A smart gas pipeline partition safety supervision method and an internet of things system

Abstract

The present application provides a kind of wisdom gas pipeline partition safety supervision method and internet of things system, the method includes based on the distribution data of gas pipeline in target area, the target area is divided, determine multiple subareas;Obtain the gas sensing data of each site of multiple subareas on gas pipeline;Based on the gas sensing data, pipeline data, determine the actual leakage point on the gas pipeline;Based on the gas sensing data, historical maintenance record, determine the potential hidden danger point on the gas pipeline.The internet of things system includes successively interactive government regulation management platform, government regulation sensing network platform, government regulation object platform, gas company sensing network platform and gas equipment object platform.The present application can realize timely processing gas leakage problem, reduce the security risk due to gas leakage, while it can ensure that potential hidden danger point can be repaired in time, and be monitored sufficiently, to effectively prevent the occurrence of gas leakage.

Description

Technical Field

[0001] This manual relates to the field of gas pipeline leak monitoring, and in particular to a smart gas pipeline zoned safety monitoring method and Internet of Things system. Background Technology

[0002] Most gas pipelines are buried underground. When a leak occurs, it can directly or indirectly generate a large amount of flammable gas, which can spread to surrounding rainwater and sewage pipe networks and other municipal manholes. Because rainwater and sewage pipes are interconnected, the gas can spread for several kilometers along the pipeline. When the concentration reaches the explosive limit, it will cause a large-scale chain explosion upon encountering an ignition source. Therefore, it is necessary to monitor gas pipelines.

[0003] To address the issue of how to regulate gas pipelines, CN110107815B proposes a method and device for detecting gas pipeline leaks, which detects the concentration of combustible gases in underground spaces surrounding multiple gas pipelines and determines whether the concentration meets the leakage conditions. However, the detection of combustible gas concentrations often occurs only when the gas pipeline has been damaged, resulting in a leak that has reached a certain level before being detected, exhibiting a lag. Furthermore, it fails to detect potential hazard points and cannot address them in a timely manner.

[0004] Therefore, it is desirable to provide an improved intelligent gas pipeline zoning safety monitoring method and system that can predict the actual locations of leaks and potential locations of leaks, so as to facilitate timely early warning and maintenance. Summary of the Invention

[0005] This invention includes a smart gas pipeline zoning safety supervision method. The method includes: dividing the target area into multiple sub-areas based on gas pipeline distribution data; acquiring gas sensor data for each point on the gas pipeline within the target area from multiple gas company management platforms via a government gas supervision sensor network platform; determining actual leak points on the gas pipeline based on the gas sensor data and pipeline data; determining potential hazard points on the gas pipeline based on the gas sensor data and historical maintenance records; responding to the existence of an actual leak point: generating a first maintenance instruction and a first adjustment instruction based on the actual location of the actual leak point on the gas pipeline; and sending the first maintenance instruction to the gas company management platform corresponding to the actual location to instruct... The gas company management platform is instructed to dispatch maintenance personnel to repair the actual location; and to send the first adjustment instruction to the gas pipeline control equipment corresponding to the actual location to adjust the operating parameters of the gas pipeline control equipment; in response to the existence of the potential hazard point: based on the potential location of the potential hazard point on the gas pipeline, a second inspection instruction and a second adjustment instruction are generated; the second inspection instruction is sent to the gas company management platform corresponding to the potential location to instruct the gas company management platform to dispatch maintenance personnel to inspect the potential location; and the second adjustment instruction is sent to the gas pipeline monitoring device corresponding to the potential location to adjust the monitoring parameters of the gas pipeline monitoring device.

[0006] This invention includes an intelligent gas pipeline zoning safety supervision IoT system. The system comprises a government regulatory management platform, a government regulatory sensor network platform, a government regulated object platform, a gas company sensor network platform, and a gas equipment object platform that interact sequentially. The government regulatory management platform includes a government gas regulatory management platform and a government safety regulatory management platform; the government regulatory sensor network platform includes a government gas regulatory sensor network platform and a government safety regulatory sensor network platform. The intelligent gas pipeline zoning safety supervision IoT system is configured to perform the following operations: based on gas pipeline distribution data in a target area, divide the target area into multiple sub-areas; through the government gas regulatory sensor network platform, obtain gas sensor data for each point on the gas pipeline in the multiple sub-areas within the target area from multiple gas company management platforms; based on the gas sensor data and pipeline data, determine the actual leak points on the gas pipeline; based on the gas sensor data and historical maintenance records, determine the potential hidden danger points on the gas pipeline. In response to the existence of the actual leak point: based on the actual location of the actual leak point on the gas pipeline, a first maintenance command and a first adjustment command are generated; the first maintenance command is sent to the gas company management platform corresponding to the actual location to instruct the gas company management platform to dispatch maintenance personnel to maintain the actual location; and the first adjustment command is sent to the gas pipeline control equipment corresponding to the actual location to adjust the operating parameters of the gas pipeline control equipment. In response to the existence of the potential hazard point: based on the potential location of the potential hazard point on the gas pipeline, a second maintenance command and a second adjustment command are generated; the second maintenance command is sent to the gas company management platform corresponding to the potential location to instruct the gas company management platform to dispatch maintenance personnel to inspect the potential location; and the second adjustment command is sent to the gas pipeline monitoring device corresponding to the potential location to adjust the monitoring parameters of the gas pipeline monitoring device. Attached Figure Description

[0007] Figure 1 This is a schematic diagram of the structure of an intelligent gas pipeline zone safety monitoring IoT system according to some embodiments of this specification;

[0008] Figure 2 This is an exemplary flowchart of a smart gas pipeline zoned safety monitoring method according to some embodiments of this specification;

[0009] Figure 3 This is an exemplary schematic diagram illustrating the determination of sub-regions and management mapping relationships according to some embodiments of this specification;

[0010] Figure 4This is a structural schematic diagram illustrating the determination of the actual leak point according to some embodiments of this specification;

[0011] Figure 5 This is a schematic diagram of the structure of a potential hazard prediction model according to some embodiments of this application specification. Detailed Implementation

[0012] Typically, one or more detection devices are installed near gas pipelines to detect leaks. For example, sensors can be embedded to detect gas concentrations. However, these detection devices have a certain time lag, often only detecting leaks after they have occurred. Furthermore, the time required from detecting a leak to having personnel arrive on-site for repairs increases safety risks.

[0013] In view of this, in some embodiments of this specification, it is desirable to provide an improved intelligent gas pipeline zoning safety monitoring method and Internet of Things system, which can identify actual leak points and potential hazard points on gas pipelines, take timely measures when gas pipeline leaks, and take corresponding maintenance measures for potential hazard points to prevent leaks from occurring at potential hazard points as well.

[0014] Figure 1 This is a schematic diagram of the structure of an intelligent gas pipeline zone safety monitoring IoT system, as shown in some embodiments of this specification. Figure 1 As shown, the smart gas pipeline zone safety supervision IoT system includes a government supervision and management platform 110, a government supervision sensor network platform 120, a government supervision object platform 130, a gas company sensor network platform 140, and a gas equipment object platform 150 that interact sequentially.

[0015] The government regulatory management platform 110 is configured by the government to coordinate and facilitate communication and collaboration among various functional platforms, and to aggregate all information on gas pipelines, providing a platform for perception management and control management functions for the operation and maintenance of gas pipelines.

[0016] In some embodiments, the government regulatory management platform 110 may include a government gas regulatory management platform and a government safety regulatory management platform, which can respectively analyze and process gas-related data and gas pipeline safety-related data. In some embodiments, the government regulatory management platform 110 can interact with the government regulatory sensor network platform 120.

[0017] The government regulatory sensor network platform 120 is configured as a functional platform for the government to manage sensor communications. In some embodiments, the government regulatory sensor network platform 120 can realize the functions of sensing and communication of perception information and control information. In some embodiments, the government regulatory sensor network platform is configured to interact with the government gas regulatory management platform and the gas company management platform. In some embodiments, the government regulatory sensor network platform 120 may include a government gas regulatory sensor network platform and a government safety regulatory sensor network platform, which can be used to acquire the operation and status information of gas pipeline-related equipment and gas pipeline safety-related equipment, respectively.

[0018] In some embodiments, the government regulatory sensor network platform 120 can engage in bidirectional information exchange with the government regulatory management platform 110 and the government-regulated object platform 130. For example, the government regulatory management platform 110 can send a data acquisition instruction to the gas company management platform via the government regulatory sensor network platform 120. After receiving the data acquisition instruction, the gas company management platform can upload gas sensor data to the government regulatory management platform 110 via the government regulatory sensor network platform 120.

[0019] The government regulatory platform 130 refers to a platform for generating government regulatory information and controlling the execution of regulatory information. In some embodiments, the government regulatory platform 130 may include a gas company management platform.

[0020] In some embodiments, the gas company management platform can receive gas sensor data uploaded by the gas company sensor network platform 140, and upload the gas sensor data to the government regulatory sensor network platform 120 upon receiving a data acquisition instruction from the government regulatory sensor network platform 120; and receive first and / or second maintenance instructions issued by the government regulatory management platform 110 via the government regulatory sensor network platform 120, generate dispatch tasks based on the first and / or second maintenance instructions, and issue the dispatch tasks to maintenance personnel to dispatch maintenance personnel to perform maintenance on the target sub-area. Further explanation regarding gas sensor data, maintenance instructions, and dispatch tasks can be found in the relevant content below. In some embodiments, each gas company has its corresponding gas company management platform.

[0021] The gas company sensor network platform 140 refers to the functional platform for the gas company to comprehensively manage sensor communication.

[0022] In some embodiments, the gas company's sensor network platform 140 can engage in bidirectional information exchange with the gas equipment object platform 150 and the gas company's management platform. For example, the gas company's sensor network platform 140 can upload gas sensor data obtained from the gas equipment object platform 150 to the gas company's management platform.

[0023] The gas equipment object platform 150 refers to a functional platform for generating sensing information and executing control information. In some embodiments, the gas equipment object platform 150 may include a gas pipeline monitoring device. The gas equipment object platform 150 can interact with the gas company's sensor network platform 140.

[0024] A gas pipeline monitoring device is used to monitor information related to gas pipelines. The device includes at least temperature sensors, humidity sensors, flow rate sensors, and pressure sensors. For example, the gas pipeline monitoring device can be deployed inside the gas pipeline in the area to be monitored to monitor gas sensor data within the pipeline and upload the obtained data to the gas company's sensor network platform 140 via the gas equipment object platform 150. For more information about the aforementioned platform, please refer to the following... Figures 2 to 4 And its related descriptions.

[0025] In some embodiments, the intelligent gas pipeline zoning safety monitoring system may also include multi-level networks, such as a primary network and a secondary network. For example, the primary network may include an intelligent gas primary network management platform, an intelligent gas primary network sensor network platform, and an intelligent gas primary network object platform. Similarly, the secondary network may include an intelligent gas secondary network management platform, an intelligent gas secondary network sensor network platform, and a gas secondary network object platform.

[0026] In some embodiments of this specification, the smart gas company management platform can acquire multi-dimensional data from other platforms, improving the accuracy of pipeline assessment for sub-regions and enabling the pipeline corridor to make faster and more accurate adjustments to sub-regions in emergency situations. Furthermore, this IoT system integrates multiple platforms, including the gas company management platform, the gas equipment object platform 150, and the government regulatory management platform 110, and can form a closed-loop information operation between these platforms. Effective collaborative operation between platforms helps improve the informatization and intelligence of smart gas pipeline zoning safety supervision. In summary, through intelligent data management and collaborative decision-making, this system can improve the supervision capabilities, safety, and management efficiency of pipeline zones.

[0027] Figure 2 This is an exemplary flowchart of a smart gas pipeline zoning safety monitoring method according to some embodiments of this specification. Figure 2 As shown, process 200 includes the following steps. In some embodiments, process 200 may be executed by a government gas regulatory management platform.

[0028] Step 210: Based on the gas pipeline distribution data in the target area, the target area is divided into multiple sub-areas.

[0029] The target area refers to the area where gas pipelines are distributed and require zoned supervision. For example, a district of a city may be designated as the target area. Gas pipeline distribution data refers to data related to the distribution of gas pipelines. In some embodiments, gas pipeline distribution data may include the density of gas pipelines, the number of gas pipelines, their length, diameter, gas flow direction, and specific routes of the gas pipelines.

[0030] In some embodiments, the government gas regulatory management platform can obtain relevant records from the database of the government regulatory management platform 110 when gas pipelines are built in the target area, and determine the content related to the distribution of gas pipelines as the required gas pipeline distribution data.

[0031] A sub-region refers to multiple regions obtained by dividing a target area. In some embodiments, the government gas regulatory management platform can determine multiple sub-regions in various ways. For example, the government gas regulatory management platform can perform gridding on the map of the target area, dividing the map of the target area into multiple grid regions based on the grid size and grid shape, and determining the resulting grid regions as the desired sub-regions. The grid size and grid shape can be preset based on prior experience.

[0032] In some embodiments, the government gas regulatory management platform can also divide gas pipelines into multiple sub-regions based on methods for constructing target pipeline maps and determining fitness levels. For more information on this part, please refer to... Figure 3 And related explanations.

[0033] Step 220: Obtain gas sensor data for each point in multiple sub-areas of the gas pipeline within the target area from multiple gas company management platforms through the government's gas regulatory sensor network platform.

[0034] Gas sensing data refers to acquired monitoring data related to gas. In some embodiments, gas sensing data may include gas temperature data, gas humidity, gas flow rate data, gas flow volume data, and pipeline pressure data, etc.

[0035] In some embodiments, gas monitoring data can be obtained from gas sensor data at various points in multiple sub-regions along a gas pipeline through pipeline monitoring devices. A location refers to the specific location on the gas pipeline in a sub-region where gas sensor data needs to be collected. In some embodiments, the government's gas regulatory management platform can define locations on the gas pipeline where pipeline monitoring devices are installed as locations.

[0036] Step 230: Based on gas sensor data and pipeline data, determine the actual leak point on the gas pipeline.

[0037] Pipeline data refers to monitoring data related to gas pipelines obtained from gas pipelines in a target area. In some embodiments, pipeline data may include information such as the pipeline's material, model, length, diameter, and service life.

[0038] In some embodiments, the government gas regulatory management platform can obtain log records from the database during the installation and construction of gas pipelines and extract the corresponding pipeline data from them.

[0039] The actual leak point refers to the specific location where a gas leak actually occurs on the gas pipeline in the target area at the current time.

[0040] In some embodiments, the government gas regulatory management platform can construct a vector database based on historical data and determine the corresponding actual leak point based on matching vector retrieval. In some embodiments, the government gas regulatory management platform can construct a target feature vector based on gas sensor data and pipeline data. There are various ways to construct the target feature vector. For example, it can be constructed using methods such as One-Hot and Word2Vec.

[0041] The vector database can include multiple reference vectors and corresponding reference leak points. Each reference vector can be constructed based on historical gas sensor data and historical pipeline data. The construction method of the reference vector is similar to that of the target feature vector. The reference leak points can be constructed based on historical actual leak points in the historical gas leak events corresponding to the reference vectors.

[0042] In some embodiments, the government gas regulatory management platform can determine the actual leak point based on the similarity between the target feature vector and multiple reference vectors in a vector database. For example, a reference vector whose similarity to the target feature vector meets a first preset condition can be used as the target vector, and the reference leak point corresponding to the target vector can be used as the final actual leak point. The first preset condition can be set according to the situation. For example, it could be the maximum similarity, or a similarity greater than a threshold. Another example is the minimum vector distance.

[0043] In some embodiments, the government gas regulatory management platform can directly identify a specific point on the pipeline as the actual leak point. In some embodiments, the gas company's management platform can also independently identify a specific point on the pipeline as the actual leak point and upload the information to the government gas regulatory management platform. In some embodiments, the gas company's management platform can also work with the government gas regulatory management platform to jointly determine whether a specific point on the pipeline is the actual leak point.

[0044] In some embodiments, the government gas regulatory management platform can also determine the actual location of leaks and the corresponding leakage results for each leak point based on identification features, current gas sensor data, and historical gas sensor data. For details on this section, please refer to [link to relevant documentation]. Figure 4 And related explanations.

[0045] Step 240: Based on gas sensor data and historical maintenance records, identify potential hazard points on the gas pipeline.

[0046] Historical maintenance records refer to the information recorded during the inspection, repair, and maintenance of gas pipelines over a historical period. In some embodiments, historical maintenance records may include at least the historical maintenance time, the historical maintenance location, and the cause of the historical failure.

[0047] In some embodiments, the government gas regulatory management platform can obtain historical records of gas pipeline maintenance from the database and extract the corresponding historical maintenance records.

[0048] Potential hazard points refer to specific locations on gas pipelines in the target area where there is a risk of gas leakage.

[0049] In some embodiments, the government gas regulatory management platform can construct a vector database based on historical data and identify corresponding potential hazard points based on matching vector retrieval. In some embodiments, the government gas regulatory management platform can construct a target feature vector based on gas sensor data and historical maintenance records. There are various ways to construct the target feature vector. For example, it can be constructed using methods such as One-Hot and Word2Vec.

[0050] The vector database can include multiple reference vectors and corresponding reference potential hazard points. Each reference vector can be constructed based on historical gas sensor data and historical maintenance records. The construction method of the reference vector is similar to that of the target feature vector. Reference potential hazard points can be obtained based on historical leak points corresponding to the reference vector before any leaks occurred.

[0051] In some embodiments, the government's gas regulatory management platform can determine potential hazard points based on the similarity between a target feature vector and multiple reference vectors in a vector database. For example, reference vectors whose similarity to the target feature vector meets preset conditions can be used as target vectors, and the reference hazard points corresponding to the target vectors can be used as the final potential hazard points. The preset conditions can be set according to the situation. For example, maximum similarity, or similarity greater than a threshold, etc. Another example is minimum vector distance.

[0052] In some embodiments, the government's gas regulatory management platform can also determine updated potential hazard points based on a hazard point prediction model. For details on this section, please refer to [link to relevant documentation]. Figure 5 And related explanations.

[0053] Step 250, in response to the existence of a leak: Based on the actual location of the leak on the gas pipeline, generate a first maintenance command and a first adjustment command; send the first maintenance command to the gas company management platform corresponding to the actual location to instruct the gas company management platform to dispatch maintenance personnel to maintain the actual location; and send the first adjustment command to the gas pipeline control equipment corresponding to the actual location to adjust the operating parameters of the gas pipeline control equipment.

[0054] The first maintenance instruction is an operational instruction used to instruct the gas company to repair any actual leaks on the gas pipeline.

[0055] In some embodiments, the government gas regulatory management platform can establish a first preset table based on the actual location and severity of historical leaks. The first preset table includes the correspondence between historical leak locations, historical leak severity, and different historical first maintenance orders. The government gas regulatory management platform can consult the first preset table to determine the current first maintenance order based on the current actual location and severity of the leak. The leak severity reflects the severity of the gas leak and is positively correlated with the gas leak rate.

[0056] The first adjustment command is an instruction used to instruct the gas company to adjust the gas pipeline control equipment within a gas pipeline where a leak has occurred. Operating parameters are the control parameters for the operation of the gas pipeline control equipment (e.g., gas pressure valves, gas flow rate valves, etc.). In some embodiments, operating parameters may include at least gas pressure.

[0057] In some embodiments, the government gas regulatory management platform can establish a second preset table based on historical actual leakage locations and historical operating parameters. The second preset table includes the correspondence between historical actual leakage locations, historical operating parameters, and different historical first adjustment instructions. The government gas regulatory management platform can determine the current first adjustment instruction by consulting the second preset table, based on the current actual leakage location and operating parameters.

[0058] In some embodiments, after the government gas regulatory management platform generates a first maintenance instruction and a first adjustment instruction, it can send the first maintenance instruction to the gas company management platform corresponding to the actual location. The gas company management platform, in accordance with the relevant requirements of the first maintenance instruction, dispatches a predetermined number of maintenance personnel with corresponding maintenance tools to the actual location and performs maintenance according to the required maintenance operation process. The gas equipment object platform 150 can receive the first adjustment instruction and adjust the gas pipeline control equipment corresponding to the actual location to reduce the pressure of the pipeline where the actual leak point is located and alleviate the gas leak.

[0059] Step 260, in response to the existence of potential hazard points: Based on the potential location of potential hazard points on the gas pipeline, generate a second maintenance instruction and a second adjustment instruction; send the second maintenance instruction to the gas company management platform corresponding to the potential location to instruct the gas company management platform to dispatch maintenance personnel to inspect the potential location; and send the second adjustment instruction to the gas pipeline monitoring device corresponding to the potential location to adjust the monitoring parameters of the gas pipeline monitoring device.

[0060] The second maintenance instruction is an operational instruction used to instruct the gas company to inspect and maintain potential hazards on gas pipelines. In some embodiments, the second maintenance instruction may include the location of the pipeline to be inspected, the required maintenance tools, the specific maintenance operations, and the number of maintenance personnel. The composition of the second maintenance instruction is similar to that of the first maintenance instruction; please refer to the relevant content of the first maintenance instruction above.

[0061] In some embodiments, the government gas regulatory management platform can establish a third preset table based on historical potential locations and historical operating parameters. The third preset table includes the correspondence between historical potential locations, historical operating parameters, and different historical second maintenance instructions. The government gas regulatory management platform can determine the current second maintenance instruction by consulting the third preset table, based on the potential location and operating parameters of the current actual leak point.

[0062] The second adjustment instruction is to regulate the gas pipeline monitoring device that has potential hidden dangers within the gas pipeline.

[0063] In some embodiments, the second adjustment instruction may include the specific location of the gas pipeline monitoring device to be adjusted, the monitoring parameters to be adjusted, and the adjustment range of the monitoring parameters. In some embodiments, the monitoring parameters may include at least the gas monitoring frequency and the monitoring accuracy. The composition of the second adjustment instruction is similar to that of the first adjustment instruction, and can be found in the relevant content of the first adjustment instruction above.

[0064] In some embodiments, the government gas regulatory management platform can establish a fourth preset table based on historical potential locations and historical monitoring parameters. The fourth preset table includes the correspondence between historical potential locations, historical monitoring parameters, and different historical second adjustment instructions. The government gas regulatory management platform can consult the fourth preset table to determine the current second adjustment instruction based on the potential location and monitoring parameters of the current actual leak point.

[0065] In some embodiments, after the government gas regulatory management platform generates a second maintenance instruction and a second adjustment instruction, it can issue the second maintenance instruction to the gas company management platform corresponding to the actual location. The gas company management platform, in accordance with the relevant requirements of the second maintenance instruction, dispatches a predetermined number of maintenance personnel with corresponding maintenance tools to the potential location and performs risk assessment and maintenance according to the required maintenance operation process. The gas equipment object platform 150 can receive the second adjustment instruction and adjust the gas pipeline monitoring device corresponding to the actual location, thereby monitoring the pipeline location with potential leakage risk more frequently.

[0066] In some embodiments of this specification, by identifying the actual leak point and potential hazard points, dispatching personnel to repair the actual leak point and reducing the gas pressure in the pipeline at the leak point can reduce gas leakage in a timely manner and lower the safety risks caused by gas leakage. Dispatching personnel to inspect potential hazard points and increasing the monitoring frequency of relevant monitoring devices can ensure that potential hazard points are repaired in a timely manner and fully monitored, thereby preventing gas leakage from occurring.

[0067] Figure 3 This is an exemplary schematic diagram illustrating the determination of sub-regions and management mapping relationships according to some embodiments of this specification. Steps 310 to 360 (not shown in the figure) can be executed by the government gas regulatory management platform.

[0068] Step 310: Based on the location data 311 and pipeline number data 312 of the gas pipelines in the target area, construct the target pipeline map 321.

[0069] Location data 311 refers to information related to the geographical location of the gas pipeline distribution. In some embodiments, location data may include the coordinates, depth, etc. of the gas pipeline.

[0070] In some embodiments, the government gas regulatory management platform can obtain log records from the database during the installation and construction of gas pipelines and extract the corresponding location data from them.

[0071] Pipeline number data 312 refers to data that uses numbers to identify, distinguish, and manage gas pipelines. By using the pipeline number data corresponding to the gas pipeline, the current position of the pipeline in the overall pipeline can be determined.

[0072] In some embodiments, the government gas regulatory management platform can use a pipeline numbering method to number the pipelines in the gas pipeline system, obtaining pipeline number data 312. Since there may be one or more gas pipelines in a gas pipeline system, and each gas pipeline is composed of multiple gas sub-pipelines connected together, the same pipeline numbering method can be used to number all gas pipelines, or each gas pipeline can be numbered using a separate pipeline numbering method. For example, if there are two gas pipelines, each consisting of 20 sub-pipelines, when using the same pipeline numbering method to number all gas pipelines, the sub-pipelines in the first pipeline can be numbered from 01 to 20, and the sub-pipelines in the second pipeline can be numbered from 21 to 40. When using a separate pipeline numbering method for each gas pipeline, the sub-pipelines in the first pipeline can be numbered from L1-01 to L1-20, and the sub-pipelines in the second pipeline can be numbered from L2-01 to L2-20. Here, L1 and L2 refer to the first pipeline and the second pipeline, respectively.

[0073] The target pipeline graph 321 refers to a knowledge graph representing the structure of gas pipelines. The target pipeline graph can reflect the characteristics of the location distribution of gas pipelines. The target pipeline graph 321 can consist of at least one node and at least one edge. In some embodiments, the node corresponds to the connection point between sub-pipes of the gas pipeline. When two sub-pipes are connected, the connection point is determined as a node. In some embodiments, the edge corresponds to a sub-pipe of the gas pipeline, and the attributes of the edge include the data generation rate, data complexity, number, and location of the sub-pipe. The length of the edge in the target pipeline graph can reflect the actual length of the sub-pipe. The edge can be undirected or directed. When the edge is directed, the direction is the direction of gas flow.

[0074] The data generation rate 313 refers to the rate at which the gas pipeline monitoring device generates monitoring data when monitoring the sub-pipeline section. In some embodiments, the data generation rate can be positively correlated with the number of gas pipeline monitoring devices and the detection frequency within the sub-pipeline. In some embodiments, the government gas regulatory management platform can obtain the data generation rate of the sub-pipeline over a historical time period and calculate the average historical data generation rate, determining the average historical data generation rate as the data generation rate of the sub-pipeline.

[0075] Data complexity 314 refers to the complexity of the monitoring data generated when monitoring sub-pipelines. Data complexity measures the resources consumed by the government gas regulatory management platform in processing monitoring-related data. In some embodiments, data complexity is positively correlated with the fluctuation range and accuracy of the monitoring data. For example, the greater the fluctuation range and the higher the accuracy of the monitoring data, the more resources the government gas regulatory management platform consumes in processing the monitoring-related data, and the higher the data complexity. In some embodiments, the government gas regulatory management platform can obtain the monitoring data generated by the sub-pipeline within a historical time period and the corresponding historical data complexity, calculate the average historical data complexity, and determine the average historical data complexity as the data complexity of the sub-pipeline.

[0076] Step 320: Based on the target pipeline map 321, use the first preset algorithm to generate candidate sub-region partitioning parameters.

[0077] The first preset algorithm 322 refers to an algorithm for generating candidate sub-region partitioning parameters. In some embodiments, the first preset algorithm can be: randomly assigning the sub-pipes corresponding to the nodes to groups, generating N candidate sub-region partitioning parameters. In some embodiments, the first preset algorithm can also be: meshing the location area of ​​the gas pipeline according to preset parameters to obtain several sub-regions, and dividing the sub-pipes within the same sub-region into the same group. The preset parameters may include the shape, size, and coordinates of the grid center point. By changing the preset parameters N times, N different candidate sub-region partitioning parameters can be obtained.

[0078] Candidate sub-region partitioning parameter 323 refers to one or more alternative sub-region partitioning parameters. In some embodiments, the candidate sub-region partitioning parameters correspond to multiple groups of gas pipelines corresponding to the edges of the target pipeline map. For example, the candidate sub-region partitioning parameters include [S1(03,06,08,11),S2(01,05),S3(02,09,16),S4(07,10,12),], which represent the length of each pipeline branch included in each candidate sub-region in candidate sub-regions S1, S2, S3, and S4. For example, the lengths of the four pipeline branches included in candidate sub-region S1 are 03, 06, 08, and 11, respectively (the larger the value, the longer the pipeline).

[0079] Step 330: Determine the data generation rate 313 and data complexity 314 for multiple groups, and based on the data generation rate 313 and data complexity 314, determine the fitness 331 of the candidate sub-region partitioning parameters corresponding to the multiple groups.

[0080] In some embodiments, the government regulatory management platform 110 can divide the gas pipeline based on candidate sub-region division parameters to obtain multiple groups corresponding to each candidate sub-region division parameter, and each group corresponds to a sub-region in the candidate sub-region division parameter. For more information on data generation rate 313 and data complexity 314, please refer to the relevant content above.

[0081] A fitness score of 331 can reflect the reasonableness of the candidate sub-region partitioning. In some embodiments, fitness can be represented by a numerical value or an rank. For example, the smaller the fitness score, the more reasonable the allocation of different regions.

[0082] In some embodiments, to ensure that the data generation rate and data complexity of each sub-region are as balanced as possible, the government gas regulatory management platform can set the fitness function as the sum of the variance of the data generation rate and the variance of the data complexity of each sub-region. The methods for obtaining the data generation rate and data complexity can be found in the relevant descriptions above.

[0083] Step 340: Based on the candidate sub-region partitioning parameters 323 and their corresponding fitness 331, determine the selected pipeline partitioning parameters.

[0084] Pipeline division parameter 341 refers to a set of division parameters selected from candidate sub-region division parameters. In some embodiments, the government gas regulatory management platform can select the candidate sub-region division parameters with the highest fitness from a large number of candidate sub-region division parameters and determine them as the required pipeline division parameters.

[0085] In some embodiments, the government gas regulatory management platform can determine pipeline division parameters 340 in multiple ways. For example, the platform randomly assigns the sub-pipelines corresponding to nodes to groups, generating N candidate sub-region division parameters. For each candidate sub-region division parameter, the average distance from all sub-pipelines to the corresponding management company is calculated. The N candidate sub-region division parameters are then sorted according to their average distance from smallest to largest. The candidate sub-region division parameter with the shortest average distance is selected as the required pipeline division parameter.

[0086] Step 350: Based on the selected pipeline division parameters 341, the gas pipeline is divided into multiple sub-regions 351.

[0087] The following provides an exemplary preset algorithm flow, detailing the specific implementation of determining the selected pipeline division parameter 341 in steps 320-350. In some embodiments, the preset algorithm flow can be executed by the aforementioned government gas regulatory management platform. The preset algorithm flow may include:

[0088] Step 1: Encode the sub-pipes of the gas pipeline according to their numbering data and determine the number of groups. In some embodiments, the encoding method may include, but is not limited to, binary encoding and real number encoding; the upper limit of the number of groups may be preset in advance based on prior experience.

[0089] Step 2: Based on the encoding in Step 1, generate N candidate sub-region partitioning parameters, i.e., N initial solutions (parents). In some embodiments, the government gas regulatory management platform can generate N candidate sub-region partitioning parameters using a first preset algorithm. More details about the first preset algorithm can be found above.

[0090] Step 3: Set the fitness function to determine the fitness of the partitioning parameters for each candidate sub-region. Please refer to the relevant content above for instructions on setting the fitness function.

[0091] Step 4: Perform optimal selection based on the selection function. In some embodiments, the government gas regulatory management platform can determine the selection function based on methods such as roulette wheel selection. The selection probability of a candidate sub-region partitioning parameter can be negatively correlated with fitness. For example, the lower the fitness, the greater the probability that the candidate sub-region partitioning parameter will be selected.

[0092] For example, the selection probability of a candidate sub-region partitioning parameter can be 1 - [the fitness of a candidate sub-region partitioning parameter / the total fitness value]. The total fitness value can be the sum of the fitness values ​​of all candidate sub-region partitioning parameters.

[0093] Step 5: Perform a crossover operation based on the parent candidate sub-region partitioning parameters to generate offspring. Crossover refers to the process of exchanging and combining two individuals as parents to generate offspring with mixed characteristics of their parents. In some embodiments, the government gas regulatory management platform can select two candidate sub-region partitioning parameters from multiple candidate sub-region partitioning parameters for crossover to generate new candidate sub-region partitioning parameters as offspring. The crossover probability refers to the probability that two individuals will crossover.

[0094] In some embodiments, the crossover probability can be preset based on prior experience. For example, the crossover probability can be set to (0.4 to 0.99). In some embodiments, the crossover probability can also be positively correlated with the number of sub-pipes; the more sub-pipes there are, the greater the crossover probability.

[0095] In some embodiments, the types of intersections may also include one or more intersection methods such as single-point intersection, multi-point intersection, and uniform intersection. Single-point intersection refers to the exchange of multiple data points among candidate sub-region partitioning parameters at a single intersection point. Multi-point intersection refers to the exchange of data at multiple intersection points among candidate sub-region partitioning parameters. Uniform intersection refers to sequentially scanning each data point (such as the length of a sub-pipe) in the candidate sub-region partitioning parameters and determining whether the data intersects with data from another candidate sub-region partitioning parameter based on the intersection probability.

[0096] As an example only, the parent candidate sub-region partitioning parameters include parameter A: [S1(03,06,08,11),S2(01,05),S3(02,09,16),S4(07,10,12),] and parameter B: [S1(02,04,07),S2(06,11),S3(03,19,16),S4(28,10,12),]. The government gas regulatory management platform can divide the sub-regions according to a single... By crossing points, swap 01 in S2 of parameter A with 06 in S2 of parameter B to obtain parameter C: [S1(03,06,08,11),S2(06,05),S3(02,09,16),S4(07,10,12),] and parameter D: [S1(02,04,07),S2(01,11),S3(03,19,16),S4(28,10,12),].

[0097] Step 6: Perform a mutation operation on the parent generation candidate sub-region partitioning parameters to generate offspring.

[0098] Mutation refers to altering the data of offspring. In some embodiments, the government gas regulatory management platform can mutate the candidate sub-region partitioning parameters of the parent generation based on the mutation probability, changing the data of the candidate sub-region partitioning parameters and generating new candidate sub-region partitioning parameters for the offspring generation. The mutation probability can be 0.05. In some embodiments, the types of mutation may also include one or more crossover methods such as basic bit mutation, uniform mutation, and non-uniform mutation.

[0099] Among them, basic positional mutation refers to the mutation of data at one or more randomly specified positions in the candidate sub-region partitioning parameters based on the mutation probability. Uniform mutation refers to replacing each data point in the candidate sub-region partitioning parameters with random numbers uniformly distributed within a preset range according to the mutation probability. Non-uniform mutation refers to randomly perturbing the data in the candidate sub-region partitioning parameters, and using the perturbed result as the mutated data.

[0100] As an example only, the parent candidate sub-region partitioning parameters include parameter A: [S1(03,06,08,11),S2(01,05),S3(02,09,16),S4(07,10,12),]. The government gas regulatory management platform can mutate 02 in S3 of parameter A according to the basic bit mutation method to obtain a random value 21, thus obtaining parameter E[S1(03,06,08,11),S2(01,05),S3(21,09,16),S4(07,10,12),].

[0101] Step 7: Select the offspring obtained in Steps 5 and 6.

[0102] In some embodiments, the government gas regulatory management platform can select candidate sub-region division parameters that meet the retention criteria. The retention criteria refer to the judgment conditions used to eliminate unreasonable candidate sub-region division parameters. In order to ensure that the pipelines in the sub-regions divided by the candidate sub-region division parameters can be connected as much as possible, and that pipelines that are too far apart are not placed in the same sub-region, the retention criteria need to be used for screening.

[0103] In some embodiments, the retention conditions may include: the number of nodes minus the number of edges in the pipeline feature map corresponding to the candidate sub-region partitioning parameters is ≤1, and the maximum positional distance between sub-pipes is within a preset distance range.

[0104] When the number of nodes minus the number of edges is greater than 1, it indicates that at least two points in the current sub-region are not connected by edges, thus failing to meet the condition that as many sub-pipes as possible in the sub-region can be connected. Considering that closely spaced sub-pipes may also have some degree of mutual influence, it is also necessary to consider whether the maximum distance between sub-pipes is within the preset distance range.

[0105] When the maximum distance between sub-pipes is less than the lower limit of a preset distance range, or greater than the upper limit of the preset distance range, the sub-region division is unreasonable. By limiting the maximum distance to within the preset distance range, it can be ensured that the area of ​​the generated sub-region is neither too large nor too small. In some embodiments, the preset distance range can be determined relative to the average evaluation score of each gas company. For more information on this section, please refer to [link to relevant documentation]. Figure 4 And related explanations.

[0106] Step 8: Replace the undesirable candidate sub-region partitioning parameters (parent generation) with the candidate sub-region partitioning parameters (child generation) that meet the conditions.

[0107] In some embodiments, the government gas regulatory management platform can use newly generated candidate sub-region partitioning parameters (such as the candidate sub-region partitioning parameters of the offspring generated in steps 5 and 6) to replace the undesirable candidate sub-region partitioning parameters with high fitness in the original population (such as the candidate sub-region partitioning parameters of the parent generation generated in steps 1 and 2).

[0108] For example, a government gas regulatory management platform can sort candidate subregion partitioning parameters (parents) in the original population and newly generated candidate subregion partitioning parameters (offspring) according to their fitness from highest to lowest. It can then select a predetermined number of newly generated candidate subregion partitioning parameters (offspring) with lower fitness to replace a predetermined number of original candidate subregion partitioning parameters (parents) with higher fitness. The predetermined number can be set based on prior experience.

[0109] Step 9: Repeat steps 3 to 8 above to continuously evolve until the iteration stopping condition is met.

[0110] In some embodiments, the iteration stopping condition may include: the maximum number of evolutions has been completed, the fitness has reached a preset expected value, the fitness value remains the same as the preset number of iterations, or the fitness difference between two iterations is lower than a preset difference threshold. The number of evolutions, the preset expected value, the preset number of iterations, and the preset difference threshold can be set using various methods such as manual experience or online queries.

[0111] In some embodiments, after the government gas regulatory management platform stops iterating, it can select the candidate sub-region division parameter with the lowest fitness as the target pipeline division parameter, determine the number of groups corresponding to the candidate sub-region division parameter with the lowest fitness as the number of groups in the target pipeline division parameter, and divide according to the number of the gas sub-pipeline corresponding to each group.

[0112] Step 360: Determine the management mapping relationship 361 between multiple sub-regions 351 and the management company, and generate a routing instruction 371 based on the management mapping relationship 361 and send it to the gas data forwarding device of the gas pipeline network to create the corresponding route.

[0113] The management mapping relationship 361 refers to the correspondence between different sub-regions and the management companies that manage those sub-regions. To monitor the gas pipelines in a sub-region, the data collected by the gas pipeline monitoring devices in that sub-region needs to be sent to the corresponding management company via a corresponding route. Here, the management company refers to the company that maintains and manages the gas pipelines, and the route refers to the control rules used to control how the monitoring data is transmitted. By creating corresponding routes through routing instructions, the gas data forwarding equipment in the gas pipeline network can forward the received monitoring data to the gas company management platform corresponding to the management company, and can also forward the received monitoring data from the sub-regions to the gas company management platform corresponding to their respective management companies.

[0114] A gas data forwarding device refers to a forwarding device that forwards monitoring data collected by a gas pipeline monitoring device to a corresponding platform. In some embodiments, the gas data forwarding device may include, but is not limited to, a router, a server, etc.

[0115] Routing instruction 371 refers to an instruction used to create a route. In some embodiments, the routing instruction can be: based on the management mapping relationship, controlling the gas data forwarding device to forward the monitoring data of the corresponding sub-area to the gas company management platform corresponding to the management company.

[0116] The government gas regulatory management platform can determine the management mapping relationship in various ways. 361. In some embodiments, the government gas regulatory management platform can calculate the distance between a sub-region and all gas companies near that sub-region, and select the management company with the shortest distance as the management company corresponding to that sub-region. After determining the management companies corresponding to all sub-regions, the management mapping relationship can be obtained.

[0117] In some embodiments, the government gas regulatory management platform may also identify candidate management companies for each of the multiple sub-regions; and from the candidate management companies, determine the management company corresponding to each sub-region based on the evaluation scores of the candidate management companies and the importance of the sub-region.

[0118] The candidate management company refers to a company that can manage a sub-region. In some embodiments, the government gas regulatory management platform can calculate the company's comprehensive index based on the candidate management company's evaluation score and the importance of the sub-region, and select the candidate management company with the highest comprehensive index as the corresponding management company.

[0119] The evaluation score reflects the management company's ability to manage gas pipelines. In some embodiments, the evaluation score may be negatively correlated with the gas failure rate within a sub-area. For further information on the evaluation score, please refer to [link to relevant documentation]. Figure 5 Related descriptions.

[0120] For example, the overall index of a management company can be calculated using formula (1):

[0121] U=k1×D+k2×Y (1)

[0122] Where U is the comprehensive index of the management company, k1 and k2 are coefficients, D is the actual distance from the management company to the sub-region, and Y is the evaluation score of the management company. The coefficients k1 and k2 can be preset based on prior experience.

[0123] Importance reflects the degree to which a sub-region needs to be given special attention. In some embodiments, importance can be positively correlated with the total number of sub-channels within the sub-region. For example, the more sub-channels a sub-region has, the greater the likelihood of anomalies occurring in that sub-region, the more special attention it needs to receive, and thus the higher its importance.

[0124] In some embodiments, the government's gas regulatory management platform can determine the importance of sub-regions based on a constructed pipeline connection map. This pipeline connection map is a drawing that reflects the connection relationships between pipelines and users.

[0125] In some embodiments, the pipeline connection graph may consist of at least one node and at least one edge. The node may further include pipeline nodes and user nodes. Pipeline nodes correspond to the branching points of the gas pipeline. When a gas pipeline needs to branch or connect, the connection point is identified as a pipeline node. User nodes are nodes with a non-zero in-degree and a zero out-degree, and they correspond to gas users.

[0126] In some embodiments, the edge corresponds to a gas pipeline. The edge is a directed edge, and its direction is the direction of gas flow.

[0127] In some embodiments, the importance of a pipeline can be positively correlated with the average proximity of the pipeline node to all user nodes and the number of user paths. The proximity of a pipeline node to a user node refers to the shortest path from the pipeline node to the user node, including the number of nodes along the path (user nodes are also counted). The number of user paths refers to the number of user nodes that the pipeline node can reach along the gas flow direction. For example, in the pipeline node A → pipeline node B → user node C, the proximity of pipeline node A and user node C is 2, and the number of user paths for pipeline node A is 1.

[0128] For example, the importance of a pipeline can be calculated using formula (2):

[0129]

[0130] Where M represents the importance of the pipeline, k1 and k2 are coefficients, and L... nLet N be the proximity of the pipeline node to the user node n, N be the total number of user nodes reachable from the pipeline node, and P be the number of user paths. The coefficients k1 and k2 can be preset based on prior experience, with k2 > k1.

[0131] In some embodiments, the statistical values ​​(e.g., mean, maximum, median, etc.) of the importance of pipes in a sub-region can be determined as the importance of the sub-region.

[0132] In some embodiments, the management company corresponding to a sub-region may be determined by following these steps:

[0133] Step A involves determining the candidate management companies for each sub-region according to the established criteria, and then ranking them from highest to lowest based on their comprehensive index. It also involves determining all sub-regions that each candidate management company could potentially manage, and then ranking them from highest to lowest based on their respective importance. The criteria may include ensuring that the distance between the sub-region and the gas company is less than a preset distance threshold.

[0134] Step B: Select the company with the highest comprehensive index ranking among the candidate management companies corresponding to the sub-region as the candidate management company for that sub-region. Determine whether the importance ranking of the sub-region exceeds the maximum management threshold of the management company. If it does not exceed the threshold, proceed to step C-1. If it does exceed the threshold, proceed to step C-2.

[0135] Step C-1: Select the candidate management company as the management company for this sub-region. Complete.

[0136] Step C-2: Remove the candidate management company from the list of candidate management companies corresponding to the sub-region, and proceed to step B.

[0137] Repeat steps B and C-2 above, and repeat the cycle until step C-1 is satisfied.

[0138] In some embodiments of this specification, by judging the comprehensive index of the candidate management companies, ranking the candidate management companies, and then ranking the sub-regions that the candidate management companies can manage according to their importance, it can be ensured that more important sub-regions are assigned to more experienced management companies, while less important sub-regions are assigned to management companies with average management capabilities, thereby reducing the failure rate of pipelines.

[0139] In some embodiments, in response to the fulfillment of replacement conditions, the government gas regulatory management platform may re-determine the management company corresponding to the sub-region.

[0140] The change condition refers to changing the judgment criteria for the management company corresponding to the sub-region. In some embodiments, the change condition may be that the frequency of abnormal situations occurring in the sub-region under the management of the management company is greater than a preset threshold.

[0141] An abnormal situation refers to an abnormality in the gas pipeline. In some embodiments, an abnormal situation may include gas leaks, gas pipeline malfunctions, and false alarms or misjudgments.

[0142] In some embodiments, when the replacement conditions are met, the government gas regulatory management platform can, based on step C-2 above, remove the company from the list of candidate management companies corresponding to the sub-region, and jump to step B to reselect the company with the highest comprehensive index ranking as the management company for that sub-region.

[0143] In some embodiments of this specification, when a sub-region under the jurisdiction of a management company repeatedly experiences malfunctions or missed judgments, it indicates that the current management company lacks the capability to manage that sub-region. In this case, updating the management mapping relationship and changing the management company managing the sub-region can be considered to ensure adequate supervision of the sub-region.

[0144] In some embodiments of this specification, by analyzing and processing sub-pipes of gas pipelines using target pipeline maps, it is possible to effectively uncover potential correlations between different pieces of information from chaotic and complex data.

[0145] Figure 4 This is a structural schematic diagram illustrating the determination of the actual leak point according to some embodiments of this specification.

[0146] like Figure 4 As shown, determining the actual leak point on the gas pipeline includes determining the identification feature 421 based on historical gas sensor data 411, historical actual leak points 412, historical weather data 413, and pipeline data 414 at each point in multiple sub-regions.

[0147] Historical gas sensor data 411 refers to gas sensor data generated within a historical period. For more information on gas sensor data, please refer to [link / reference needed]. Figure 2 Related descriptions.

[0148] Historical weather data 413 refers to weather-related data generated within a historical period. In some embodiments, historical weather data 413 may include at least one of temperature data, precipitation data, humidity data, etc.

[0149] In some embodiments, the government gas regulatory management platform may establish a database, store historical weather data 413 in the database, and read historical weather data 413 from the database.

[0150] For more information on pipeline data 414, please refer to [link / reference]. Figure 2 Description of pipeline data.

[0151] Identification feature 421 refers to a feature that can be used for identification and is distinct from one another. In some embodiments, identification feature 421 may include different types of data. For example, identification feature 421 may include at least one of flow rate, pressure, temperature, etc. Identification feature data corresponding to identification feature 421 may include at least one of flow rate data, pressure data, temperature data, etc.

[0152] In some embodiments, the government gas regulatory management platform can determine the identification feature 421 through various methods. For example, the government gas regulatory management platform can determine the identification feature 421 through feature extraction and / or feature selection. Feature extraction and / or feature selection refers to removing redundant data from multiple datasets and extracting the required data.

[0153] In some embodiments, feature extraction may include at least one of Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), and Singular Value Decomposition (SVD). Feature selection may include at least one of Information Gain, Relief algorithm, and Chi Squares algorithm.

[0154] In some embodiments, the government gas regulatory management platform may use feature selection based on clustering analysis algorithms to determine identification feature 421. In some embodiments, feature selection based on clustering analysis algorithms may include the following steps:

[0155] Step S1: Select at least one subset of features from the original features.

[0156] The original features refer to the data set formed by raw data generated over multiple historical periods. In some embodiments, the original features may include at least one of historical gas sensor data 411, historical weather data 413, pipeline data 414, etc. In some embodiments, the government gas regulatory management platform may determine the original features through various methods, such as reading from historical data or obtaining manual input.

[0157] A feature subset refers to a set of data selected from the original features. For example, a feature subset may include at least one of the following: (historical gas sensor data, historical weather data), (historical gas sensor data, pipeline data), (historical weather data, pipeline data), etc.

[0158] In some embodiments, the government gas regulatory management platform may select at least one data point from the original features to form a feature subset.

[0159] Step S2: Perform cluster analysis on the feature subset.

[0160] In some embodiments, the government gas regulatory management platform may use clustering algorithms to perform clustering analysis on selected feature subsets. The clustering algorithm may include at least one of K-means clustering, hierarchical clustering, etc. In some embodiments, the government gas regulatory management platform may cluster the data corresponding to each feature subset.

[0161] Step S3: Evaluate the feature subset.

[0162] In some embodiments, the government gas regulatory management platform can use various methods to score the clustering effect of feature subsets. For example, it can compare the similarity between the results of clustering using the original features and the results of clustering using feature subsets; the higher the similarity, the higher the clustering score.

[0163] The similarity can include at least one of the following: the similarity of the number of clusters obtained by clustering, the similarity of the elements included in each cluster, etc. For example, the original features include 10 data points, numbered 1, 2, 3, ..., 10. When clustering using the original features, the original features are divided into three clusters: (1, 2, 5, 7), (3, 9), and (4, 6, 8, 10). The similarity between the clustering results of the feature subsets and the three clusters is compared.

[0164] Step S4: Determine the identification features.

[0165] In some embodiments, the government gas regulatory management platform may calculate a comprehensive score using various methods; based on the comprehensive score, identification features may be determined.

[0166] The overall score is a rating used to evaluate a subset of features. In some embodiments, the overall score may be positively correlated with the clustering score and negatively correlated with the number of features in the feature subset. A higher clustering score and fewer features in the feature subset result in a higher overall score.

[0167] In some embodiments, the government gas regulatory management platform may use a third preset algorithm to calculate a comprehensive score. For example, the third preset algorithm may be: Comprehensive Score = First Coefficient × Cluster Score - Second Coefficient × Number of Features in the Feature Subset. Wherein, the first coefficient and the second coefficient may be preset values.

[0168] In some embodiments, such as Figure 4As shown, the government gas regulatory management platform can determine the actual location 431 of the actual leak point and the corresponding leak result 432 of each actual leak point based on the identification feature 421, the current gas sensor data 422, and the historical gas sensor data 411.

[0169] Actual location 431 refers to the actual location where a leak occurred on the gas pipeline. In some embodiments, actual location 431 can be characterized by at least one of coordinates, latitude and longitude, gas pipeline number, etc.

[0170] Leakage result 432 refers to data characterizing whether a leak has occurred on the gas pipeline. In some embodiments, leakage result 432 includes whether a leak has occurred and the extent of the leak.

[0171] Leakage intensity refers to the degree to which a leak actually occurs in a gas pipeline. In some embodiments, leakage intensity can be expressed in grades, such as Class I leak, Class II leak, Class III leak, etc.

[0172] In some embodiments, the government gas regulatory management platform can determine the leakage intensity based on leakage data from when gas leaks from the gas pipeline at the actual point of origin 431. The leakage data may include at least one of gas pressure, flow velocity, and flow rate. For example, the government gas regulatory management platform can compare one of the leakage data points with its corresponding preset leakage threshold. The preset leakage threshold may include a first leakage threshold, a second leakage threshold, a third leakage threshold, etc., that increase sequentially. When at least one of the leakage data points corresponds to a value greater than the first leakage threshold, the second leakage threshold, and / or the third leakage threshold, the government gas regulatory management platform can determine the leakage intensity as a Level 1, Level 2, and / or Level 3 leakage. In some embodiments, the government gas regulatory management platform can also comprehensively consider multiple leakage data points to obtain a comprehensive consideration result. The comprehensive consideration result is then compared with its corresponding preset leakage threshold. The comprehensive consideration may include using a fourth preset algorithm to calculate multiple leakage data points. For example, calculating the average value of multiple leakage data points.

[0173] In some embodiments, leakage data can be acquired through sensors, such as at least one of pressure sensors, flow rate sensors, and flow sensors. In some embodiments, the preset leakage threshold can be acquired based on at least one of experience, historical data, or human input.

[0174] In some embodiments, the government gas regulatory management platform can determine the identification feature data corresponding to identification feature 421 based on identification feature 421. For example, when identification feature 421 includes flow rate and pressure, the identification feature data includes flow rate data and pressure data.

[0175] In some embodiments, the government gas regulatory management platform can select gas sensor data corresponding to the identified feature data from the current gas sensor data 422 based on the identified feature data, and construct a leakage feature vector based on the selected gas sensor data. For example, the identified feature 421 includes flow rate and pressure. The current flow rate data and pressure data are selected from the current gas sensor data 422, and a vector (x, y) is constructed based on the flow rate data and pressure data. Here, x can represent the flow rate data, and y can represent the pressure data.

[0176] In some embodiments, the government gas regulatory management platform can construct a vector database based on historical data. For example, the vector database can be constructed based on historical gas sensor data 411, historical actual leak points, historical leak intensities, and other data. In some embodiments, the government gas regulatory management platform can match the leak feature vectors in the vector database, select the historical leak vector with the highest similarity to the leak feature vector, and use the historical actual location and historical leak result corresponding to the historical leak vector as the actual location and leak result corresponding to the leak feature vector.

[0177] The embodiments provided in this specification, by identifying identification features, enable targeted selection of collected data, which can effectively reduce the data dimensionality, reduce the amount of data that needs to be analyzed when determining whether a leak has occurred, thereby reducing the amount of computation, speeding up the calculation, avoiding redundant data from interfering with the calculation process, and preventing delays in subsequent personnel dispatch for maintenance due to untimely judgment.

[0178] The second preset algorithm refers to the algorithm that determines the correlation between historical potential hazard points and historical actual leakage points.

[0179] A strong association rule is an association rule that satisfies both support and confidence conditions. In some embodiments, a strong association rule may include (location 1, potential vulnerability point exists), (location 2, potential vulnerability point exists) → (location 3, current actual leakage point). That is, if potential vulnerability points exist at locations 1 and 2, it can be inferred that an actual leakage point exists at location 3.

[0180] In some embodiments, the support condition may include a support level greater than or equal to a support threshold. In some embodiments, the support threshold may be obtained based on at least one of experience, historical data, or obtaining human input. In some embodiments, the confidence condition may include a confidence level greater than or equal to a confidence threshold. In some embodiments, the confidence threshold may be obtained based on at least one of experience, historical data, or obtaining human input.

[0181] The support of a strong association rule is represented by S(A→B) in the algorithm, which refers to the proportion of data that covers the association rule. In some embodiments, the formula (3) for calculating the support is:

[0182]

[0183] Here, `itemset including A∪B` refers to the amount of data where both events A and B occur. For example, the amount of data where potential vulnerabilities are found at point 1 and point 2, while a leak actually occurs at point 3. `totalitemset` refers to the total amount of data. For example, the total amount of data for historically acquired potential vulnerabilities and actual leaks.

[0184] The confidence score of a strong association rule is represented as C(A→B) in the algorithm. The confidence score measures the credibility of the strong association rule. In some embodiments, the confidence score is calculated using formula (4):

[0185]

[0186] Here, S(A∪B) refers to the amount of data where both events A and B occur. S(A) refers to the amount of data where event A occurs. C(A→B) refers to the proportion of the amount of data where both events A and B occur to the amount of data where event A occurs.

[0187] In some embodiments, the support threshold and confidence threshold may be related to the frequency of historical missed detections and the frequency of historical actual leaks.

[0188] Historical missed detection frequency refers to the frequency with which leaks actually occurred but were not detected within a certain historical period. In some embodiments, the historical missed detection frequency can be determined based on multiple methods, such as obtaining it based on historical data, table lookup, or at least one of these methods.

[0189] Historical actual leakage frequency refers to the frequency of actual leaks occurring over a historical period. In some embodiments, the historical actual leakage frequency can be determined based on a variety of methods, such as obtaining it based on historical data, table lookup, or at least one of these methods.

[0190] In some embodiments, the support threshold and confidence threshold can be negatively correlated with the frequency of historical missed detections and the frequency of historical actual leaks.

[0191] By relating the support threshold and confidence threshold to the frequency of historical missed detections and the frequency of historical actual leaks, the accuracy of the confirmed support threshold and confidence threshold can be improved.

[0192] The actual leak points to be updated refer to the actual leak points calculated using strong association rules.

[0193] In some embodiments, the government gas regulatory management platform can construct a dataset table based on historical potential hazard points and historical actual leak points.

[0194] The dataset table may include the historical actual status of each point in multiple sub-regions along the gas pipeline within the target area. In some embodiments, the dataset table (1) may include:

[0195]

[0196]

[0197] The time point represents the time when the gas sensor data was collected. x1, x2, x3, and x4 can represent the first, second, third, and fourth locations, respectively. -1 indicates that the location is identified as a potential hazard point. 0 indicates that the location is normal. 1 indicates that the location is identified as an actual leak point.

[0198] In some embodiments, the government gas regulatory management platform can determine the actual leak points to be updated based on strong association rules using the Apriori algorithm.

[0199] For example, the preset support threshold and confidence threshold are 0.6 and 0.7, respectively. The number of iterations in the Apriori algorithm can be k, where k = 1, 2, 3...n.

[0200] When k=1, at least one-itemset is generated. One-itemsets with support less than the support threshold are discarded, and the remaining one-itemsets are identified as frequent one-itemsets L1.

[0201] For example, based on Table (1), eight one-itemsets can be generated: {(x1, -1)}, {(x1, 1)}, {(x2, -1)}, {(x2, 1)}, {(x3, -1)}, {(x3, 1)}, {(x4, -1)}, and {(x4, 1)}. The support of each of the eight one-itemsets is calculated. For example, {(x1, -1)} appears at times 1, 2, 4, and 5, so its support is 4 / 5 = 0.8. {(x4, 1)} appears only at time 4, so its support is 1 / 5 = 0.2. Normal points do not need to be considered; therefore, it is not necessary to generate one-itemsets such as {(x1, 0)}, {(x2, 0)}, {(x3, 0)}, and {(x4, 0)}.

[0202] In some embodiments, the government gas regulatory management platform can construct a one-itemset support table based on the support of one-itemsets. In some embodiments, the one-itemset support table (2) may include:

[0203] 0.8 0 0.4 0.2 0.2 0.6 0.6 0.2

[0204] If only the support of {(x1, -1)}, {(x3, 1)}, and {(x4, -1)} is greater than or equal to the support threshold, then the determined frequent 1-itemset L1 = By determining strong association rules, potential hazard points that meet the requirements can be identified as actual leak points to be updated based on these rules. The actual leak points to be updated and the actual leak points can then be confirmed as the updated actual leak points, improving the accuracy of actual leak point confirmation. When dispatching maintenance personnel, potential hazard points can be repaired, reducing the risk of gas pipeline leaks.

[0205] Figure 5 This is a structural schematic diagram of a potential hazard prediction model based on some embodiments of this application specification.

[0206] The hazard prediction model 520 refers to a model that can be used to determine identified potential hazard points 531 at future points in time. The hazard prediction model 520 can be a machine learning model, such as a Long Short-Term Memory (LSTM) recurrent neural network.

[0207] In some embodiments, the inputs to the hazard prediction model 520 may include historical maintenance records 511, current gas sensor data 512, historical potential hazard points 513, historical actual leak points 514, and future weather data 515, and the output may be a identified potential hazard point 531 and the actual leak time 532 at a future time.

[0208] Historical maintenance records 511 refer to relevant records generated during maintenance of gas pipelines within a historical period. In some embodiments, the government gas regulatory management platform can obtain historical maintenance records 511 through various methods, such as from historical data. In some embodiments, historical maintenance records 511 may include at least one of the following: maintenance time, number of maintenance operations, and maintenance location.

[0209] Current gas sensor data 512 refers to gas sensor data generated within the current time period. Historical potential hazard points 513 refer to potential hazard points identified within a historical time period. Historical actual leak points 514 refer to actual leak points identified within a historical time period. For more information on gas sensor data, potential hazard points, actual leak points, and pipeline data 516, please refer to [link to relevant documentation]. Figure 2 Related descriptions.

[0210] Future weather data 515 refers to weather-related data for a future time period. In some embodiments, future weather data 515 is similar to historical weather data. For more information on future weather data 515, please refer to [link / reference needed]. Figure 4 Related descriptions.

[0211] The actual occurrence time of 532 refers to the time it takes for a potential hazard point to become an actual leak point.

[0212] The stronger the management capabilities of the management company, the lower the probability or number of potential hidden dangers will be in the sub-regions corresponding to the management company.

[0213] In some embodiments, the input to the hazard prediction model 520 may also include the evaluation score of the management company corresponding to the sub-region.

[0214] The evaluation score is a score derived from evaluating the management capabilities of a management company, and it can characterize the management company's capabilities. In some embodiments, the government gas regulatory management platform can determine the evaluation score through various methods based on the number of sub-regions managed by the management company corresponding to the sub-region, the frequency of historical missed detections (missed judgment frequency), the average command response speed, and the average fault frequency of the managed sub-regions. In some embodiments, the evaluation score can be negatively correlated with the number of sub-regions, the frequency of historical missed detections, the average command response speed, and the average fault frequency.

[0215] In some embodiments, the government gas regulatory management platform can determine an evaluation score based on a preset algorithm. For example, the evaluation score = 100 - (coefficient 1 × number of managed sub-regions + coefficient 2 × historical missed detection frequency + coefficient 3 × average fault frequency of the managed sub-regions + coefficient 4 × average command response speed). Here, coefficients 1, 2, 3, and 4 are preset values ​​and can be obtained based on at least one of experience, historical data, etc. For more information on the historical missed detection frequency, please refer to [link to relevant documentation]. Figure 5 Related explanations.

[0216] Command response speed refers to the time interval between the government issuing a command to the management company and the management company commencing action. Average command response speed refers to the average of multiple command response speeds over a period of time. In some embodiments, the government gas regulatory management platform can obtain command response speed through various methods, such as historical data or manual input. Fault frequency refers to the frequency of gas pipeline failures. In some embodiments, the government gas regulatory management platform can determine fault frequency through various methods, such as historical data or manual input. Using evaluation scores as input to the hazard prediction model allows the management company's capabilities to be considered when identifying potential hazard points, which helps improve the accuracy of potential hazard point identification.

[0217] In some embodiments, the method for obtaining the hazard point prediction model includes: performing reinforcement training on the initial hazard point prediction model based on the target training samples corresponding to the target area, and obtaining the trained hazard point prediction model.

[0218] In some embodiments, the hazard prediction model can be trained using multiple labeled target training samples. The government gas regulatory management platform can input these labeled target training samples into the initial hazard prediction model, construct a loss function using the labels and the results of the initial hazard prediction model, and iteratively update the parameters of the initial hazard prediction model based on the loss function. The model training is complete when the loss function of the initial hazard prediction model meets preset conditions, resulting in a trained hazard prediction model. These preset conditions may include loss function convergence, the number of iterations reaching a threshold, etc.

[0219] In some embodiments, when training the hazard prediction model, a first hazard prediction model can be obtained by training with multiple labeled initial training samples. Then, based on the target training sample data corresponding to the target area, the first hazard prediction model is subjected to secondary reinforcement training to obtain a second hazard prediction model, which serves as the final hazard prediction model.

[0220] This ensures the basic accuracy of the first hazard prediction model. Furthermore, considering the characteristics of different target areas and the potential for significant differences in training samples due to varying environments, a second hazard prediction model is specifically trained for each target area. This improves the accuracy and reliability of the final hazard prediction model and saves training time, eliminating the need to retrain the first hazard prediction model for each target area.

[0221] In some embodiments, the target training samples may include historical maintenance records, gas sensor data, historical potential hazard points, historical actual leak points, weather data, and pipeline data at a given moment. In some embodiments, the target training samples are acquired based on historical data of the target area.

[0222] In some embodiments, the label corresponding to the target training sample includes the actual leak point and the actual time of the leak at a second time point; wherein the second time point is later than the first time point. In some embodiments, the target training sample may also include the evaluation score of the management company corresponding to the sub-region. For more information on the evaluation score, please refer to the relevant description above.

[0223] In some embodiments, to enhance the adaptability and generality of the hazard prediction model, the target training samples include multiple training samples from multiple different occurrence times. Training the hazard prediction model using multiple training samples from multiple different occurrence times can improve the accuracy of the hazard prediction model in predicting potential hazard points in different time periods.

[0224] In some embodiments, the government gas regulatory management platform may determine target training samples based on a preset range.

[0225] The preset range refers to the proportion of samples corresponding to a specific actual time period within all samples. In some embodiments, the government gas regulatory management platform obtains target training samples from historical data of the target area. For example, the time period from the moment the sample data was collected to the point where a leak occurred is called the actual time period and serves as the target training sample. Historical data within a certain period is divided into multiple segments according to the number of time periods and the preset range, thereby obtaining multiple training samples corresponding to multiple time periods. The number of time periods refers to the number of segments after dividing a period. The durations of multiple time periods can be the same or different.

[0226] In some embodiments, the preset range can be calculated based on a fifth preset algorithm. The fifth preset algorithm can be: Preset Range = (100% / Number of Time Periods). For example, starting from a certain moment, samples of all labels corresponding to actual occurrence time periods within two hours are obtained from historical data. The preset number of time periods is 4. Each preset range is 25%. That is, the proportion of historical data within each time period that falls within the two-hour time range is 25%. The two hours are divided into segments according to the preset range, with historical data within each of the four time periods (0-15min, 15-30min, 30-60min, and 60-120min) accounting for 25%. Historical data within the four ranges (0-15min, 15-30min, 30-60min, and 60-120min) can be used as training samples, and labels are set based on the actual occurrence time periods (0-15min, 15-30min, 30-60min, and 60-120min) corresponding to the training samples. In some embodiments, the length of each time period can be the same or different. For example, starting from a certain moment, historical data within a two-hour time range is obtained. The preset number of time periods is 4. Historical data from four time periods—0-30min, 30-60min, 60-90min, and 90-120min—can be used as training samples.

[0227] In some embodiments, the government gas regulatory management platform can determine the number of training samples corresponding to each actual occurrence time in the target training sample based on the historical leakage frequency of the target area, gas company size data, gas pipeline data, and bandwidth data.

[0228] In some embodiments, the government gas regulatory management platform can construct a training feature vector (x0, y0, z0, t0) based on historical leakage frequency, gas company size data, gas pipeline data, and bandwidth data for the target area. Based on this training feature vector, the similarity between the training feature vector and standard vectors in the vector database is compared. The standard vector with the highest similarity is determined, and the number of training samples corresponding to the highest similarity standard vector is determined as the target number of training samples for each actual leakage time period. In some embodiments, the government gas regulatory management platform can determine the number of training samples corresponding to the standard vector through various methods, such as obtaining them from historical data.

[0229] Gas company size data refers to data related to the size of management companies within a target area. For example, it may include at least one of the following: the number of management companies corresponding to the target area, the number of sub-areas managed by each management company, and the number of employees (e.g., maintenance personnel) of the management company. In some embodiments, gas company size data may be obtained based on historical data. Gas pipeline data refers to data related to gas pipelines. For example, it may include at least one of the following: the number of gas pipelines, the size of their distribution area, and the area they occupy. In some embodiments, gas pipeline data may be obtained based on historical data. Bandwidth data refers to data related to the transmission network. For example, it may include at least one of the following: data transmission rate and transmission volume. In some embodiments, bandwidth data may be obtained from the vendor.

[0230] If a target area has a large number of management companies with large scale and wide bandwidth, and the number and distribution area of ​​gas pipelines in that area are small, then the average number of gas pipelines managed by each management company is less and the area is smaller, resulting in less management pressure on the management companies. In this case, the number of training samples for time periods closer to the present can be reduced, while the number of training samples for time periods farther from the present can be increased. Conversely, the number of training samples for time periods closer to the present should be increased, while the number of training samples for time periods farther from the present should be decreased. By obtaining multiple training samples for multiple time periods, different training samples can be selectively used to train the hazard prediction model, thereby obtaining different training effects. For example, if the management companies in the target area have sufficient management capabilities, they can focus on potential hazard points corresponding to recent (e.g., within the next 30 minutes) and long-term (e.g., within the next 90-120 minutes) time points. Training samples from multiple time periods, including recent and long-term, can be selected to train the hazard prediction model, improving the model's prediction accuracy for potential hazard points corresponding to recent and long-term time points. For example, if the management capabilities of the management company in the target area are insufficient, it can focus only on potential hidden dangers corresponding to recent time points. It can select recent training samples to train the hidden danger prediction model, thereby improving the prediction accuracy of the hidden danger prediction model for potential hidden dangers corresponding to recent time points.

[0231] Based on the current actual leakage points, a second preset algorithm is used to estimate potential hidden danger points to be updated at future time points; and based on the identified potential hidden danger points and the potential hidden danger points to be updated, the updated potential hidden danger points are determined.

[0232] In some embodiments, the government gas regulatory management platform can use a second preset algorithm based on strong correlation rules and the current actual leak point to estimate potential hazard points to be updated at future time points. For example, the government gas regulatory management platform determines the strong correlation rule {(x3, 1)} → {(x1, -1)}. When x3 is confirmed as an actual leak point, point x1 will be identified as a potential hazard point to be updated. The method for determining potential hazard points to be updated is similar to the method for determining actual leak points to be updated. For more information on strong correlation rules, the second preset algorithm, and how to determine potential hazard points to be updated, please refer to [link to relevant documentation]. Figure 4 Related descriptions.

[0233] In some embodiments, the government gas regulatory management platform can determine updated potential hazard points based on potential hazard points and potential hazard points to be updated. In some embodiments, the updated potential hazard points can be the union of potential hazard points and potential hazard points to be updated. By using a hazard point prediction model to predict potential hazard points, the prediction accuracy of potential hazard points can be improved, thereby facilitating subsequent monitoring of potential hazard points and ensuring the safe use of gas pipelines.

Claims

1. A smart gas pipeline zoned safety supervision method, characterized in that, The method is executed by the government's gas regulatory management platform and includes: Based on the gas pipeline distribution data in the target area, the target area is divided into multiple sub-regions, including: Based on the gas pipeline distribution data, a target pipeline map is constructed; the gas pipeline distribution data includes the location data and pipeline number data of the gas pipelines. Based on the target pipeline map, a first preset algorithm is used to generate candidate sub-region partitioning parameters. Each candidate sub-region partitioning parameter corresponds to multiple groups obtained by partitioning the sub-pipes corresponding to the edges of the target pipeline map. The first preset algorithm includes at least one of the following: Randomly assign the sub-pipes corresponding to the edges of the target pipeline map to multiple groups to generate multiple candidate sub-region partitioning parameters; or The location area of ​​the gas pipeline is gridded according to preset parameters to obtain multiple sub-regions. Sub-pipes within the same sub-region are divided into the same group. Multiple candidate sub-region division parameters are generated by changing the preset parameters multiple times. The preset parameters include the shape, size and center point coordinates of the grid. Determine the data generation rate and data complexity of the multiple groups, and based on the data generation rate and data complexity, determine the fitness of the candidate sub-region partitioning parameters corresponding to the multiple groups; Based on the candidate sub-region partitioning parameters and their corresponding fitness, the selected pipeline partitioning parameters are determined; Based on the selected pipeline division parameters, the gas pipeline is divided into multiple sub-regions; Determine the management mapping relationship between the plurality of sub-regions and management companies, wherein the management mapping relationship includes the management company corresponding to each sub-region; and Based on the management mapping relationship, a routing instruction is generated and sent to the gas data forwarding device of the gas pipeline network to create the corresponding route; Through the government's gas regulatory sensor network platform, gas sensor data of each point in the multiple sub-areas on the gas pipeline within the target area is obtained from the management platforms of multiple gas companies. Based on the gas sensor data and pipeline data, the actual leak point on the gas pipeline is determined; Based on the gas sensor data and historical maintenance records, potential hidden danger points on the gas pipeline were identified. In response to the existence of the actual leak point: based on the actual location of the actual leak point on the gas pipeline, a first maintenance command and a first adjustment command are generated; the first maintenance command is sent to the gas company management platform corresponding to the actual location to instruct the gas company management platform to dispatch maintenance personnel to maintain the actual location; and the first adjustment command is sent to the gas pipeline control equipment corresponding to the actual location to adjust the operating parameters of the gas pipeline control equipment. In response to the existence of the potential hazard point: based on the potential location of the potential hazard point on the gas pipeline, a second maintenance instruction and a second adjustment instruction are generated; the second maintenance instruction is sent to the gas company management platform corresponding to the potential location to instruct the gas company management platform to dispatch maintenance personnel to inspect the potential location; and the second adjustment instruction is sent to the gas pipeline monitoring device corresponding to the potential location to adjust the monitoring parameters of the gas pipeline monitoring device.

2. The method as described in claim 1, characterized in that, The determination of the actual leak point on the gas pipeline based on the gas sensor data and pipeline data includes: Based on historical gas sensor data, historical leak locations, historical weather data, and pipeline data at each point in the multiple sub-regions, identification features are determined. These identification features characterize the data type corresponding to the gas sensor data used to determine whether a gas pipeline leak has occurred. Based on the identification features, current gas sensor data, and historical gas sensor data, the actual location of the leak point and the corresponding leakage result for each leak point are determined. The leakage result includes whether a leak has occurred and the leakage intensity.

3. The method as described in claim 2, characterized in that, The method further includes: Based on historical maintenance records, current gas sensor data, historical potential hazard points, historical actual leak points, future weather data, and pipeline data, a hazard point prediction model is used to predict the identified potential hazard points and actual leak times at future time points. The hazard point prediction model is a machine learning model. The method for obtaining the hidden danger point prediction model includes: performing reinforcement training on the initial hidden danger point prediction model based on the target training samples corresponding to the target area to obtain the trained hidden danger point prediction model. Based on strong association rules and the current actual leakage points, a second preset algorithm is used to predict potential vulnerability points to be updated at future time points. The strong association rules are association rules that satisfy support and confidence conditions, and the second preset algorithm is the Apriori algorithm. Based on the identified potential hazard points and the potential hazard points to be updated, the updated potential hazard points are determined.

4. A smart gas pipeline zoned safety monitoring IoT system, characterized in that, The system includes a government regulatory management platform, a government regulatory sensor network platform, a government regulatory object platform, a gas company sensor network platform, and a gas equipment object platform that interact sequentially. The government regulatory management platform includes a government gas regulatory management platform and a government safety regulatory management platform. The government regulatory sensor network platform includes a government gas regulatory sensor network platform and a government safety regulatory sensor network platform. The government gas regulatory management platform is configured to perform the following operations: Based on the gas pipeline distribution data in the target area, the target area is divided into multiple sub-regions, including: Based on the gas pipeline distribution data, a target pipeline map is constructed; the gas pipeline distribution data includes the location data and pipeline number data of the gas pipelines. Based on the target pipeline map, a first preset algorithm is used to generate candidate sub-region partitioning parameters. Each candidate sub-region partitioning parameter corresponds to multiple groups obtained by partitioning the sub-pipes corresponding to the edges of the target pipeline map. The first preset algorithm includes at least one of the following: Randomly assign the sub-pipes corresponding to the edges of the target pipeline map to multiple groups to generate multiple candidate sub-region partitioning parameters; or The location area of ​​the gas pipeline is gridded according to preset parameters to obtain multiple sub-regions. Sub-pipes within the same sub-region are divided into the same group. Multiple candidate sub-region division parameters are generated by changing the preset parameters multiple times. The preset parameters include the shape, size and center point coordinates of the grid. Determine the data generation rate and data complexity of the multiple groups, and based on the data generation rate and data complexity, determine the fitness of the candidate sub-region partitioning parameters corresponding to the multiple groups; Based on the candidate sub-region partitioning parameters and their corresponding fitness, the selected pipeline partitioning parameters are determined; Based on the selected pipeline division parameters, the gas pipeline is divided into multiple sub-regions; Determine the management mapping relationship between the plurality of sub-regions and management companies, wherein the management mapping relationship includes the management company corresponding to each sub-region; and Based on the management mapping relationship, a routing instruction is generated and sent to the gas data forwarding device of the gas pipeline network to create the corresponding route; Through the government's gas regulatory sensor network platform, gas sensor data of each point in the multiple sub-areas on the gas pipeline within the target area is obtained from the management platforms of multiple gas companies. Based on the gas sensor data and pipeline data, the actual leak point on the gas pipeline is determined; Based on the gas sensor data and historical maintenance records, potential hidden danger points on the gas pipeline were identified. In response to the existence of the actual leak point: based on the actual location of the actual leak point on the gas pipeline, a first maintenance command and a first adjustment command are generated; the first maintenance command is sent to the gas company management platform corresponding to the actual location to instruct the gas company management platform to dispatch maintenance personnel to maintain the actual location; and the first adjustment command is sent to the gas pipeline control equipment corresponding to the actual location to adjust the operating parameters of the gas pipeline control equipment. In response to the existence of the potential hazard point: based on the potential location of the potential hazard point on the gas pipeline, a second maintenance instruction and a second adjustment instruction are generated; the second maintenance instruction is sent to the gas company management platform corresponding to the potential location to instruct the gas company management platform to dispatch maintenance personnel to inspect the potential location; and the second adjustment instruction is sent to the gas pipeline monitoring device corresponding to the potential location to adjust the monitoring parameters of the gas pipeline monitoring device.

5. The Internet of Things system as described in claim 4, characterized in that, The gas equipment object platform also includes a gas pipeline monitoring device, which is deployed in the gas pipeline in the target area and configured to monitor gas sensing data in the gas pipeline and upload the gas sensing data to the gas company's sensor network platform via the gas equipment object platform. The gas pipeline monitoring device includes at least a temperature sensor, a humidity sensor, a flow rate sensor, and a pressure sensor. The government-regulated platform includes at least the gas company management platform, which is configured as follows: Receive the gas sensing data uploaded by the gas company's sensor network platform, and upload the gas sensing data to the government regulatory sensor network platform when receiving the data acquisition instruction issued by the government regulatory sensor network platform. as well as The system receives the first maintenance instruction and / or the second maintenance instruction issued by the government regulatory management platform via the government regulatory sensor network platform, generates a dispatch task based on the first and / or the second maintenance instruction, and issues the dispatch task to the maintenance personnel to instruct the gas company management platform to dispatch the maintenance personnel to carry out maintenance on the actual location and / or the potential location.

6. The Internet of Things system as described in claim 4, characterized in that, The government gas regulatory management platform is further configured to perform the following operations: Based on historical gas sensor data, historical leak points, historical weather data, and pipeline data at each point in the multiple sub-regions, identification features are determined. These identification features characterize the data types corresponding to the gas sensor data used to determine whether the gas pipeline has leaked. as well as Based on the identification features, current gas sensor data, and historical gas sensor data, the actual location of the leak point and the corresponding leakage result for each leak point are determined. The leakage result includes whether a leak has occurred and the leakage intensity.

7. The Internet of Things system as described in claim 6, characterized in that, The government gas regulatory management platform is further configured to perform the following operations: Based on historical maintenance records, current gas sensor data, historical potential hazard points, historical actual leak points, future weather data, and pipeline data, a hazard point prediction model is used to predict the identified potential hazard points and actual leak times at future time points. The hazard point prediction model is a machine learning model. The method for obtaining the hidden danger point prediction model includes: performing reinforcement training on the initial hidden danger point prediction model based on the target training samples corresponding to the target area to obtain the trained hidden danger point prediction model. Based on strong association rules and the current actual leakage points, a second preset algorithm is used to predict potential vulnerability points to be updated at future time points. The strong association rules are association rules that satisfy support and confidence conditions, and the second preset algorithm is the Apriori algorithm. Based on the identified potential hazard points and the potential hazard points to be updated, the updated potential hazard points are determined.

Images