Intelligent Internet of Things System, Method and Storage Medium for Differential Pressure Safety Monitoring of Gas Pipelines

Abstract

The present invention provides an intelligent Internet of Things system, method and storage medium for differential pressure safety monitoring of gas pipelines. The Internet of Things system includes a gas company management platform and multiple other platforms. The method is executed by the gas company management platform and includes obtaining a differential pressure data matrix of the current gas pipeline network; determining an abnormal judgment result of the current gas pipeline network based on the differential pressure data matrix; in response to the existence of an abnormality, determining an abnormal probability distribution based on the differential pressure data matrix; generating an operation instruction set based on the abnormal probability distribution; generating a pressure regulating instruction based on the differential pressure data matrix, and the pressure regulating instruction is configured to control a pressure regulating device to regulate the pressure. The method can also be run after computer instructions stored in a computer-readable storage medium are read. The Internet of Things system and method can realize real-time and accurate monitoring and analysis of the differential pressure of gas pipelines, and ensure the safe and stable operation of the gas pipeline network.

Description

Technical Field

[0001] This specification relates to the field of pipeline monitoring, and particularly to an intelligent gas pipeline differential pressure safety monitoring Internet of Things system, method, and storage medium. Background Art

[0002] As an important infrastructure for modern energy transportation, gas pipelines bear the heavy responsibility of efficiently transporting natural gas. Monitoring the pressure changes in gas pipelines is crucial for the safety assessment of gas pipe networks. If the pressure differences between different gas pipelines at the same time and / or between different times of the same gas pipeline are abnormal, it may indicate problems such as gas leakage.

[0003] Currently, there are many ways to monitor differential pressure (such as using pressure sensors for monitoring), but there are lack of effective means for analyzing whether the differential pressure is abnormal and the causes of the differential pressure.

[0004] Therefore, providing an intelligent gas pipeline differential pressure safety monitoring Internet of Things system, method, and storage medium can achieve real-time and accurate monitoring and analysis of the differential pressure in gas pipelines, ensure the safe and stable operation of gas pipe networks, and guarantee the continuity and reliability of energy supply. Summary of the Invention

[0005] In order to solve problems such as the analysis of abnormal differential pressure and the determination of the causes of differential pressure, the present invention provides an intelligent gas pipeline differential pressure safety monitoring Internet of Things system, method, and storage medium.

[0006] The invention content includes an Internet of Things system for intelligent gas pipeline differential pressure safety monitoring. The system includes a government safety supervision management platform, a government safety supervision sensing network platform, a government safety supervision object platform, a gas company sensing network platform, a gas equipment object platform, and a gas maintenance object platform; the government safety supervision object platform includes a gas company management platform; the gas company management platform is configured to: obtain a differential pressure data matrix of the current gas pipeline network, the differential pressure data matrix includes differential pressure data of the gas pipelines in the current gas pipeline network at different time periods; based on the differential pressure data matrix, determine an abnormal judgment result of the current gas pipeline network; in response to the abnormal judgment result being abnormal, based on the differential pressure data matrix, determine the abnormal probability distribution of the current gas pipeline network; based on the abnormal probability distribution, generate an operation instruction set, the operation instruction set includes operation instructions corresponding to the gas pipelines, and the operation instructions include: the monitoring regulation instruction is configured to regulate the monitoring frequency of pipeline monitoring equipment; the movement control instruction is configured to instruct pipeline inspection equipment to inspect the target gas pipeline at a preset frequency; the reporting instruction is configured to receive and record the abnormal situation of the current gas pipeline network; the maintenance instruction is configured to allocate staff to maintain the current gas pipeline network; based on the differential pressure data matrix, generate a pressure regulating instruction, and the pressure regulating instruction is configured to control the pressure regulating equipment to regulate the abnormal gas pipelines in the current gas pipeline network.

[0007] The invention content includes a method for intelligent gas pipeline differential pressure safety monitoring. The method is executed by a gas company management platform. The method includes: obtaining a differential pressure data matrix of the current gas pipeline network, the differential pressure data matrix includes differential pressure data of the gas pipelines in the current gas pipeline network at different time periods; based on the differential pressure data matrix, determine an abnormal judgment result of the current gas pipeline network; in response to the abnormal judgment result being abnormal, based on the differential pressure data matrix, determine the abnormal probability distribution of the current gas pipeline network; based on the abnormal probability distribution, generate an operation instruction set, the operation instruction set includes operation instructions corresponding to the gas pipelines, and the operation instructions include: the monitoring regulation instruction is configured to regulate the monitoring frequency of pipeline monitoring equipment; the movement control instruction is configured to instruct pipeline inspection equipment to inspect the target gas pipeline at a preset frequency; the reporting instruction is configured to receive and record the abnormal situation of the current gas pipeline network; the maintenance instruction is configured to allocate staff to maintain the current gas pipeline network; based on the differential pressure data matrix, generate a pressure regulating instruction, and the pressure regulating instruction is configured to control the pressure regulating equipment to regulate the abnormal gas pipelines in the current gas pipeline network.

[0008] The invention content includes a computer-readable storage medium that stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes the intelligent gas pipeline differential pressure safety monitoring method.

[0009] The beneficial effects brought by the above invention content include but are not limited to: (1) By constructing a differential pressure data matrix from the differential pressure data of each gas pipeline at different times, it is possible to accurately determine whether there is an abnormality in the current gas pipeline network and accurately determine the abnormal probability distribution of the gas pipeline network, so as to further generate corresponding operation instructions and pressure regulation instructions, and be able to timely and accurately evaluate the abnormal situation of the pipeline network and perform targeted operations such as regulation, inspection, and maintenance to ensure the normal operation of the gas pipeline network; (2) According to the flow data and differential pressure data matrix of the current gas pipeline network, obtain the flow characteristic map and differential pressure characteristic map of the current gas pipeline network, and further determine the abnormal probability distribution of the current gas pipeline network, which can fully consider the gas flow situation and differential pressure situation in the gas pipeline, effectively avoid misjudgment of abnormal situations, and improve the recognition accuracy of abnormal gas pipelines; (3) Through the evaluation model, based on the pipeline position characteristics and abnormal probability distribution, determine the operation instruction set, and be able to accurately evaluate the operation instruction set using the learning ability of the machine learning model, thereby improving the accuracy of abnormal repair of gas pipelines. Brief Description of the Drawings

[0010] This specification will further illustrate in the form of exemplary embodiments, and these exemplary embodiments will be described in detail through the drawings. These embodiments are not restrictive. In these embodiments, the same numbers represent the same structures, where:

[0011] Figure 1 is a schematic diagram of the platform structure of the intelligent gas pipeline differential pressure safety monitoring Internet of Things system shown in some embodiments of this specification;

[0012] Figure 2 is an exemplary flowchart of the intelligent gas pipeline differential pressure safety monitoring method shown in some embodiments of this specification;

[0013] Figure 3 is an exemplary flowchart of determining the abnormal probability distribution shown in some embodiments of this specification;

[0014] Figure 4 is an exemplary flowchart of determining the operation instruction set shown in some embodiments of this specification;

[0015] Figure 5 is an exemplary schematic diagram of the evaluation model shown in some embodiments of this specification. Detailed Description of the Invention

[0016] The accompanying drawings required for the description of the embodiments will be briefly introduced below. The accompanying drawings do not represent all embodiments.

[0017] As used herein, "system", "device", "unit", and / or "module" are a way to distinguish different components, elements, parts, portions, or assemblies at different levels. If other words can achieve the same purpose, the said words can be replaced by other expressions.

[0018] Unless the context clearly indicates an exception, words such as "a", "an", "one", and / or "the" are not specifically singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of the clearly identified steps and elements, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0019] When describing the operations performed step by step in the embodiments of this specification, unless otherwise specified, the order of the steps can be adjusted, the steps can be omitted, and other steps can also be included during the operation process.

[0020] Figure 1 It is a schematic diagram of the platform structure of the intelligent gas pipeline differential pressure safety monitoring Internet of Things system shown in some embodiments of this specification.

[0021] In some embodiments, as Figure 1 shown, the intelligent gas pipeline differential pressure safety monitoring Internet of Things system 100 may include a government safety supervision management platform 110, a government safety supervision sensing network platform 120, a government safety supervision object platform 130, a gas company sensing network platform 140, a gas equipment object platform 150, and a gas maintenance object platform 160.

[0022] The government safety supervision management platform 110 refers to a platform for supervising and managing the gas pipeline network, which can be used to coordinate and cooperate among various functional platforms, and provide perception management and control management functions for the Internet of Things operation system.

[0023] The government safety supervision sensing network platform 120 refers to a functional platform for managing the sensing communication of the government, which can be configured as a communication network or a gateway, etc.

[0024] In some embodiments, the government safety supervision sensing network platform 120 can interact with the government safety supervision management platform 110 upward and with the government safety supervision object platform 130 downward. For example, the government safety supervision object platform 130 can send a reporting instruction to the government safety supervision management platform 110 through the government safety supervision sensing network platform 120.

[0025] The government security supervision object platform 130 refers to an object platform for generating perception information and executing control information.

[0026] In some embodiments, the government safety supervision object platform 130 may include a gas company management platform 131 .

[0027] The gas company management platform 131 refers to a comprehensive management platform for gas company related information, which can be used to manage gas companies and pipeline pressure difference safety monitoring related parameters.

[0028] In some embodiments, the gas company management platform 131 can be configured to: obtain the pressure difference data matrix of the current gas pipeline network; determine the abnormality judgment result of the current gas pipeline network based on the pressure difference data matrix; in response to the abnormality judgment result being that an abnormality exists, determine the abnormality probability distribution of the current gas pipeline network based on the pressure difference data matrix; generate an operation instruction set based on the abnormality probability distribution; and generate a pressure regulation instruction based on the pressure difference data matrix.

[0029] In some embodiments, the gas company management platform 131 may be further configured to: in response to the number of abnormal gas pipelines in the pressure difference data matrix exceeding a first preset threshold, determine that the abnormality judgment result is that an abnormality exists.

[0030] In some embodiments, the gas company management platform 131 can be further configured to: obtain the flow data of the current gas pipeline network; determine the flow characteristic map of the current gas pipeline network based on the flow data; determine the pressure difference characteristic map of the current gas pipeline network based on the pressure difference data matrix and the flow data; determine the abnormal probability distribution of the current gas pipeline network based on the flow characteristic map and the pressure difference characteristic map.

[0031] In some embodiments, the gas company management platform 131 can be further configured to: determine a first probability distribution based on a flow characteristic map; determine a second probability distribution based on a pressure difference characteristic map; and determine the abnormal probability distribution of the current gas pipeline network based on the first probability distribution and the second probability distribution.

[0032] In some embodiments, the gas company management platform 131 can be further configured to: generate multiple candidate instruction sets; for a candidate instruction set, determine the operating cost and operating efficiency of the candidate instruction set based on the pipeline location characteristics of the gas pipeline in the current gas pipeline network, the candidate instruction set and the abnormal probability distribution; determine the operating instruction set based on the operating cost and operating efficiency of the candidate instruction set.

[0033] In some embodiments, the gas company management platform 131 may be further configured to determine an operation instruction set through an evaluation model based on pipeline location characteristics and abnormal probability distribution of gas pipelines in the current gas pipeline network.

[0034] In some embodiments, the gas company management platform 131 may be further configured to: update the learning rate of the evaluation model based on an attenuation factor in response to the evaluation model being trained for a preset number of rounds.

[0035] In some embodiments, the gas company management platform 131 may further include a processor. In some embodiments, the processor may include one or more sub-processing devices (e.g., a single-core processing device or a multi-core and multi-chip processing device). By way of example only, the processor may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), etc., or any combination thereof.

[0036] The gas company sensing network platform 140 refers to an integrated management platform for the sensing information of the gas company and may be configured as a communication network or a gateway, etc., for realizing the functions of sensing communication of sensing information and sensing communication of control information.

[0037] In some embodiments, the gas company sensing network platform 140 may interact with the government safety supervision object platform 130 upward and interact with the gas equipment object platform 150 and the gas maintenance object platform 160 downward. For example, the government safety supervision object platform 130 may send the obtained monitoring and control instructions, motion control instructions, and pressure regulation instructions to the gas equipment object platform 150 through the gas company sensing network platform 140, or send the obtained reporting instructions and maintenance instructions to the gas maintenance object platform 160.

[0038] The gas equipment object platform 150 refers to a function platform that performs pipeline differential pressure monitoring, differential pressure regulation, and pipeline inspection and may include at least one pipeline monitoring device, at least one pipeline inspection device, and at least one pressure regulating device.

[0039] The pipeline monitoring device refers to a function device that monitors the differential pressure in the gas pipeline. For example, a pressure sensor, etc. In some embodiments, the pipeline monitoring device may be used to obtain the differential pressure data matrix of the gas pipeline network.

[0040] The pipeline inspection device refers to a function device that inspects inside the gas pipeline. For example, a pipeline crawling robot, etc. In some embodiments, the pipeline inspection device may be used to inspect the target gas pipeline at a preset frequency.

[0041] The pressure regulating device refers to a function device that regulates the pressure in the gas pipeline. For example, a gas pipeline pressure regulator, etc. In some embodiments, the pressure regulating device may be used to regulate the abnormal gas pipeline in the current gas pipeline network.

[0042] The gas maintenance object platform 160 refers to a platform that interacts with gas workers. Workers refer to those engaged in work related to the gas pipeline network. For example, gas pipeline network safety officers, repairmen, etc.

[0043] The gas maintenance object platform 160 may include at least one interaction device. For example, mobile phones, computers, etc. In some embodiments, the gas maintenance object platform may be used to receive maintenance instructions to deploy staff to maintain the current gas pipeline network.

[0044] For more information about the above platforms, please refer to Figures 2 - 5 and its related descriptions.

[0045] In some embodiments of this specification, based on the intelligent gas pipeline differential pressure safety monitoring Internet of Things system 100, an information operation closed-loop can be formed between each functional platform, and coordinated and regularly operated under the unified management of the gas company management platform, so as to realize the informatization and intelligence of intelligent gas pipeline differential pressure safety monitoring.

[0046] Figure 2 is an exemplary flowchart of the intelligent gas pipeline differential pressure safety monitoring method shown in some embodiments of this specification. As Figure 2 shown, process 200 includes the following steps. In some embodiments, process 200 can be implemented based on the intelligent gas pipeline differential pressure safety monitoring Internet of Things system 100 and executed by the gas company management platform 131. For example, executed by the processor in the gas company management platform 131.

[0047] For more information about the gas company management platform 131 and the processor, please refer to Figure 1 the relevant descriptions.

[0048] Step S210, obtain the differential pressure data matrix of the current gas pipeline network.

[0049] The current gas pipeline network refers to the gas pipeline network that needs to be monitored for differential pressure safety currently. The current gas pipeline network may include multiple gas pipelines.

[0050] The differential pressure data matrix refers to a matrix composed of differential pressure data. In some embodiments, the differential pressure data matrix may include the differential pressure data of each gas pipeline in the current gas pipeline network at different time periods.

[0051] Among them, different time periods can be default set by the processor or preset by technicians according to requirements. For example, multiple time periods within 24 hours.

[0052] The differential pressure data refers to data related to the difference in gas pipeline pressure. In some embodiments, the gas pipeline pressure can be monitored and obtained by pipeline monitoring equipment (such as pressure sensors).

[0053] In some embodiments, each time period may include multiple time points, and each adjacent pair of time points within a time period constitutes a sub-time period. For a gas pipeline, the differential pressure data for a time period may refer to the average of the differential pressure data for all sub-time periods within that time period. The differential pressure data for a sub-time period may refer to the gas pipeline pressure at the end time point of the sub-time period minus the gas pipeline pressure at the initial time point.

[0054] In some embodiments, the processor may obtain the differential pressure data of the gas pipelines in the current gas pipeline network at different time periods from the gas equipment object platform, and construct a differential pressure data matrix with the differential pressure data of a gas pipeline in a time period as an element, the differential pressure data of different gas pipelines in the same time period as a row vector, and the differential pressure data of the same gas pipeline in different time periods as a column vector. For example, the differential pressure data matrix may be expressed as , where represents the number of gas pipelines in the current gas pipeline network, represents the number of time periods, represents the gas pipeline at time period The differential pressure data , the row vector represents the differential pressure data of different gas pipelines (gas pipeline 1 - gas pipeline ) within time period , and the column vector represents the differential pressure data of gas pipeline at different time periods (time period 1 - time period ).

[0055] Step S220, based on the differential pressure data matrix, determine the abnormal judgment result of the current gas pipeline network.

[0056] The abnormal judgment result refers to the judgment result of whether the differential pressure data of the gas pipeline network is abnormal, and may include there being an abnormality and there being no abnormality.

[0057] In some embodiments, the processor may determine the abnormal judgment result of the gas pipeline network based on the differential pressure data matrix through various methods.

[0058] In some embodiments, in response to the presence of differential pressure data higher than a preset differential pressure threshold within the differential pressure data matrix, the processor may determine that the abnormal judgment result is there being an abnormality. Among them, the preset differential pressure threshold may refer to the maximum value of the differential pressure data when there is no abnormality, and may be set by default by the processor or set by a technician based on historical experience.

[0059] In some embodiments, in response to the number of abnormal gas pipelines within the differential pressure data matrix exceeding a first preset threshold, it is determined that the abnormal judgment result is there being an abnormality.

[0060] An abnormal gas pipeline refers to a gas pipeline with abnormal differential pressure data. For a gas pipeline, if its differential pressure data is abnormal in one or more time periods, it is an abnormal gas pipeline.

[0061] In some embodiments, the abnormal differential pressure data may refer to the differential pressure data exceeding the normal differential pressure range. Among them, the normal differential pressure range can be determined by the processor and / or technicians based on historical data statistics, and the normal differential pressure ranges of different gas pipelines in different time periods may be the same or different. For example, for (the differential pressure data of the gas pipeline in the time period ), in multiple historical monitors, taking the mean of multiple historical differential pressure data of the gas pipeline in the time period as the center of the range, the corresponding normal differential pressure range is determined according to the preset range size. The preset range size can be default set by the processor or set by the technician based on historical experience. Exemplarily, the center of the range is and the preset range size is , and the normal differential pressure range is .

[0062] The first preset threshold refers to the maximum value of the number of abnormal gas pipelines allowed to exist when there is no abnormality.

[0063] In some embodiments, the first preset threshold can be default set by the processor or set by the technician based on historical experience.

[0064] In some embodiments, the first preset threshold can also be related to the number of abnormalities of the current gas pipeline network in the historical time period.

[0065] The historical time period refers to the time period before the current time period. Among them, the current time period can be the time period in the differential pressure data matrix, and the historical time period can be the time period before the earliest time period in the differential pressure data matrix.

[0066] The number of abnormalities refers to the number of times when the abnormality judgment result of the gas pipeline network is abnormal. The number of abnormalities of the current gas pipeline network in the historical time period can be obtained by the processor or technician based on historical data statistics.

[0067] In some embodiments, the first preset threshold can be negatively correlated with the number of abnormalities. The more the number of abnormalities of the current gas pipeline network in the historical time period, the more unstable the current gas pipeline network is, and the corresponding first preset threshold can be smaller to ensure that the abnormality of the current gas pipeline network can be detected in time.

[0068] In some embodiments of the present specification, by determining whether the number of abnormal gas pipelines in the differential pressure data matrix exceeds a first preset threshold, the abnormal judgment result of the current gas pipeline network can be accurately determined, which is beneficial to timely discovering pipeline network abnormalities.

[0069] Step S230, in response to the abnormal judgment result indicating the existence of an abnormality, based on the differential pressure data matrix, determine the abnormal probability distribution of the current gas pipeline network.

[0070] The abnormal probability distribution refers to the distribution of abnormal situations existing in the current gas pipeline network. In some embodiments, the abnormal probability distribution may include multiple combinations of abnormal situations in the current gas pipeline network and their corresponding multiple probabilities.

[0071] Among them, an abnormal situation refers to an abnormal event that may occur during the transportation of gas in the pipeline network. For example, gas pipeline leakage, gas pipeline blockage, and related equipment failures. An abnormal situation combination refers to a combination of multiple abnormal gas pipelines and their corresponding multiple abnormal situations.

[0072] Only as an example, if the current gas pipeline network includes abnormal gas pipeline 1 and abnormal gas pipeline 2, and the abnormal situations include abnormal situation 1 and abnormal situation 2 (such as gas pipeline leakage and gas pipeline blockage), then the abnormal situation combinations may include , , , , representing abnormal gas pipeline 1 and abnormal gas pipeline 2 respectively, representing abnormal situation 1 and abnormal situation 2 respectively, representing 4 abnormal situation combinations respectively, representing that abnormal gas pipeline 1 has abnormal situation 1 and abnormal gas pipeline 2 has abnormal situation 1, and the rest are similar.

[0073] Correspondingly, the abnormal probability distribution can be expressed as , , , , ] representing the abnormal situation combinations occurring probabilities, can be the product of the probability of abnormal gas pipeline 1 having abnormal situation 1 and the probability of abnormal gas pipeline 2 having abnormal situation 1, and the rest are similar.

[0074] In some embodiments, the processor may determine the abnormal probability distribution through various methods based on the differential pressure data matrix. For example, the processor may determine the corresponding abnormal probability distribution by looking up a first preset table based on the differential pressure data matrix. The first preset table may include the correspondence between the differential pressure data matrix and the abnormal probability distribution. The first preset table may be constructed by the processor and / or technicians according to historical data.

[0075] In some embodiments, the processor may also obtain the flow data of the current gas pipeline network from the government safety supervision and management platform; determine the flow characteristic map of the current gas pipeline network based on the flow data; determine the differential pressure characteristic map of the current gas pipeline network based on the differential pressure data matrix and the flow data; and determine the abnormal probability distribution of the current gas pipeline network based on the flow characteristic map and the differential pressure characteristic map. For more content on this part, reference can be made to Figure 3 the relevant description.

[0076] Step S240: Generate an operation instruction set based on the abnormal probability distribution.

[0077] The operation instruction set refers to a set of operation instructions.

[0078] An operation instruction refers to an instruction for operating the differential pressure safety monitoring of a gas pipeline, which may include a monitoring and control instruction, a motion control instruction, a reporting instruction, and / or a maintenance instruction.

[0079] A monitoring and control instruction refers to an instruction for regulating a pipeline monitoring device. The monitoring and control instruction may be sent to the gas equipment object platform to regulate the monitoring frequency of the pipeline monitoring device.

[0080] In some embodiments, the monitoring and control instruction may also be used to add pipeline monitoring devices.

[0081] A motion control instruction refers to an instruction for controlling the movement of a pipeline inspection device. The motion control instruction may be sent to the gas equipment object platform to instruct the pipeline inspection device to inspect the target gas pipeline at a preset frequency.

[0082] The target gas pipeline refers to the gas pipeline that needs to be inspected. In some embodiments, the processor may determine the abnormal gas pipeline and the upstream and downstream gas pipelines connected to the abnormal gas pipeline as the target gas pipeline.

[0083] A reporting instruction refers to an instruction for conveying relevant information. The reporting instruction may be sent to the government safety supervision and management platform to receive and record the abnormal conditions of the current gas pipeline network.

[0084] A maintenance instruction refers to an instruction related to the maintenance of a gas pipeline. It may include the geographical location of an abnormal gas pipeline and the corresponding number of allocated staff. The maintenance instruction can be sent to a gas maintenance object platform to deploy staff to maintain the current gas pipeline network.

[0085] In some embodiments, the processor may generate an operation instruction set through various methods based on the abnormal probability distribution.

[0086] In some embodiments, the processor may determine the combination of abnormal situations with the highest occurrence probability based on the abnormal probability distribution, and determine the corresponding operation instruction set by querying a second preset table. The second preset table may include the corresponding relationship between the combination of abnormal situations and the operation instruction set. The second preset table may be pre-constructed by the processor and / or technicians based on historical data and / or historical experience.

[0087] In some embodiments, the processor may generate multiple candidate instruction sets; for a candidate instruction set, based on the pipeline location characteristics of the gas pipelines in the current gas pipeline network, the candidate instruction set, and the abnormal probability distribution, determine the operation cost and operation efficiency of the candidate instruction set; based on the operation cost and operation efficiency of the candidate instruction set, determine the operation instruction set. For more content on this part, reference can be made to Figure 4 and its related descriptions.

[0088] Step S250: Generate a pressure regulation instruction based on the differential pressure data matrix.

[0089] A pressure regulation instruction refers to an instruction for regulating the pressure of a gas pipeline. The pressure regulation instruction can be sent to a gas equipment object platform to control the pressure regulation equipment to regulate the pressure of the abnormal gas pipelines in the current gas pipeline network. For more content on the pressure regulation equipment, reference can be made to Figure 1 its related descriptions.

[0090] In some embodiments, the processor may determine the abnormal gas pipelines and the time period corresponding to the abnormal differential pressure data based on the differential pressure data matrix, and then generate the corresponding pressure regulation instruction to continuously control the differential pressure of the abnormal gas pipelines within the normal differential pressure range during the time period corresponding to the abnormal differential pressure data.

[0091] In some embodiments of this specification, a differential pressure data matrix is constructed through the differential pressure data of each gas pipeline at different time periods, thereby accurately determining whether there is an abnormality in the current gas pipeline network and accurately determining the abnormal probability distribution of the gas pipeline network, so as to further generate the corresponding operation instruction and pressure regulation instruction, and be able to timely and accurately evaluate the abnormal situation of the pipeline network and perform targeted operations such as regulation, inspection, and maintenance to ensure the normal operation of the gas pipeline network.

[0092] Figure 3is an exemplary flowchart for determining an abnormal probability distribution according to some embodiments of this specification. As Figure 3 shown, process 300 may include the following steps. In some embodiments, process 300 may be executed by the gas company management platform 131. For example, it is executed by a processor in the gas company management platform 131.

[0093] Step S310, obtain the flow rate data of the current gas pipeline network.

[0094] The flow rate data refers to the data related to the gas flow rate in the gas pipeline. For example, the flow rate data of the current gas pipeline network may include the average flow rate data of each gas pipeline at different time periods.

[0095] The average flow rate data refers to the average value of the gas flow rate in a gas pipeline within a time period. The average flow rate data of a gas pipeline within a time period can be represented by the average value of the instantaneous flow rate data at multiple time points within that time period. Among them, the instantaneous flow rate data may refer to the amount of gas passing through the cross-section of the gas pipeline at a certain moment, and can be obtained through a flow sensor (such as an ultrasonic flowmeter, a turbine flowmeter, etc.).

[0096] Exemplarily, the average flow rate data of the gas pipeline can be expressed as , representing the average flow rate data of the gas pipeline within the time period . Regarding can refer to the relevant description in the differential pressure data matrix. Figure 2

[0097] In some embodiments, the processor may obtain the flow rate data of each gas pipeline from the government safety supervision management platform through the government safety supervision sensing network platform.

[0098] Step S320, based on the flow rate data, determine the flow rate characteristic map of the current gas pipeline network.

[0099] The flow rate characteristic map refers to a map representing the relevant characteristics of the flow rate data of the current gas pipeline network, and can be composed of multiple nodes and multiple edges. In some embodiments, the processor may establish a flow rate characteristic map based on the connection relationship of each gas pipeline in the current gas pipeline network and the flow rate data of each gas pipeline.

[0100] Among them, the nodes of the flow rate characteristic map may be each gas pipeline in the current gas pipeline network, and one node of the flow rate characteristic map corresponds to one gas pipeline. The node characteristic of one node of the flow rate characteristic map may be the average flow rate data of the gas pipeline corresponding to that node at different time periods.

[0101] In some embodiments, if the gas pipelines corresponding to two nodes in the flow feature map are connected, there is an edge in the flow feature map between the two nodes. The edge in the flow feature map is a directed edge, and the direction of the edge can be from the upstream gas pipeline to the downstream gas pipeline.

[0102] The edge feature of the flow feature map can be the influence degree sequence of the upstream node corresponding to the edge on the downstream node.

[0103] The influence degree sequence refers to a sequence composed of multiple influence degrees corresponding to multiple time periods. The influence degree refers to the influence degree of the upstream node on the downstream node in a certain time period.

[0104] In some embodiments, for an edge in the flow feature map, the influence degree corresponding to a certain time period can be represented by the ratio of the average flow data of the upstream node of the edge in that time period to the sum of the average flow data of all upstream nodes of the downstream node of the edge.

[0105] For example, for the gas pipeline pointing to the gas pipeline , the gas pipeline is the upstream node, and the gas pipeline is the downstream node. There are 3 edges with the gas pipeline as the downstream node, and the upstream nodes corresponding to the other 2 edges are the gas pipelines and the gas pipeline , then the influence degree of the edge of the gas pipeline pointing to the gas pipeline in the time period can be , , , respectively represent the average flow data of the gas pipelines , , in the time period , and the gas pipelines , , are all the upstream gas pipelines of the gas pipeline .

[0106] Step S330, based on the pressure difference data matrix and the flow data, determine the pressure difference feature map of the current gas pipeline network.

[0107] The pressure difference feature map refers to a map that characterizes the relevant features of the pressure difference data of the gas pipelines in the current gas pipeline network, and can be composed of multiple nodes and multiple edges. In some embodiments, the processor can establish a pressure difference feature map based on the connection relationship of each gas pipeline in the current gas pipeline network, the flow data of the current gas pipeline network, and the pressure difference data matrix.

[0108] Among them, the nodes of the differential pressure characteristic map are the same as those of the flow characteristic map. The node characteristic of a node in the differential pressure characteristic map can be the differential pressure data of the corresponding gas pipeline at different time periods.

[0109] In some embodiments, the processor can directly extract the differential pressure data of the gas pipeline at different time periods from the differential pressure data matrix as the differential pressure node characteristics.

[0110] The edges of the differential pressure characteristic map are the same as those of the flow characteristic map. The edge characteristic of the differential pressure characteristic map can be the sensitivity sequence of the upstream node corresponding to the edge to the downstream node.

[0111] Among them, the sensitivity sequence refers to a sequence composed of multiple sensitivities corresponding to multiple time periods. Sensitivity refers to the sensitive degree of the downstream node to the change of the upstream node at a certain time period. The greater the sensitivity, the greater the influence of the flow of the upstream gas pipeline on the pressure of the downstream gas pipeline.

[0112] In some embodiments, for an edge of the differential pressure characteristic map, the sensitivity corresponding to a certain time period can be represented by the ratio of the differential pressure data of the downstream node of the edge at this time period to the average flow data of the upstream node of the edge at this time period.

[0113] For example, for the gas pipeline pointing to the gas pipeline , the gas pipeline is the upstream node, and the gas pipeline is the downstream node. Then the sensitivity of this edge in the time period can be , representing the average flow data of the gas pipeline in the time period , representing the differential pressure data of the gas pipeline in the time period .

[0114] Step S340, based on the flow characteristic map and the differential pressure characteristic map, determine the abnormal probability distribution of the current gas pipeline network.

[0115] In some embodiments, the processor can determine the abnormal probability distribution of the current gas pipeline network based on the flow characteristic map and the differential pressure characteristic map through an abnormal monitoring algorithm. Among them, the abnormal monitoring algorithm can include but is not limited to clustering algorithms, machine learning, density statistics, etc.

[0116] In some embodiments, the processor may determine a first probability distribution based on the flow feature map, determine a second probability distribution based on the differential pressure feature map, and determine the abnormal probability distribution of the current gas pipeline network based on the first probability distribution and the second probability distribution.

[0117] The first probability distribution refers to the abnormal probability distribution of the nodes in the flow feature map, that is, the distribution of abnormal conditions existing in the gas pipelines in the flow feature map. The first probability distribution may include multiple combinations of abnormal conditions in the flow feature map and their respectively corresponding multiple probabilities.

[0118] The combination of abnormal conditions in the flow feature map refers to the combination of multiple first abnormal nodes and their respectively corresponding multiple abnormal conditions. For more content about abnormal conditions and combinations of abnormal conditions, reference can be made to Figure 2 the relevant description.

[0119] Among them, the first abnormal node refers to the abnormal gas pipeline with abnormal conditions in the flow feature map. In some embodiments, the processor may determine the nodes in the flow feature map that meet at least two preset conditions in the first preset condition group as the first abnormal nodes.

[0120] The first preset condition group may correspond one-to-one with the abnormal conditions, and each abnormal condition has its own corresponding flow abnormal range.

[0121] The flow abnormal range refers to the range where the average flow data is not within the normal flow range, that is, the flow abnormal range is included in the complement of the normal flow range. The flow abnormal ranges of different abnormal conditions are different. The determination of the normal flow range can refer to the determination process of the normal differential pressure range above, which will not be elaborated here. The flow abnormal range can be determined by the processor and / or technicians based on historical data and / or historical experience.

[0122] In some embodiments, the first preset condition group may include a first preset condition, a second preset condition, and a third preset condition. For a node in an abnormal condition, the corresponding first preset condition may be that in a preset number of adjacent time periods, multiple average flow data of the node are all within the flow abnormal range corresponding to the abnormal condition; the second preset condition may be that in the most recent time period, the average flow data of at least one downstream node of the node is within the flow abnormal range corresponding to the abnormal condition; the third preset condition may be that the upstream node of the edge with the greatest influence corresponding to the node has an average flow data in the most recent time period within the flow abnormal range corresponding to the abnormal condition.

[0123] The adjacent time period refers to the time period adjacent to the current moment. For example, if the current time period is time period 10, the adjacent time periods may include time period 9, time period 8, etc. that are before time period 10 and adjacent to time period 10.

[0124] In some embodiments, the preset number of adjacent time periods in the first preset condition can be set by default by the processor or preset by a technician based on experience.

[0125] In some embodiments, the processor can determine the preset number of adjacent time periods in the first preset condition based on the operation instruction sequence of the gas pipeline corresponding to the node of the traffic feature map.

[0126] The operation instruction sequence refers to a sequence composed of multiple time periods and their corresponding operation instructions. In some embodiments, the processor can determine the operation instruction sequence based on historical data.

[0127] In some embodiments, for a node in the traffic feature map, the processor can determine the historical time period corresponding to the last execution of the motion control instruction and the historical time period corresponding to the last execution of the maintenance instruction based on the operation instruction sequence, determine the historical time period closest to the current time period between the two, and determine the number of time periods between it and the current time period as the preset number. For more content about operation instructions, motion control instructions, maintenance instructions, etc., reference can be made to [[ID= the relevant description.

[0128] In some embodiments of this specification, determining the preset number of adjacent time periods based on the operation instruction sequence can cover the time periods not monitored after the last execution of the operation instruction (including motion control instructions and / or maintenance instructions), avoid misjudgment due to lack of relevant data, and improve the judgment accuracy of the first abnormal node in the traffic feature map.

[0129] In some embodiments, a first abnormal node in the traffic feature map can have one or more abnormal situations at the same time. If a first abnormal node has one abnormal situation, the probability corresponding to this abnormal situation is 1. If a first abnormal node has multiple abnormal situations, the probability of each abnormal situation corresponding to this node is the same. For example, if there are 10 abnormal situations in Gas Pipeline 1 in the traffic feature map, the probability of each abnormal situation is 1 / 10.

[0130] Merely as an example, the traffic feature map includes a first abnormal node 1 and a first abnormal node 2 (e.g., abnormal Gas Pipeline 1 and abnormal Gas Pipeline 2). The abnormal situations of the first abnormal node 1 include abnormal situation 1 and abnormal situation 2, and the abnormal situation of the first abnormal node 2 includes abnormal situation 1. Then the abnormal situation combinations can include 、 , represent the first abnormal node 1 and the second abnormal node 2 respectively, represent abnormal situation 1 and abnormal situation 2 respectively, 、 represent 2 abnormal situation combinations respectively, It represents that there is an abnormal situation 1 for the first abnormal node 1 and an abnormal situation 1 for the first abnormal node 2, and the rest are similar.

[0131] Correspondingly, the first probability distribution can be expressed as , 、 respectively represent combinations of abnormal situations 、 The probabilities of occurrence, can be the product of the probability of the first abnormal node 1 having an abnormal situation 1 and the probability of the first abnormal node 2 sending an abnormal situation 1, and the rest are similar.

[0132] The second probability distribution refers to the abnormal probability distribution of the nodes in the pressure difference characteristic map, that is, the distribution of abnormal situations existing in the abnormal gas pipelines in the pressure difference characteristic map. In some embodiments, the second probability distribution may include multiple combinations of abnormal situations in the pressure difference characteristic map and their corresponding multiple probabilities.

[0133] The combination of abnormal situations in the pressure difference characteristic map refers to the combination of multiple second abnormal nodes and their corresponding multiple abnormal situations.

[0134] Among them, the second abnormal node refers to an abnormal gas pipeline with an abnormal situation in the pressure difference characteristic map. In some embodiments, the processor may determine, as the second abnormal node, a node in the pressure difference characteristic map that meets at least two preset conditions in the second preset condition group.

[0135] The second preset condition group may correspond one-to-one with the abnormal situations, and each abnormal situation has its corresponding pressure difference abnormal range. That is, one abnormal situation may correspond to a flow rate abnormal range and a pressure difference abnormal range.

[0136] The pressure difference abnormal range refers to the range where the pressure difference data is not within the normal pressure difference range, that is, the pressure difference abnormal range is included in the complement of the normal pressure difference range. The pressure difference abnormal ranges for different abnormal situations are different, and the pressure difference abnormal range can be determined by the processor and / or technical personnel based on historical data and / or historical experience. For more information about the normal pressure difference range and pressure difference data, reference can be made to ​ the relevant description.

[0137] In some embodiments, the second preset condition group may include a fourth preset condition, a fifth preset condition, and a sixth preset condition. For a node in an abnormal situation, the corresponding fourth preset condition may be that, in multiple adjacent time periods, the multiple differential pressure data of the node are all within the differential pressure abnormal range corresponding to the abnormal situation; the fifth preset condition may be that, in the most recent time period, the differential pressure data of at least one downstream node of the node are within the differential pressure abnormal range corresponding to the abnormal situation; the sixth preset condition may be that the upstream node of the edge with the maximum sensitivity corresponding to the node has differential pressure data within the differential pressure abnormal range corresponding to the abnormal situation in the most recent time period.

[0138] In some embodiments, a second abnormal node in the differential pressure feature map may also have one or more abnormal situations at the same time. Regarding the probability corresponding to the abnormal situation of the second abnormal node, reference may be made to the probability corresponding to the abnormal situation of the first abnormal node in the flow rate feature map, that is, the content for determining the second probability distribution based on the differential pressure feature map may refer to the relevant description for determining the first probability distribution based on the flow rate feature map. The principle is the same and will not be elaborated here.

[0139] In some embodiments, the processor may determine the abnormal probability distribution of the current gas pipeline network through weighted processing based on the first probability distribution and the second probability distribution.

[0140] In some embodiments, the weight magnitude relationship of the weighted processing may be determined based on the flow rate feature map and the differential pressure feature map. For example, for each time period, the processor may count the number of nodes with average flow rate data within the flow rate abnormal range in the flow rate feature map as the number of flow rate abnormal nodes in that time period, and then calculate the variance of the multiple numbers of flow rate abnormal nodes corresponding to each time period, denoted as the first variance; for each time period, the processor may count the number of nodes with differential pressure data within the differential pressure abnormal range in the differential pressure feature map as the number of differential pressure abnormal nodes in that time period, and then calculate the variance of the multiple numbers of differential pressure abnormal nodes corresponding to each time period, denoted as the second variance; the processor may determine the weight magnitude relationship of the first probability distribution and the second probability distribution in the weighted processing based on the magnitude relationship between the first variance and the second variance.

[0141] Exemplarily, if the first variance is greater than the second variance, it means that the number of differential pressure abnormal nodes corresponding to each time period fluctuates more than the number of flow rate abnormal nodes, then the weight of the first probability distribution may be appropriately increased, that is, the weight of the first probability distribution is greater than the weight of the second probability distribution; similarly, if the first variance is less than (or equal to) the second variance, then the weight of the first probability distribution is less than (or equal to) the weight of the second probability distribution.

[0142] In some embodiments of the present specification, the first probability distribution and the second probability distribution are respectively obtained based on the flow characteristic map and the pressure difference characteristic map, and accurate weights are assigned to the first probability distribution and the second probability distribution based on the magnitude relationship between the variances of the number of abnormal pressure difference nodes and the variances of the number of abnormal flow nodes corresponding to each time period, which is beneficial to accurately determining the abnormal probability distribution of the current gas pipeline network.

[0143] In some embodiments of the present specification, according to the flow data and the pressure difference data matrix of the current gas pipeline network, the flow characteristic map and the pressure difference characteristic map of the current gas pipeline network are obtained, and the abnormal probability distribution of the current gas pipeline network is further determined, which can fully consider the gas flow condition and the pressure difference condition in the gas pipeline, effectively avoid misjudgment of abnormal conditions, and improve the recognition accuracy of abnormal gas pipelines.

[0144] ​ is an exemplary flowchart for determining an operation instruction set shown in some embodiments of the present specification. As ​ shown, process 400 includes the following steps. In some embodiments, process 400 may be executed by the gas company management platform 131. For example, it is executed by a processor in the gas company management platform 131.

[0145] Step S410, generate a plurality of candidate instruction sets.

[0146] A candidate instruction set refers to a set composed of a plurality of candidate instructions, and the candidate instruction set can be used for a candidate operation instruction set.

[0147] In some embodiments, the processor may randomly form one or more candidate instruction sets based on one or more candidate instructions.

[0148] A candidate instruction refers to an operation instruction for candidate. A candidate instruction set may include one or more candidate instructions such as a candidate monitoring and control instruction, a candidate motion control instruction, a candidate maintenance instruction, a candidate reporting instruction, etc. In some embodiments, the candidate instruction may be randomly generated by the processor or set by a technician according to experience.

[0149] For more content about the above-mentioned multiple instructions, reference can be made to ​ the relevant description.

[0150] Step S420, for a candidate instruction set, determine the operation cost and operation efficiency of the candidate instruction set based on the pipeline location characteristics of the gas pipelines in the current gas pipeline network, the candidate instruction set, and the abnormal probability distribution.

[0151] The pipeline location characteristic refers to the location characteristic of the gas pipeline, which may include the geographical location of the current gas pipeline.

[0152] In some embodiments, the processor can directly retrieve the geographical location of the gas pipeline uploaded in advance.

[0153] The operating cost refers to the cost required to execute the candidate instruction set, which can include the total number of staff and / or working equipment, the total working time, the total distance between each gas pipeline and the work center, etc.

[0154] Among them, the staff can include pipeline maintenance personnel, and the working equipment can include pipeline inspection equipment. In some embodiments, the processor can determine the number of working equipment based on the motion control instruction and determine the allocated number of staff based on the maintenance instruction, so as to determine the total number of staff and / or working equipment.

[0155] The total working time can be the sum of the maintenance duration of the staff and / or the inspection duration of the working equipment. In some embodiments, the processor can statistically calculate the average value of multiple historical total working times and use it as the total working time.

[0156] The work center can be the gathering and distribution location of the staff and / or working equipment. The work center can include a maintenance center and an inspection center. The distance between the gas pipeline and the work center can be determined by the processor based on the geographical location of the work center and the pipeline location characteristics.

[0157] In some embodiments, the processor can determine the operating cost by weighted summation based on the total number of staff and / or working equipment, the total working time, and the total distance between each gas pipeline and the work center. Among them, the weighting factors can be default settings of the processor or preset by technicians according to experience.

[0158] The operating efficiency refers to the ability of the candidate instruction set to detect abnormalities and / or solve risks in the gas pipeline network.

[0159] In some embodiments, the processor can determine the operating efficiency based on the operating coefficient, the standard instruction set, and the candidate instruction set. Among them, for each combination of abnormal situations in the abnormal probability distribution, there is a corresponding standard instruction set. Both the standard instruction set and the operating coefficient can be default settings of the processor or preset by technicians according to experience.

[0160] Exemplarily, for a combination of abnormal situations, the processor can determine the gap between the standard instruction set and a certain candidate instruction set, and determine the ratio of the operating coefficient to the gap as the operating efficiency of the candidate instruction set.

[0161] In some embodiments, the gap between the standard instruction set and the candidate instruction set can be represented by a vector distance. For example, for a gas pipeline, the processor can calculate the vector distance between the standard operation instructions corresponding to the gas pipeline in the standard instruction set and the candidate operation instructions corresponding to the gas pipeline in the candidate instruction set, and statistically calculate the mean value of the respective vector distances corresponding to each gas pipeline in the current gas pipeline network, and use it as the gap between the standard instruction set and the candidate instruction set. Among them, the vector distance can refer to the distance between the standard vector corresponding to the standard operation instruction and the candidate vector corresponding to the candidate operation instruction. For example, the Euclidean distance, etc. The vector elements of the standard vector can include the monitoring frequency in the standard monitoring and control instructions and the standard quantity of additional pipeline monitoring devices, the preset frequency in the standard motion control instructions, the quantization value of the standard reporting instructions (quantized to 1 for having standard reporting instructions and quantized to 0 for having no standard reporting instructions), the geographical location of the gas pipeline and the allocated quantity of staff in the standard maintenance instructions, and the candidate vector is the same.

[0162] In some embodiments, for a combination of abnormal situations, the processor can determine the product of its corresponding probability and the operation efficiency, and determine the sum of the respective products corresponding to various combinations of abnormal situations included in the candidate instruction set as the operation efficiency corresponding to the candidate instruction set.

[0163] Step S430, determine the operation instruction set based on the operation cost and operation efficiency of the candidate instruction set.

[0164] In some embodiments, the processor can perform a weighted sum of the operation cost and operation efficiency of the candidate instruction set, and determine the candidate instruction set with the highest weighted sum value as the operation instruction set. Among them, the weight coefficient of the operation cost is negative, and the weights of both the operation cost and the operation efficiency can be default set by the processor or preset by technicians according to experience.

[0165] In some embodiments of this specification, by evaluating the operation cost and operation efficiency of the candidate instruction set, the operation instruction set can be accurately and quickly determined from the candidate instructions, which is beneficial to the safe operation of the gas pipeline network.

[0166] In some embodiments, the operation instruction set can further include the operation instruction sequences of each gas pipeline in the current gas pipeline network at different time periods, and the processor can also determine the operation instruction set through an evaluation model based on the pipeline position characteristics and abnormal probability distribution of the gas pipelines in the current gas pipeline network.

[0167] For more content about the operation instruction set, the current gas pipeline network, different time periods, operation instruction sequences, pipeline position characteristics, and abnormal probability distribution, reference can be made to ​ the relevant description.

[0168] An evaluation model refers to a model used to determine an operation instruction set. In some embodiments, the evaluation model can be a machine learning model. For example, a deep neural network (DNN), a graph neural network (GNN), etc.

[0169] In some embodiments, as ​ shown, the input of the evaluation model 550 can include the pipeline location feature 510 and the abnormal probability distribution 520, and the output can be the operation instruction set 560.

[0170] In some embodiments, the evaluation model can be obtained through training in various ways. For example, the evaluation model can be obtained by training with multiple training samples with training labels, etc. A set of training samples for training the evaluation model can include the sample pipeline location features of the sample gas pipelines and the sample abnormal conditions. The training label corresponding to a set of training samples is the actual operation instruction set with the best operation effect.

[0171] In some embodiments, the processor can obtain multiple sets of training samples based on historical data, where the historical pipeline location features and historical abnormal conditions of each historical gas pipeline in a historical gas pipeline network can be used as a set of training samples; construct multiple historical vectors based on the multiple sets of training samples, and the historical vectors include historical pipeline location features and historical abnormal conditions; cluster the multiple historical vectors to determine one or more clusters; for each cluster, the processor can use the historical actual operation instruction set with the best operation effect among the multiple historical actual operation instruction sets corresponding to the multiple sets of training samples in the cluster as the training label corresponding to all the training samples in the cluster. Among them, the clustering methods include but are not limited to K-means clustering, mean shift clustering, etc. The best operation effect can mean that within multiple time periods after executing the historical actual operation instruction set, the historical differential pressure data of each historical gas pipeline in the corresponding historical gas pipeline network are all within the normal differential pressure range, and the fluctuation of the historical differential pressure data is the smallest.

[0172] For more information about abnormal conditions, differential pressure data, and normal differential pressure range, please refer to ​ the relevant description.

[0173] In some embodiments, the processor may perform multiple rounds of iterative training on the initial evaluation model based on multiple sets of training samples with training labels (including historical sample pipeline location features and historical sample anomalies). One round of iterative training includes: inputting a set of training samples with training labels into the initial evaluation model, determining a loss function value based on the training labels and the output of the initial evaluation model, and iteratively updating the parameters of the initial evaluation model based on the loss function value. Multiple rounds of iterations are performed until an iteration condition is met, completing the training of the initial evaluation model and obtaining a trained evaluation model. The iteration method may include, but is not limited to, gradient descent, and the iteration condition may include convergence of the loss function and reaching a preset number of iterations.

[0174] In some embodiments, as ​ As shown, the input of the evaluation model 550 may further include a flow characteristic map 530 and a pressure difference characteristic map 540 .

[0175] For more information about flow rate characteristic graphs and pressure difference characteristic graphs, please refer to ​ Related description.

[0176] In some embodiments, the training samples for training the evaluation model may further include a sample flow characteristic map and a sample pressure difference characteristic map, and the sample flow characteristic map and the sample pressure difference characteristic map may be acquired based on historical data.

[0177] In some embodiments, the processor can train the evaluation model based on training samples including sample flow characteristic maps, sample pressure difference characteristic maps, sample pipeline position characteristics, and sample abnormality probability distribution. For the training method of the evaluation model, please refer to the relevant description above.

[0178] In some embodiments of the present specification, using the flow characteristic map and the pressure difference characteristic map as inputs of the evaluation model can improve the accuracy of the evaluation model in determining the operating instruction set.

[0179] In some embodiments, in response to the evaluation model being trained for a preset number of rounds, the processor may update the learning rate of the evaluation model based on a decay factor.

[0180] The preset number of rounds refers to the pre-set number of training rounds, for example, the preset number of iterations.

[0181] In some embodiments, the processor may determine the preset number of rounds based on the complexity of the pressure difference characteristic map.

[0182] Since the higher the complexity of the differential pressure characteristic map, the more complex the connection mode of the gas pipelines in the current gas pipeline network, the higher the potential instability in the current gas pipeline network, and the greater the amount of information included in the training samples. Therefore, the complexity of the differential pressure characteristic map can have a positive correlation with the preset number of rounds, that is, the higher the complexity of the differential pressure characteristic map, the larger the preset number of rounds, so as to delay the time of learning rate decay and enable the evaluation model to accurately learn the data characteristics of the training samples.

[0183] The complexity of the differential pressure characteristic map refers to the degree of complexity of the differential pressure characteristic map. The complexity of the differential pressure characteristic map can be represented by a value from 0 to 10, and the larger the value, the higher the complexity.

[0184] In some embodiments, the processor can determine the in-degree and out-degree of the pipelines of each node in the differential pressure characteristic map, further calculate the variance of the in-degree of all nodes and the variance of the out-degree of all nodes, and determine the mean value of the above two variances as the complexity of the differential pressure characteristic map. Among them, the in-degree of the pipeline refers to the number of pipelines flowing into the node, and the out-degree of the pipeline refers to the number of pipelines flowing out of the node. The in-degree and out-degree of the pipeline can be obtained by the processor based on the differential pressure characteristic map.

[0185] The decay factor is a parameter that decays the learning rate of the evaluation model and can be represented by a value between 0 and 1. In some embodiments, the decay factor can be default set by the processor and / or preset by the technician based on experience.

[0186] The learning rate is a parameter that controls the update of the evaluation model. For example, the search step size of the gradient descent method in iterative training. In some embodiments, in response to the evaluation model being trained for a preset number of rounds, the processor can update the learning rate of the evaluation model based on the decay factor. For example, multiplying the learning rate by the decay factor to update it to a new learning rate.

[0187] In some embodiments of this specification, determining the preset number of rounds according to the complexity of the differential pressure characteristic map and updating the learning rate based on the decay factor after being trained for the preset number of rounds can help the model better converge to the optimal solution, avoid oscillations or non-convergence during the model training process, and enable the model to be fully learned, which is beneficial to accurately obtaining the operation instruction set.

[0188] In some embodiments of this specification, by using the evaluation model, based on the pipeline position characteristics and the abnormal probability distribution, to determine the operation instruction set, it can accurately evaluate the operation instruction set by using the learning ability of the machine learning model, thereby improving the accuracy of abnormal maintenance of gas pipelines.

[0189] One or more embodiments of this specification provide an intelligent gas pipeline differential pressure safety monitoring device, which includes at least one memory and at least one processor. The at least one memory is used to store computer instructions, and the at least one processor executes the computer instructions or partial instructions to implement an intelligent gas pipeline differential pressure safety monitoring method.

[0190] One or more embodiments of this specification provide a computer-readable storage medium that stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes an intelligent gas pipeline differential pressure safety monitoring method.

[0191] The embodiments in the present invention are only for illustration and explanation, and do not limit the scope of application of the present invention. For those skilled in the art, various modifications and changes that can be made under the guidance of the present invention are still within the scope of the present invention. In addition, certain features, structures, or characteristics in one or more embodiments of the present invention can be appropriately combined.

[0192] If there are inconsistencies or conflicts between the descriptions, definitions, and / or uses of terms in the materials cited in the present invention and the content described in the present invention, the descriptions, definitions, and / or uses of terms in the present invention shall prevail.

[0193] Finally, it should be understood that the embodiments described in the present invention are only used to illustrate the principles of the embodiments of the present invention. Other variations may also fall within the scope of the present invention. Therefore, by way of example and not limitation, alternative configurations of the embodiments of the present invention may be regarded as consistent with the teachings of the present invention. Accordingly, the embodiments of the present invention are not limited to the embodiments explicitly presented and described in the present invention.

Claims

1. An Internet of Things system for intelligent safety monitoring of pressure difference in gas pipelines, characterized in that, The system includes a gas company management platform; The gas company management platform is configured to: Obtain a differential pressure data matrix of the current gas pipeline network, where the differential pressure data matrix includes differential pressure data of gas pipelines in the current gas pipeline network at different time periods; Based on the differential pressure data matrix, determine an abnormal judgment result of the current gas pipeline network; In response to the abnormal judgment result indicating an abnormality, Obtain the flow data of the current gas pipeline network; Based on the flow data, determine a flow characteristic map of the current gas pipeline network; Based on the differential pressure data matrix and the flow data, determine a differential pressure characteristic map of the current gas pipeline network; Based on the flow characteristic map and the differential pressure characteristic map, determine an abnormal probability distribution of the current gas pipeline network; Based on the abnormal probability distribution, generate an operation instruction set, where the operation instruction set includes operation instructions corresponding to the gas pipelines, and the operation instructions include: A monitoring and control instruction, which is configured to control the monitoring frequency of pipeline monitoring equipment; A motion control instruction, which is configured to instruct pipeline inspection equipment to inspect a target gas pipeline at a preset frequency; A reporting instruction, which is configured to receive and record the abnormal conditions of the current gas pipeline network; A maintenance instruction, which is configured to allocate staff to maintain the current gas pipeline network; Based on the differential pressure data matrix, generate a pressure regulation instruction, which is configured to control pressure regulation equipment to regulate the pressure of abnormal gas pipelines in the current gas pipeline network.

2. The Internet of Things system according to claim 1, wherein The gas company management platform is further configured to: Generate multiple candidate instruction sets; For one of the candidate instruction sets, based on the pipeline position characteristics of the gas pipelines in the current gas pipeline network, the candidate instruction set, and the abnormal probability distribution, determine the operation cost and operation efficiency of the candidate instruction set; Based on the operation cost and operation efficiency of the candidate instruction set, determine the operation instruction set.

3. The Internet of Things system according to claim 1, characterized in that, The operation instruction set further includes an operation instruction sequence of the gas pipelines in the current gas pipeline network at different time periods, and the gas company management platform is further configured to: Based on the pipeline position characteristics of the gas pipelines in the current gas pipeline network and the abnormal probability distribution, determine the operation instruction set through an evaluation model, where the evaluation model is a machine learning model.

4. The Internet of Things system according to claim 1, wherein The system further includes a government safety supervision management platform, a government safety supervision sensor network platform, a government safety supervision object platform, a gas company sensor network platform, a gas equipment object platform, and a gas maintenance object platform; where the government safety supervision object platform includes the gas company management platform; The government safety supervision management platform is communicatively connected to the government safety supervision object platform through the government safety supervision sensor network platform; The government safety supervision object platform is communicatively connected to the gas equipment object platform and the gas maintenance object platform through the gas company sensor network platform; The monitoring and control instruction is sent to the gas equipment object platform, the motion control instruction is sent to the gas equipment object platform, the reporting instruction is sent to the government safety supervision and management platform, the maintenance instruction is sent to the gas maintenance object platform, and the pressure regulating instruction is sent to the gas equipment object platform.

5. A method for intelligent differential pressure safety monitoring of gas pipelines, characterized in that, The method is executed by a gas company management platform, and the method includes: Obtain the differential pressure data matrix of the current gas pipeline network, where the differential pressure data matrix includes the differential pressure data of the gas pipelines in the current gas pipeline network at different time periods; Based on the differential pressure data matrix, determine the abnormal judgment result of the current gas pipeline network; In response to the abnormal judgment result indicating an abnormality, Obtain the flow data of the current gas pipeline network; Based on the flow data, determine the flow characteristic map of the current gas pipeline network; Based on the differential pressure data matrix and the flow data, determine the differential pressure characteristic map of the current gas pipeline network; Based on the flow characteristic map and the differential pressure characteristic map, determine the abnormal probability distribution of the current gas pipeline network; Based on the abnormal probability distribution, generate an operation instruction set, where the operation instruction set includes operation instructions corresponding to the gas pipelines, and the operation instructions include: A monitoring and control instruction, configured to regulate the monitoring frequency of pipeline monitoring equipment; A motion control instruction, configured to instruct pipeline inspection equipment to inspect the target gas pipeline at a preset frequency; A reporting instruction, configured to receive and record the abnormal conditions of the current gas pipeline network; A maintenance instruction, configured to allocate staff to maintain the current gas pipeline network; Based on the differential pressure data matrix, generate a pressure regulating instruction, where the pressure regulating instruction is configured to control a pressure regulating device to regulate the pressure of the abnormal gas pipelines in the current gas pipeline network.

6. The method according to claim 5, characterized in that The generating the operation instruction set based on the abnormal probability distribution includes: Generate a plurality of candidate instruction sets; For one of the candidate instruction sets, based on the pipeline position characteristics of the gas pipelines in the current gas pipeline network, the candidate instruction set, and the abnormal probability distribution, determine the operation cost and operation efficiency of the candidate instruction set; Based on the operation cost and the operation efficiency of the candidate instruction set, determine the operation instruction set.

7. The method according to claim 5, characterized in that, The operation instruction set further includes an operation instruction sequence of the gas pipelines in the current gas pipeline network at different time periods, and the generating the operation instruction set based on the abnormal probability distribution further includes: Based on the pipeline position characteristics of the gas pipelines in the current gas pipeline network and the abnormal probability distribution, determine the operation instruction set through an evaluation model, where the evaluation model is a machine learning model.

8. A computer-readable storage medium, characterized in that, The storage medium stores computer instructions, and when a computer reads the computer instructions in the storage medium, the computer executes the intelligent gas pipeline differential pressure safety monitoring method according to any one of claims 5-7.

Images