6. The satellite remote sensing based gas pipeline network leak source quantification inversion method of claim 5, wherein: The probability extreme points in the leakage source probability graph are identified by searching for the regional maximum value of the leakage source probability graph, and a leakage source probability extreme point set is generated. Based on the leakage source probability extreme point set, the probability extreme points are converted into geographic coordinate points by coordinate mapping to form a leakage source candidate position coordinate set. The leakage source candidate position coordinate set is spatially matched with the gas pipeline network vector topological data to determine the geographic coordinate point closest to the gas pipeline network as the gas pipeline network leakage source position. Based on the gas pipeline network leakage source position, the gas column concentration value at the corresponding position in the gas column concentration distribution data is extracted and integrated with the gas diffusion dominant parameter to generate a leakage flux inversion data set.

7. The satellite remote sensing based gas pipeline network leak source quantification inversion method according to claim 6, characterized in that: The leakage flux inversion data set is calculated by the Gauss integral method to generate a preliminary leakage flux value. The preliminary leakage flux value is time-scaled with the wind speed data in the gas diffusion dominant parameter to output the gas pipeline network leakage amount. The concentration distribution module is configured to collect satellite remote sensing spectral data and gas pipeline network vector topological data, perform radiation calibration and atmospheric correction on the satellite remote sensing spectral data by a differential absorption spectrum algorithm, and generate gas column concentration distribution data. The abnormal identification module is configured to identify the abnormal gas concentration in the satellite remote sensing observation area based on the gas column concentration distribution data, and output a gas anomaly area graph.

8. A satellite remote sensing based gas pipeline network leakage source quantitative inversion system based on any one of the satellite remote sensing based gas pipeline network leakage source quantitative inversion methods of claims 1-7, characterized in that: ​ ​ ​ a probability and statistics module, configured to perform spatial superposition of the gas anomaly area map and gas pipe network vector topological data, output gas superposition data, and perform weighted statistics on the gas superposition data to output a leakage source probability distribution map; a dominant parameter module, configured to perform geometric feature extraction on the gas anomaly area map by a morphological analysis method, obtain a gas diffusion characteristic parameter, and perform ground flow field inversion on the gas diffusion characteristic parameter to output a gas diffusion dominant parameter; a leakage position module, configured to perform coupling analysis on the gas diffusion dominant parameter and the leakage source probability distribution map to form a leakage source probability map, and perform extreme value positioning on the leakage source probability map to determine a gas pipe network leakage source position; a flux inversion module, configured to perform leakage flux inversion on gas column concentration distribution data and the gas diffusion dominant parameter based on the gas pipe network leakage source position to output a gas pipe network leakage amount.

Images