A method for identifying high-consequence areas in gas pipelines driven by multi-source data and knowledge

Abstract

The present invention discloses a method for identifying high-consequence areas of gas pipelines driven by multi-source data and knowledge, which collects vector line data with coordinate information of pipelines to be identified; identifies all areas with building clusters in the buffer zone of the pipeline to be identified and uses drones to photograph the building cluster areas; pre-processes the multi-view images obtained by drones; establishes a spatial information database; constructs a building attribute information database; constructs a space-attribute integrated database; quantitatively calculates the grade of the area around the pipeline and determines the grade of the high-consequence area. The present invention combines data from different sources, respectively utilizes their advantages and application methods, improves the accuracy of high-consequence area identification; constructs a space-attribute integrated database to improve the efficiency of management and subsequent calculations; establishes a quantitative calculation model for regional grades to reduce subjective errors in human judgment; and uses a rule knowledge base to realize automated high-consequence area grade calculation.

Description

Technical Field

[0001] The invention relates to a method for identifying high-consequence areas of a gas pipeline driven by multi-source data and knowledge, and belongs to the field of gas pipeline integrity management. Background Art

[0002] The identification of high consequence areas of long-distance gas pipelines is to find out the areas where pipeline leakage will seriously endanger public safety and (or) cause great environmental damage by analyzing the basic data of the pipeline, and identify the potential risks in the consequence areas. With the rapid development of my country's economy, the number of buildings around long-distance gas pipelines is increasing. The identification of high consequence areas has become an indispensable part of the national and corporate risk prevention and control measures: in 2015, the state issued GB 32167-2015 "Oil and Gas Pipeline Integrity Management Specifications", which clearly stated that the identification of high consequence areas of oil and gas pipelines is the basis for subsequent pipeline risk assessment and integrity evaluation, and clearly stipulated that the identification cycle of high consequence areas cannot exceed 18 months. Based on the identification results of high consequence areas, pipeline managers can timely formulate and adjust risk assessment, integrity evaluation plans or other control measures, which is conducive to the key management of high consequence area sections. At the same time, it can also reduce resource investment on the basis of effective prevention of dangers, and ultimately form a set of efficient and low-cost pipeline management solutions.

[0003] At present, manual field survey is the main method for identifying high consequence areas, but this method has high labor costs and low efficiency. Remote sensing technology has the characteristics of long distance and large range. With the development of remote sensing technology and the progress of data processing, remote sensing data can meet the requirements of high consequence area identification in long-distance gas pipelines. Among them, satellite remote sensing data has a wide coverage, but the accuracy of high consequence area identification will be reduced due to the large weather images and low resolution; drone images have high resolution and low-altitude flight is not affected by clouds, but drone batteries have low capacity, and their application in long-distance and long-length pipelines will be greatly limited.

[0004] At the same time, in order to completely identify the high-consequence areas of long-distance gas pipelines, it is also necessary to obtain the property information of the building such as the storey height and type. However, for areas that cannot be reached by personnel, the reliability of the results of manual investigation is low. Therefore, how to efficiently and accurately obtain the building attributes is also a major problem in the identification of high-consequence areas. Therefore, it is very necessary to explore a new method to identify the high-consequence areas of long-distance gas pipelines.

[0005] At present, research on the field of high-consequence areas of long-distance gas pipelines mainly focuses on the use of single remote sensing images and data for identification. For example, the "Intelligent Identification Method for High-Consequence Areas of Gas Pipelines Based on Satellite Image Recognition" with application number 202210789232.4 uses satellite images, but due to the low resolution of satellite remote sensing images, the method of extracting building outlines is limited and the accuracy is not high; and due to the influence of clouds, it is difficult and costly to obtain remote sensing images that are timely and have clear identification areas and suitable resolution; although most of the attribute information of buildings can be obtained remotely, important information such as the number of floors and the number of households still needs to be obtained after field investigation by dispatched staff, and the building space and attribute information are not integrated, and data management is relatively troublesome; the regional level and high-consequence area level cannot be calculated quantitatively, and still require manual qualitative judgment, which is highly subjective and has a low degree of automation. There are few studies on the use of drones to identify high-consequence areas, and the degree of automation and intelligence in acquiring building attributes is low. For example, "A UAV system and method for identifying high-consequence areas in pipelines" with application number 202110056426.9, this study uses images transmitted in real time by drones to identify buildings around the pipeline, and proposes a method for calculating the straight-line distance from the building to the pipeline based on drone-mounted equipment. However, it does not process and utilize the high resolution and other characteristics of drone images. In addition, the use of drones to fly the entire gas pipeline places high demands on the drone's endurance and the flight skills of personnel. The acquisition of building attributes is done by directly shouting and asking through the drone, which is inefficient and difficult to implement. After obtaining the attributes, manual processing and identification are still required, and the automation and accuracy are not high. Therefore, we combine the advantages of UAV and satellite remote sensing data, and use multi-source data to conveniently and accurately obtain building-related attributes, and establish a space-attribute integrated database; automatically identify the knowledge entities in the relevant consequence areas and convert them into logical expressions, thereby constructing a rule knowledge base. Combining the above two, we can automatically identify the high-consequence areas of gas pipelines and quantitatively calculate the regional level, which can greatly improve the efficiency and accuracy of identifying high-consequence areas of long-distance gas pipelines. Summary of the invention

[0006] In order to overcome the problems in the prior art, the present invention provides a method for identifying high-consequence areas of gas pipelines driven by multi-source data and knowledge. The method uses remote sensing images covering the entire long-distance gas pipeline to identify the target area that needs to be finely identified in the surrounding area, and then uses a drone to fly over the target area to obtain high-resolution multi-view images of the drone with coordinate information. The orthophoto of the target area is obtained by preprocessing all drone images, joint adjustment, aerial triangulation, dense matching, image matching and other steps. Taking advantage of its high resolution, the edge detection based on image texture features and connectivity analysis are used to extract the outline of the building, and a spatial information database is established. A three-dimensional model of the target area is established by three-dimensional reconstruction and combined with the crawled Amap data to obtain the attributes of the building in the area, such as the floor height, name, type, and number of households, collect and summarize the data to form a building attribute database, extract the relevant data in the above database, and automatically complete the conversion process based on predefined relevant rules to form a space-attribute integrated database; a rule knowledge base is constructed based on the relevant consequence area identification specification, and combined with the integrated database, the grade of the area around the pipeline and the grade of the high consequence area are calculated to accurately identify the high consequence area of ​​the long-distance gas pipeline.

[0007] The present invention provides a technical solution to solve the above technical problems: a method for identifying high-consequence areas of a gas pipeline driven by multi-source data and knowledge, comprising the following steps:

[0008] S1. Collect vector line data with coordinate information of the pipeline to be identified;

[0009] S2, preprocessing the vector line data;

[0010] S3, identifying all areas where building clusters exist in the buffer zone of the pipeline to be identified and photographing the building cluster areas using a drone;

[0011] S4, obtaining orthophotos of the building cluster area by preprocessing, joint adjustment, aerial triangulation, dense matching, and image matching the multi-view images obtained by the drone;

[0012] S5. Use edge detection method based on image texture features to extract the edge of the orthophoto and remove non-building edges through connectivity analysis to obtain building outlines and establish a spatial information database;

[0013] S6. Perform three-dimensional reconstruction of the UAV flight area through dense matching, TIN construction, and texture mapping, and obtain relevant attributes of buildings within the buffer zone in combination with Amap data, and then construct a database of building attribute information after sorting;

[0014] S7, constructing a space-attribute integrated database based on predefined relevant rules and a space information database and a building attribute information database;

[0015] S8. Construct a rule knowledge base based on the quantitative standards for regional grades and high-consequence area grades in relevant specifications, and calculate the regional grades and high-consequence area grades around the pipeline in combination with the spatial-attribute integrated database.

[0016] A further technical solution is that the vector line data includes relevant basic data on the length, operating pressure, and pipe diameter of the pipeline to be identified, satellite remote sensing image data with cloud cover <5% in the area where the pipeline to be identified is located, and building point data with relevant attribute information such as location, type, and name around the pipeline to be identified.

[0017] A further technical solution is that the specific process of step S2 is:

[0018] S21. Calculate the potential impact radius of the pipeline to be identified based on relevant basic data;

[0019] S22. Establish pipeline buffer zones based on potential impact radius as the primary area for identifying high consequence areas;

[0020] S23, clipping the satellite remote sensing image data and building point data according to the buffer size.

[0021] A further technical solution is that the calculation formula in step S21 is:

[0022]

[0023] Where: d is the outer diameter of the pipeline, in millimeters; p is the maximum allowable operating pressure of the pipe section, in MPa; r is the radius of the affected area, in meters.

[0024] A further technical solution is that the specific process of step S4 is:

[0025] S41, obtain high-precision UAV flight attitude POS data by performing joint adjustment between the RTK positioning unit carried by the UAV and the ground RTK positioning unit;

[0026] S42, performing aerial triangulation in combination with the adjusted POS data to obtain the geodetic coordinates of the unknown points and the image exterior orientation elements;

[0027] S43, densely matching the aerial triangulation results, matching a large number of three-dimensional discrete points to establish a digital surface model DSM of the target area, and then obtaining a digital elevation model DEM through automatic filtering processing;

[0028] S44. Obtain an orthophoto of the building cluster area based on the digital elevation model DEM, the digital surface model DSM of the target area and image matching.

[0029] A further technical solution is that the specific process of step S5 is:

[0030] S51, converting the color orthophoto into a grayscale image through standard floating point conversion;

[0031] S52, preprocessing the orthophoto using Gaussian smoothing filter;

[0032] S53, calculating the gradient and direction of the pixel by using the Canny operator and filtering by using a preset threshold to obtain an edge detection result, wherein the result includes both building and non-building outlines;

[0033] S54, establishing a connected area constraint condition to eliminate non-building contours and obtain an extraction result;

[0034] S55. Output the extraction result in the form of vector data and input it into the established spatial information database.

[0035] A further technical solution is that the specific process of step S6 is:

[0036] S61, based on the densely matched three-dimensional point cloud, establish an irregular triangulated network TIN, and use photos to perform texture mapping on it to obtain a three-dimensional model with real texture;

[0037] S62. Obtain the building structure, floor area, number of floors, and number of window openings on the facade through the three-dimensional model;

[0038] S63, using the building point data to obtain the type and location of the building;

[0039] S64. Collect and organize the above attributes, establish a building attribute information database, and automatically calculate and count the number of households in the residential building.

[0040] A further technical solution is that the specific process of step S7 is: extracting data from the spatial information database and the building attribute information database respectively according to the fields; and converting them, processing the extracted data according to predetermined rules, including merging or splitting fields, sorting, filling missing values, and merging with spatial data.

[0041] A further technical solution is that the specific process of step S8 is:

[0042] S81. Extract relevant rule entities from multiple specifications by using keyword information extraction methods and combining constraints; eliminate entity ambiguity by clustering methods and then summarize to form a rule data set;

[0043] S82, importing the rule data set into the rule engine, automatically processing and converting it into a computer-recognizable logical expression to form a rule knowledge base;

[0044] S83. Combine the rule knowledge base and the spatial-attribute integrated database to calculate the level of the area around the pipeline. The first and second level areas are calculated according to the following formula;

[0045]

[0046] Where: n is the number of buildings in a high consequence area; , is the quantitative standard extracted from the rule knowledge base; M is the first and second level quantitative evaluation index, M>0 is the first level area, and M<0 is the second level area;

[0047]

[0048] Where: is the number of floors of a building in a high-consequence area in the space-attribute integrated database; The building storey standard extracted from the rule knowledge base; is the maximum number of floors of buildings in the high consequence area; is the minimum number of floors of buildings in the high consequence area; S is the first and second level quantitative assessment index, S<0 is the third level area, and the rest are the fourth level areas;

[0049] S84. Based on the above calculation results, the level of the high consequence area is determined in combination with the rule knowledge base.

[0050] The present invention has the following beneficial effects:

[0051] Compared with the traditional method of field identification of high-consequence areas of long-distance gas pipelines, the method of the present invention combines satellite remote sensing data and drone image data to extract building outlines, uses building point data, building three-dimensional models and other data to obtain building-related attributes, and constructs spatial information databases and building attribute databases respectively, and integrates them into a spatial-attribute integrated database, which is convenient for enterprise management while saving a lot of manpower and material resources. A rule knowledge base is constructed based on the consequence area identification specification, and the integrated database is combined to realize quantitative calculation of the area grade around the gas pipeline and the high-consequence area grade, which reduces human subjective errors and improves the degree of intelligence. Compared with other high-consequence area identification technologies, it combines data from different sources and uses their advantages and application methods to improve the accuracy of high-consequence area identification; constructs a spatial-attribute integrated database to improve the efficiency of management and subsequent calculations; and uses the rule knowledge base to realize quantitative and automated calculation of area grades and high-consequence area grades.

[0052] The method provided by the present invention utilizes the wide coverage of satellite images to accurately identify building cluster areas, reduces the flight area of ​​drones, and solves the problem of difficulty in application in high-consequence area identification due to its weak endurance; orthophotos constructed based on drone images have the advantages of high resolution, large scale, high currentness, etc. compared to satellite images, which can greatly improve the timeliness and accuracy of the extracted building outlines. Using the three-dimensional model of the building and the AutoNavi map point data, the two complement each other and can intuitively and clearly obtain the building's floor height, type, number of households and other attributes, without the need for external personnel to conduct on-site investigations, thereby improving recognition efficiency and reducing personnel and time costs. A spatial attribute database and rule knowledge base constructed through multi-source data and related knowledge are used to construct a quantitative regional grade calculation model, which reduces errors caused by human subjective judgment and improves the accuracy of high-consequence area grade determination. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a flowchart of the present invention;

[0054] Figure 2 It is an orthophoto map of the implementation target area;

[0055] Figure 3 It is a three-dimensional model diagram of the implementation target area. DETAILED DESCRIPTION

[0056] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0057] like Figure 1 As shown, a method for identifying high-consequence areas of a gas pipeline driven by multi-source data and knowledge of the present invention comprises the following steps:

[0058] S1. Collect vector line data with coordinate information of the pipeline to be identified;

[0059] The vector line data includes: relevant basic data such as the length, operating pressure, and diameter of the pipeline to be identified; satellite remote sensing images of low cloud cover (cloud cover <5%) in the area where the pipeline to be identified is located; and building point data with relevant attribute information such as location, type, and name around the pipeline to be identified;

[0060] The vector data of the pipeline to be identified in the example is collected. The starting point is Anju Gas Distribution Station (30°18'47.29"N, 105°28'31.45"E) and the end point is Xiaoyangjiawan Gas Distribution Station (30°18'35.15"N, 105°25'55.25"E). The pipeline passes through towns, villages and mountainous areas along the way.

[0061] The Sentinel-2 optical image with a resolution of 10 meters and a width of 290 km released by the European Space Agency is used. The image time selected in the example is the image of December 2022, which can completely cover the study area without any cloud cover above.

[0062] The length of the pipeline in the example obtained from the pipeline ledger is 5.31 km, the diameter is 219 mm, and the operating pressure is 4 MPa.

[0063] After obtaining the personal key from the Amap open API platform, use Python to crawl the building point data of types such as life services, commercial residences, public facilities, and medical care on the Amap to form an EXCEL table, and then use ArcGIS software and combine the latitude and longitude coordinates to convert the EXCEL into point data.

[0064] S2. In order to reduce the amount of data and improve recognition accuracy, the above images are imported and preprocessed.

[0065] The potential impact radius of the pipeline in the example is calculated by using the pipeline basic data to be 43.362 meters.

[0066] Import the pipeline vector data in the example into ArcGIS software and input 43.362 meters as the linear unit to establish the buffer zone.

[0067] Using the Clip tool in ArcGIS software, the satellite image is clipped to raster data with the buffer range as the boundary; the building point data and its attributes within the range are retained, and the type field is displayed using the annotation function.

[0068] S3. Identify the building cluster area in the buffer zone of the pipeline to be identified through satellite images and use a drone to photograph the building cluster area.

[0069] According to the actual situation of the area, the UAV flight parameters are set, the intelligent flight function is used to reduce the possibility of accidents caused by flight, and the UAV flight attitude POS data is obtained through the RTK high-precision positioning unit. The flight altitude of the area in the example is selected as 120 meters, the lateral overlap is set to 60%, and the heading overlap is 80%, and the UAV image with a resolution of 0.1m is obtained.

[0070] S4, the orthophoto of the building cluster area is obtained by preprocessing, joint adjustment, aerial triangulation, dense matching, and image matching of the multi-view images obtained by the drone, such as Figure 2 shown.

[0071] Export the POS data from the aircraft's RTK high-precision positioning unit and the ground RTK positioning unit and organize them into txt files, and delete the attitude data and photos taken before the drone takes off.

[0072] The sorted POS data are jointly adjusted and imported into 3D Smart software together with photos for aerial triangulation.

[0073] Through dense matching of aerial triangulation results, a large number of three-dimensional discrete points are matched to establish the digital surface model DSM of the target area, and then the digital elevation model DEM is obtained through automatic filtering.

[0074] Based on the digital elevation model DEM, the digital surface model DSM of the target area and image matching, the orthophoto of the entire area is obtained.

[0075] S5. Taking advantage of the high resolution of orthophotos, edge detection methods based on image texture features were used to extract edges and non-building edges were removed through connectivity analysis to obtain building outlines and establish a spatial information database. Finally, a total of 4 high-consequence areas around the gas pipeline were obtained.

[0076] The color orthophoto is converted to its grayscale image through standard floating point conversion.

[0077] Use Gaussian filtering to smooth the grayscale image. The larger the Gaussian filter convolution kernel, the larger the scale of noise filtering by the Gaussian filter. Since the resolution and texture of each image are different, the size of the Gaussian filter convolution kernel also needs to be adjusted accordingly. In the example, through multiple experiments, a Gaussian filter composed of 5×5 convolution kernels is selected to smooth the image, and the noise of the smoothing result is significantly reduced.

[0078] The Canny algorithm uses dual thresholds to determine the edge. Pixels with gradients less than the low threshold are considered non-edge points, and those with gradients greater than the high threshold are considered strong edge points. Those in the middle are considered weak edge points, and those connected to strong edge points are considered edge points. In this example, the low threshold is set to 50, the high threshold is set to 120, and the output result shows that the building edge is complete. On this basis, the image is binarized to highlight the edge.

[0079] In order to eliminate edge contours outside buildings, the length threshold is set to 60 pixels, the number of straight lines is set to 3, and small edges and non-building contours are filtered out.

[0080] The extraction results are output in vector form and input into the spatial information database.

[0081] S6. Based on the densely matched point cloud, TIN construction and texture mapping are performed to perform 3D reconstruction of the UAV flight area and combine the Amap data to obtain the relevant attributes of the buildings within the buffer zone. After sorting, a database of building attribute information is constructed. The 3D model is as follows: Figure 3 shown.

[0082] Based on the densely matched three-dimensional point cloud, an irregular triangulated network TIN is established, and texture mapping is performed on it using photos to obtain a three-dimensional model with real texture.

[0083] In the example, the 3D Smart software is used to visualize the 3D model of the building, and the building's floor area and volume can be directly measured. The types and names of different buildings can be obtained by directly observing the 3D model. By enlarging the 3D model for investigation, the number of floors, number of windows on the facade, and structure of the building can be obtained. The fields such as building type, name, number of floors, number of windows on the facade, etc. are sorted out to establish a building attribute database, and based on these data, the number of households inside the building is automatically calculated as the data basis for subsequent work.

[0084] S7. Extract relevant data from the spatial and attribute databases, automatically complete the conversion process based on predefined relevant rules, and form a spatial-attribute integrated database.

[0085] In the example, data in the spatial and attribute databases are matched and extracted according to the FID field, and then converted, which mainly involves processing the extracted data according to predetermined rules, including operations such as merging or splitting fields, sorting, filling missing values, and merging with spatial data.

[0086] S8. Based on the quantitative standards for regional grades and high-consequence area grades in the "Oil and Gas Pipeline Integrity Management Code" (GB32167-2015) and "Operational Procedures for Identifying the Consequence Impact Area of ​​Gas Pipeline Accidents" (RQ-ZY-0207), a rule knowledge base is constructed, and combined with the integrated database, the grade of the area around the pipeline and the grade of the high-consequence area are calculated;

[0087] S81. Extract relevant rule entities from multiple specifications by using keyword information extraction methods and combining constraints; eliminate entity ambiguity by clustering methods and then summarize to form a rule data set;

[0088] S82, importing the rule data set into the rule engine, automatically processing and converting it into a computer-recognizable logical expression to form a rule knowledge base;

[0089] S83. Combine the rule knowledge base and the spatial-attribute integrated database to calculate the level of the area around the pipeline. The first and second level areas are calculated according to the following formula;

[0090]

[0091] Where: n is the number of buildings in a high consequence area; , is the quantitative standard extracted from the rule knowledge base; M is the first and second level quantitative evaluation index, M>0 is the first level area, and M<0 is the second level area;

[0092]

[0093] Where: is the number of floors of a building in a high-consequence area in the space-attribute integrated database; The building storey standard extracted from the rule knowledge base; is the maximum number of floors of buildings in the high consequence area; is the minimum number of floors of buildings in the high consequence area; S is the first and second level quantitative assessment index, S<0 is the third level area, and the rest are the fourth level areas;

[0094] S84. Based on the above calculation results, the level of the high consequence area is determined in combination with the rule knowledge base.

[0095] The keyword information extraction method is used in combination with constraint conditions to extract relevant rule entities from multiple specifications; the entity ambiguity is eliminated through clustering methods and then summarized to form a rule data set. The rule data set is imported into the rule engine, automatically processed and converted into a computer-recognizable logical expression to form a rule knowledge base. Combined with the rule knowledge base and the spatial-attribute integrated database, the grade of the area around the pipeline and the grade of the high-consequence area are calculated.

[0096] The above description is not intended to impose any form of limitation on the present invention. Although the present invention has been disclosed through the above embodiments, it is not intended to limit the present invention. Any technician familiar with the profession can make some changes or modifications to equivalent embodiments of equivalent changes using the technical contents disclosed above without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the scope of the technical solution of the present invention.

Claims

1. A method for identifying high-consequence areas in gas pipelines driven by multi-source data and knowledge, characterized in that: The following steps are involved: S1. Collect vector line data with coordinate information of the pipeline to be identified; S2, preprocessing the vector line data; S3, identifying all areas where building clusters exist in the buffer zone of the pipeline to be identified and photographing the building cluster areas using a drone; S4, obtaining orthophotos of the building cluster area by preprocessing, joint adjustment, aerial triangulation, dense matching, and image matching the multi-view images obtained by the drone; S41, obtain high-precision UAV flight attitude POS data by performing joint adjustment between the RTK positioning unit carried by the UAV and the ground RTK positioning unit; S42, performing aerial triangulation in combination with the adjusted POS data to obtain the geodetic coordinates of the unknown points and the image exterior orientation elements; S43, densely matching the aerial triangulation results, matching a large number of three-dimensional discrete points to establish a digital surface model DSM of the target area, and then obtaining a digital elevation model DEM through automatic filtering processing; S44, obtaining an orthophoto of the building cluster area based on the digital elevation model DEM, the digital surface model DSM of the target area and image matching; S5. Use edge detection method based on image texture features to extract the edge of the orthophoto and remove non-building edges through connectivity analysis to obtain building outlines and establish a spatial information database; S6. Perform three-dimensional reconstruction of the UAV flight area through dense matching, TIN construction, and texture mapping, and obtain relevant attributes of buildings within the buffer zone in combination with Amap data, and then construct a database of building attribute information after sorting; S61, based on the densely matched three-dimensional point cloud, establish an irregular triangulated network TIN, and use photos to perform texture mapping on it to obtain a three-dimensional model with real texture; S62. Obtain the building structure, floor area, number of floors, and number of window openings on the facade through the three-dimensional model; S63, using the building point data to obtain the type and location of the building; S64, collecting and arranging the above attributes, establishing a database of building attribute information, and automatically calculating and computing the number of households in the residential building; S7, constructing a space-attribute integrated database based on predefined relevant rules and a space information database and a building attribute information database; S8. Construct a rule knowledge base based on the quantitative standards for regional grades and high-consequence area grades in relevant specifications, and calculate the regional grades and high-consequence area grades around the pipeline in combination with the spatial-attribute integrated database.

2. According to claim 1, a method for identifying high-consequence areas of gas pipelines driven by multi-source data and knowledge is characterized in that: The vector line data includes relevant basic data on the length, operating pressure, and diameter of the pipeline to be identified, satellite remote sensing image data with cloud cover <5% in the area where the pipeline to be identified is located, and building point data with relevant attribute information such as location, type, and name around the pipeline to be identified.

3. According to claim 2, a method for identifying high-consequence areas of gas pipelines driven by multi-source data and knowledge is characterized in that: The specific process of step S2 is: S21. Calculate the potential impact radius of the pipeline to be identified based on relevant basic data; S22. Establish pipeline buffer zones based on potential impact radius as the primary area for identifying high consequence areas; S23, clipping the satellite remote sensing image data and building point data according to the buffer size.

4. According to claim 3, a method for identifying high-consequence areas of gas pipelines driven by multi-source data and knowledge is characterized in that: The calculation formula in step S21 is: Where: d is the outer diameter of the pipeline, in millimeters; p is the maximum allowable operating pressure of the pipe section, in MPa; r is the radius of the affected area, in meters.

5. According to claim 3, a method for identifying high-consequence areas of gas pipelines driven by multi-source data and knowledge is characterized in that: The specific process of step S5 is as follows: S51, converting the color orthophoto into a grayscale image through standard floating point conversion; S52, preprocessing the orthophoto using Gaussian smoothing filter; S53, calculating the gradient and direction of the pixel by using the Canny operator and filtering by using a preset threshold to obtain an edge detection result, wherein the result includes both building and non-building outlines; S54, establishing a connected area constraint condition to eliminate non-building contours and obtain an extraction result; S55. Output the extraction result in the form of vector data and input it into the established spatial information database.

6. The method for identifying high-consequence areas of gas pipelines driven by multi-source data and knowledge according to claim 3 is characterized in that: The specific process of step S7 is: extracting data from the spatial information database and the building attribute information database respectively according to the fields; and converting them, processing the extracted data according to predetermined rules, including merging or splitting fields, sorting, filling missing values, and merging with spatial data.

7. The method for identifying high-consequence areas of gas pipelines driven by multi-source data and knowledge according to claim 3 is characterized in that: The specific process of step S8 is as follows: S81. Extract relevant rule entities from multiple specifications by using keyword information extraction methods and combining constraints; eliminate entity ambiguity by clustering methods and then summarize to form a rule data set; S82, importing the rule data set into the rule engine, automatically processing and converting it into a computer-recognizable logical expression to form a rule knowledge base; S83. Combine the rule knowledge base and the spatial-attribute integrated database to calculate the level of the area around the pipeline. The first and second level areas are calculated according to the following formula; M=(n-n s1 )(n-n s2 ) Where: n is the number of buildings in a high consequence area; n s1 、n s2 It is the quantitative standard extracted from the rule knowledge base; M is a first- and second-level quantitative evaluation index, M>0 is a first-level region, and M<0 is a second-level region; Where: x i is the number of floors of a building in a high-consequence area in the space-attribute integrated database; standard The building storey standard extracted from the rule knowledge base; S is a first- and second-level quantitative assessment index, S<0 is a third-level region, and the rest are fourth-level regions; S84. Based on the above calculation results, the level of the high consequence area is determined in combination with the rule knowledge base.

Images