Natural gas pipeline network simulation prediction method, device and equipment and storage medium

Abstract

The invention discloses a natural gas pipeline network simulation prediction method and device, equipment and a storage medium. The method comprises the steps of determining parameter dynamic data of a plurality of nodes in a target natural gas pipeline network at the current moment based on a mechanism simulation model according to historical system data and current system data; determining boundary node prediction data of a pipe network boundary node layer in the target natural gas pipe network according to the current structure data, preset system data and parameter dynamic data of the target natural gas pipe network based on a graph neural network model and a mechanism simulation model; on the basis of a neural operator model and a mechanism simulation model, according to the parameter dynamic data and the candidate node prediction data, determining pipe section node prediction data of an internal node layer of the pipe section; and summarizing the boundary node prediction data and the pipe section node prediction data to obtain pipe network simulation prediction data of the target natural gas pipe network in a future preset time period. According to the technical scheme, the online simulation efficiency and stability of the pipe network can be improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of data simulation technology, in particular to the field of natural gas data simulation technology, and specifically to a natural gas pipeline network simulation prediction method, device, equipment and storage medium. Background Art

[0002] The natural gas pipeline network system is an energy artery. People's daily life and normal industrial operation are inseparable from the normal operation of natural gas pipelines. However, the current design planning and daily operation and maintenance of natural gas pipelines rely on simulation engines. In order to efficiently carry out the operation and maintenance of natural gas pipelines, it is key to establish accurate and fast simulation algorithms.

[0003] At present, the traditional steady-state simulation engine of natural gas pipeline network based on numerical solution faces high performance requirements, especially in terms of speed. The existing simulation engine is difficult to meet the real-time online simulation requirements of large pipeline networks with a calculation frequency of 3 minutes during daily operation and maintenance. In addition, as the complexity of operation and maintenance gradually increases, the online simulation frequency in the future is expected to reach the 1-minute level. Therefore, a faster simulation engine alternative technology is needed, among which the rapidity of deep learning technology is favored by people.

[0004] However, although data-driven technologies represented by deep learning have advantages such as fast computing speed, due to the lack of mechanism, simple data models are prone to produce prediction results that violate reality and have poor stability. Summary of the invention

[0005] The present application provides a natural gas pipeline network simulation prediction method, device, equipment and storage medium to improve the efficiency and stability of online pipeline network simulation.

[0006] According to one aspect of the present application, a natural gas pipeline network simulation prediction method is provided, the method comprising:

[0007] Based on the mechanism simulation model, the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment are determined according to the historical system data and the current system data; the historical system data refers to the system data of the historical time period from the candidate historical moment to the current moment obtained from the intermediate database at a preset frequency; the current system data refers to the system data at the current moment obtained from the intermediate database at a preset frequency;

[0008] Based on the graph neural network model and the mechanism simulation model, the boundary node prediction data of the boundary node layer of the target natural gas pipeline network is determined according to the current structure data of the target natural gas pipeline network, the preset system data and the parameter dynamic data; the preset system data refers to the preset values ​​of the gas source pressure and user flow in the target natural gas pipeline network within a preset time period in the future;

[0009] Based on the neural operator model and the mechanism simulation model, the pipe segment node prediction data of the internal node layer of the pipe segment is determined according to the parameter dynamic data and the candidate node prediction data; the candidate node prediction data is the prediction value of the nodes at both ends of each pipe segment in the target natural gas pipeline network screened from the boundary node prediction data;

[0010] The boundary node prediction data and the pipe section node prediction data are aggregated to obtain the pipe network simulation prediction data of the target natural gas pipe network within the future preset time period.

[0011] According to another aspect of the present application, a natural gas pipeline network simulation prediction device is provided, the device comprising:

[0012] A data determination module is used to determine the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment based on the mechanism simulation model, according to the historical system data and the current system data; the historical system data refers to the system data of the historical time period from the candidate historical moment to the current moment obtained from the intermediate database at a preset frequency; the current system data refers to the system data at the current moment obtained from the intermediate database at a preset frequency;

[0013] A boundary node simulation module is used to determine the boundary node prediction data of the boundary node layer of the target natural gas pipeline network based on the graph neural network model and the mechanism simulation model according to the current structure data of the target natural gas pipeline network, the preset system data and the parameter dynamic data; the preset system data refers to the preset values ​​of the gas source pressure and user flow in the target natural gas pipeline network within a preset time period in the future;

[0014] A pipe segment node simulation module, for determining the pipe segment node prediction data of the internal node layer of the pipe segment based on the neural operator model and the mechanism simulation model according to the parameter dynamic data and the candidate node prediction data; the candidate node prediction data is the prediction value of the nodes at both ends of each pipe segment in the target natural gas pipeline network screened from the boundary node prediction data;

[0015] The data aggregation module is used to aggregate the boundary node prediction data and the pipe segment node prediction data to obtain the pipeline network simulation prediction data of the target natural gas pipeline network within the future preset time period.

[0016] According to another aspect of the present application, an electronic device is provided, the electronic device comprising:

[0017] one or more processors;

[0018] A memory for storing one or more programs;

[0019] When the one or more programs are executed by the one or more processors, the one or more processors implement any one of the natural gas pipeline network simulation prediction methods provided in the embodiments of the present application.

[0020] According to another aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the program is executed by a processor, any one of the natural gas pipeline network simulation prediction methods provided in the embodiments of the present application is implemented.

[0021] According to another aspect of the present application, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the computer program implements any one of the natural gas pipeline network simulation prediction methods provided in the embodiments of the present application.

[0022] The present application determines the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment based on the mechanism simulation model, historical system data and current system data; historical system data refers to the system data of the historical time period from the candidate historical moment to the current moment obtained from the intermediate database at a preset frequency; current system data refers to the system data of the current moment obtained from the intermediate database at a preset frequency; based on the graph neural network model and the mechanism simulation model, the boundary node prediction data of the pipeline boundary node layer in the target natural gas pipeline network is determined according to the current structure data, preset system data and parameter dynamic data of the target natural gas pipeline network; the preset system data refers to the preset values ​​of the gas source pressure and user flow in the target natural gas pipeline network within a preset time period in the future; based on the neural operator model and the mechanism simulation model, the pipe segment node prediction data of the internal node layer of the pipe segment is determined according to the parameter dynamic data and the candidate node prediction data; the candidate node prediction data is the prediction value of the nodes at both ends of each pipe segment in the target natural gas pipeline network screened out from the boundary node prediction data; the boundary node prediction data and the pipe segment node prediction data are summarized to obtain the pipeline simulation prediction data of the target natural gas pipeline network within the preset time period in the future. The above-mentioned technical scheme, based on the combination of mechanism model and GNN (Graph Neural Network) data model, has the effect of improving the speed, accuracy and stability of simulation calculation. By integrating the traditional mechanism model with modern data-driven technology, this method can significantly speed up the calculation process while ensuring the prediction accuracy, and meet the real-time and frequent online simulation needs. In addition, the data model effectively improves the prediction ability of future states, while the mechanism model helps to correct the prediction deviation that may occur in the data model, ensuring the reliability and consistency of the simulation results. Through these technical means, the present invention realizes high-speed and stable online simulation of natural gas pipeline networks, which can meet the high-frequency computing needs of real-time operation and maintenance, and has good economy and practicality, and has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is a flow chart of a natural gas pipeline network simulation prediction method provided according to Example 1 of the present application;

[0024] Figure 2 This is a flow chart of a natural gas pipeline network simulation prediction method provided according to Embodiment 2 of the present application;

[0025] Figure 3 It is a structural schematic diagram of a natural gas pipeline network simulation prediction device provided according to the third embodiment of the present application;

[0026] Figure 4 It is a structural schematic diagram of an electronic device for implementing the natural gas pipeline network simulation prediction method of an embodiment of the present application. DETAILED DESCRIPTION

[0027] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0028] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] In addition, it should be noted that the collection, storage, use, processing, transmission, provision and disclosure of relevant data such as historical system data and current system data involved in the technical solution of this application are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0030] Embodiment 1

[0031] Figure 1This is a flowchart of a natural gas pipeline network simulation prediction method provided in accordance with the first embodiment of the present application. This embodiment is applicable to the case of online simulation of hydraulic and thermal parameter data of a natural gas pipeline network, and can be executed by a natural gas pipeline network simulation prediction device. The natural gas pipeline network simulation prediction device can be implemented in the form of hardware and / or software, and the natural gas pipeline network simulation prediction device can be configured in a computer device, such as a server. Figure 1 As shown, the method includes:

[0032] S110. Based on the mechanism simulation model, according to the historical system data and the current system data, determine the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment.

[0033] In this embodiment, the mechanism simulation model refers to a simulation model constructed based on physical principles and mathematical equations (such as fluid mechanics, thermodynamic equations, etc.), which is used to describe and predict the behavior of the natural gas pipeline network under different working conditions; the model takes into account multiple factors such as natural gas flow, pressure change, flow rate, etc., and simulates the operation status of the natural gas pipeline network through calculation. Historical system data refers to system data of the historical time period from the candidate historical moment to the current moment obtained from the intermediate database at a preset frequency; the data includes system operation data from a certain candidate historical moment to the current moment, and the data may include pressure data, flow data and temperature data of points such as gas sources, users, and compressor stations. Current system data refers to system data at the current moment obtained from the intermediate database at a preset frequency; it represents the operating status and data of the natural gas pipeline network at the current moment; the data may include pressure data, flow data and temperature data of points such as gas sources, users, and compressor stations. The target natural gas pipeline network refers to the natural gas pipeline network currently being studied or simulated and predicted; the structure, nodes, flow rate and pressure of the pipeline network are considered in the prediction and used to generate the future operation status of the pipeline network. Parameter dynamic data refers to the changes in system parameters at a certain moment or time period obtained through dynamic simulation calculation; these data reflect the operating status of the system at different moments, such as the instantaneous flow and pressure of each node in the natural gas pipeline network; the parameter dynamic data may include the pressure, flow and temperature of multiple nodes in the target natural gas pipeline network.

[0034] Optionally, based on the mechanism simulation model, the historical system data corresponding to the candidate historical moment is used as the boundary value, and steady-state simulation calculations are performed on multiple nodes in the target natural gas pipeline network to obtain parameter steady-state data of multiple nodes in the target natural gas pipeline network at the candidate historical moment; based on the mechanism simulation model, the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment are determined according to the candidate system data and the parameter steady-state data.

[0035] In this embodiment, the boundary value refers to the external or initial conditions of the system set in the simulation model to limit the solution space of the model; in dynamic simulation, the boundary value usually represents the known parameters of the system at a certain moment (such as the input pressure and flow of the pipeline network), which provides preliminary constraints for the simulation process. Steady-state simulation calculation refers to the state of the simulation system when it reaches equilibrium without considering the change of time. Under steady-state conditions, the various parameters of the system (such as pressure, flow, etc.) remain constant; through steady-state simulation, the static parameter data of the pipeline network at a certain moment can be obtained. Parameter steady-state data refers to the values ​​of the state parameters (such as pressure, flow, temperature, etc.) of each node in the system under stable conditions during the steady-state simulation calculation process; these parameters refer to the performance of the system when it is in equilibrium at a given historical moment or boundary conditions. Candidate system data refers to the historical system data corresponding to the time period from the next moment of the candidate historical moment to the current moment.

[0036] Furthermore, based on the mechanism simulation model, according to the candidate system data and the parameter steady-state data, determining the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment can be, based on the mechanism simulation model, using the candidate system data as the boundary value, and the parameter steady-state data as the initial value, performing dynamic simulation calculations on multiple nodes in the target natural gas pipeline network to obtain the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment.

[0037] In this embodiment, the initial value refers to the initial state data of each node in the system when the simulation calculation starts; these initial states are usually obtained by historical data, experimental data or theoretical model predictions, and provide a starting reference for dynamic simulation; in this situation, steady-state data usually serves as the initial value.

[0038] For example, the intermediate database is accessed to obtain the SCADA (Supervisory Control and Data) of the natural gas pipeline network at a frequency of 3 minutes. Acquisition, monitoring and data acquisition) system data, the data includes the SCADA system data at the current time t and the SCADA system data of a historical period of time (tk, t); taking the pressure of the pipeline gas source at time tk and the user's flow as boundary values, a simulation model based on a mechanism model is used to perform steady-state simulation calculations to obtain the steady-state values ​​of hydraulic and thermal parameters of all nodes of the pipeline network at time tk; taking the pressure of the pipeline gas source at time {t-k+1, t-k+2, ..., t} and the user's flow as boundary values, and the steady-state initial values ​​of all nodes of the pipeline network at time tk as initial values, a simulation model based on the mechanism model is used to perform dynamic simulation calculations to obtain the online simulation initial values ​​of the pipeline network at time t under real operation conditions, including the dynamic values ​​of hydraulic and thermal parameters of all nodes; wherein the dynamic values ​​of hydraulic and thermal parameters of all nodes may be the pressure, flow and temperature of the gas source node, the user node, the valve room inlet and outlet node, the compressor station inlet and outlet node, the sub-transmission station inlet and outlet node, and the internal nodes of the pipeline section.

[0039] In an optional implementation, after obtaining the historical system data and the current system data, a wavelet threshold denoising method may be used to filter the historical system data and the current system data to obtain filtered historical system data and the current system data.

[0040] In this embodiment, the wavelet threshold denoising method refers to a signal denoising method based on wavelet transform; it decomposes the signal in the wavelet domain, removes the noise component, and then reconstructs the denoised signal.

[0041] S120. Based on the graph neural network model and the mechanism simulation model, according to the current structure data, preset system data and parameter dynamic data of the target natural gas pipeline network, determine the boundary node prediction data of the pipeline boundary node layer in the target natural gas pipeline network.

[0042] In this embodiment, the graph neural network model refers to a deep learning model that specializes in processing graph structure data. In the natural gas pipeline network, the graph structure represents the pipeline network nodes (such as pipelines and valves at different locations) and the connections between them; the graph neural network infers the state of the system through the relationship between the nodes and can process complex, spatially distributed data. Current structure data refers to the key information of the operation of each part of the pipeline network; specifically, the current structure data may include at least one of the pipe section data, compressor data, valve data and pipeline network node data; wherein the pipe section data includes pipe length, pipe diameter, elevation, roughness, etc.; compressor data includes the pressure ratio of the compressor station; valve data includes the valve coefficient of the valve; node data includes the connection relationship between each node and the pipe section, compressor, valve, etc. The preset system data refers to the preset values ​​of certain key parameters in the target natural gas pipeline network within a preset time period in the future; specifically, the data refers to the preset values ​​of the gas source pressure and user flow in the target natural gas pipeline network within a preset time period in the future. Parameter dynamic data refers to the dynamic change data of each node in the target natural gas pipeline network (such as pressure, flow, temperature, etc. in the pipeline); these data reflect the state of the pipeline network that changes over time; the dynamic data can be the hydraulic parameters and thermal parameters of each node in the pipeline network. The boundary node layer refers to the boundary nodes in the natural gas pipeline network. These nodes are usually located at the input or output end of the pipeline network, and their status directly affects the overall operation of the pipeline network system; these nodes can be located at the entrance, exit or connection to the external system (such as gas source, user) of the pipeline network; the boundary node layer can include at least one of the gas source node, user node, valve room inlet and outlet node, compressor station inlet and outlet node and distribution station inlet and outlet node. Boundary node prediction data refers to the predicted values ​​of the hydraulic parameters and thermal parameters of each node in the boundary node layer.

[0043] Exemplarily, according to the current structural data of the target natural gas pipeline network, the nodes of the pipeline network are divided into two levels, the first level is the boundary node layer, which does not include all nodes including the internal nodes of the pipeline segment; the second level is the internal node layer of the pipeline segment, which is the preset pipeline segment nodes; the parameter dynamic data at time t is used as the initial condition, the preset values ​​of the gas source pressure and user flow in the next 12 hours are used as the boundary conditions, and the pressure ratio of the compressor station is set as the attribute value of the edge established by the compressor station in the graph neural network; the multi-step GNN model and the single-step GNN model are repeatedly called to predict the complete prediction value every 3 minutes in the next 12 hours, for a total of 240 time steps; the predicted value obtained in the graph neural network GNN each time is used as the initial value, and the mechanism model is called to perform an iteration of dynamic simulation to obtain the predicted value after the correction of the mechanism model, and the full-factor pipeline network node simulation combining mechanism and data is completed to obtain the boundary node prediction data of the pipeline boundary node layer in the target natural gas pipeline network.

[0044] S130, based on the neural operator model and the mechanism simulation model, according to the parameter dynamic data and the candidate node prediction data, determine the pipe segment node prediction data of the internal node layer of the pipe segment.

[0045] In this embodiment, the neural operator model refers to a model based on deep learning, which aims to approximate complex mathematical operators through neural networks, especially in the application of physical and engineering problems; here, the neural operator model is combined with a mechanism simulation model to predict the behavior of the pipeline network, especially through historical data and dynamic parameters to improve the accuracy of the prediction. The candidate node prediction data is the predicted value of the nodes at both ends of each pipe section in the target natural gas pipeline network screened from the boundary node prediction data. The internal node layer of the pipe section is composed of preset pipe section segment nodes. The pipe section node prediction data refers to the predicted values ​​of the hydraulic parameters and thermal parameters of each node in the internal node layer of the pipe section.

[0046] Exemplarily, the boundary node prediction data for the next 12 hours are collated, and the prediction values ​​of the import and export nodes of each pipe segment are extracted; the internal node value of each pipe segment at time t obtained by the mechanism model is used as the initial condition, and the pipe segment inlet pressure and outlet flow in the boundary node prediction data are used as boundary conditions, and the neural operator model is simulated in a distributed manner to obtain the prediction value of the internal node of each pipe segment in the next 12 hours; then, the prediction value of the internal node of each pipe segment obtained by the neural operator model is used as the initial condition, and the mechanism model is called to perform an iteration of dynamic simulation to obtain the prediction value after the mechanism model is corrected, and the pipe segment simulation combining the mechanism and data is completed to obtain the pipe segment node prediction data of the internal node layer of the pipe segment.

[0047] S140: Summarize the boundary node prediction data and the pipe segment node prediction data to obtain the pipe network simulation prediction data of the target natural gas pipe network within a preset time period in the future.

[0048] In this embodiment, the pipeline network simulation prediction data refers to the overall simulation prediction data of the target natural gas pipeline network in a preset time period in the future obtained by integrating the boundary node prediction data and the pipeline segment node prediction data; this data provides a prediction of the operating status of the pipeline network in the future period of time, and is usually used for scheduling, optimization and safety management.

[0049] Exemplarily, boundary node prediction data and pipe segment node prediction data are collated and aggregated to obtain complete prediction values ​​for all nodes of the entire pipe network for the next 12 hours.

[0050] The embodiment of the present application determines the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment based on the mechanism simulation model, according to the historical system data and the current system data; determines the boundary node prediction data of the pipeline boundary node layer in the target natural gas pipeline network based on the graph neural network model and the mechanism simulation model according to the current structure data, preset system data and parameter dynamic data of the target natural gas pipeline network; determines the segment node prediction data of the internal node layer of the pipeline segment based on the neural operator model and the mechanism simulation model according to the parameter dynamic data and the candidate node prediction data; summarizes the boundary node prediction data and the segment node prediction data to obtain the pipeline simulation prediction data of the target natural gas pipeline network within a preset time period in the future. The above-mentioned technical scheme, based on the combination of mechanism model and GNN (Graph Neural Network) data model, has the effect of improving the speed, accuracy and stability of simulation calculation. By integrating the traditional mechanism model with modern data-driven technology, this method can significantly speed up the calculation process while ensuring the prediction accuracy, and meet the real-time and frequent online simulation needs. In addition, the data model effectively improves the prediction ability of future states, while the mechanism model helps to correct the prediction deviation that may occur in the data model, ensuring the reliability and consistency of the simulation results. Through these technical means, the present invention realizes high-speed and stable online simulation of natural gas pipeline networks, which can meet the high-frequency computing needs of real-time operation and maintenance, and has good economy and practicality, and has broad application prospects.

[0051] Embodiment 2

[0052] Figure 2 It is a flow chart of a natural gas pipeline network simulation prediction method provided in accordance with Example 2 of the present application. Based on the technical solutions of the above-mentioned embodiments, this embodiment refines "based on the graph neural network model and the mechanism simulation model, according to the current structure data, preset system data and parameter dynamic data of the target natural gas pipeline network, determine the boundary node prediction data of the pipeline network boundary node layer in the target natural gas pipeline network" into "using the pipe section data, compressor data and valve data to form the edges of the graph neural network, and using the pipeline node data to form the nodes of the graph neural network, to obtain the target graph neural network; based on the graph neural network model and the target graph neural network, using the parameter dynamic data as the initial value and the preset system data as the boundary condition, predict the data of the pipeline network boundary nodes of the pipeline network boundary node layer, and obtain candidate prediction data of multiple nodes in the pipeline network boundary node layer; substitute the candidate prediction data as the initial value into the mechanism simulation model to correct the candidate node prediction data, and obtain the boundary node prediction data of the pipeline network boundary node layer in the target natural gas pipeline network". It should be noted that for the parts not described in detail in the embodiments of the present application, please refer to the relevant descriptions of other embodiments. Figure 2 As shown, the method includes:

[0053] S210. Based on the mechanism simulation model, according to the historical system data and the current system data, determine the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment.

[0054] S220. Use the pipe section data, compressor data and valve data to form the edges of the graph neural network, and use the pipe network node data to form the nodes of the graph neural network, so as to obtain the target graph neural network.

[0055] In this embodiment, the pipe section data refers to the relevant data of different pipe sections in the natural gas pipeline network, which usually includes information such as the pressure, flow, temperature, pipeline material, length, diameter, etc. of the pipeline; wherein the pipe section data can specifically refer to the length, diameter, elevation and roughness of the pipe section. Compressor data refers to the data of the compressor equipment in the natural gas pipeline network. The compressor is used to increase the pressure of natural gas in the pipeline system to ensure the transportation of gas; the compressor data includes the working status, input and output pressure, power, operating temperature, etc. of the compressor; wherein the compressor data can be the pressure ratio of the compressor station. Valve data refers to the relevant data of valves in the natural gas pipeline network. Valves are used to control the flow and direction of airflow; valve data can include parameters such as valve opening, pressure, flow, temperature, etc.; wherein the valve data can specifically refer to valve coefficients. The target graph neural network refers to a complex relationship structure constructed by nodes (such as connection points in the pipeline network) and edges (such as connections between pipe sections) in the natural gas pipeline network.

[0056] Specifically, the graph neural network represents the boundary nodes of the pipeline network with graph data according to the topological structure, wherein the official gas source, users, valve inlets and outlets, and compressor inlets and outlets represent nodes in the graph neural network, while the pipe sections, compressor stations, and valves represent edges in the graph neural network, and the nodes are connected by edges; wherein, the nodes and edges in the graph neural network have different parameters, the nodes have pressure, flow, and temperature attribute values, the edges formed by the pipe sections have pipe length, pipe diameter, elevation, and roughness attribute values, the edges formed by the compressor stations have pressure ratio attributes, and the edges formed by the valves have valve coefficient attributes.

[0057] For example, the natural gas pipeline network represented by the target graph neural network can be expressed by the following formula:

[0058]

[0059] Among them, G n×m Represents a graph-structured pipe network with n nodes and m edges. n Represents the boundary node, including attribute values ​​such as pressure P, flow Q, temperature T, etc. E represents the edge of the graph-structured pipe network. pipe Refers to the edge formed by the pipe segment, including the pipe length L, pipe diameter D, elevation Δ, roughness e and other attribute values. compRefers to the changes in the compressor structure, including the pressure ratio ε attribute value. valve Refers to the edge formed by the valve, including the valve coefficient δ attribute value.

[0060] S230. Based on the graph neural network model and the target graph neural network, with the parameter dynamic data as the initial value and the preset system data as the boundary condition, data prediction is performed on the pipeline network boundary nodes in the pipeline network boundary node layer to obtain candidate prediction data for multiple nodes in the pipeline network boundary node layer.

[0061] In this embodiment, the graph neural network model may include at least one of a single-step graph neural network model and a multi-step graph neural network model. Among them, the single-step graph neural network model refers to completing the update of the node in one information transmission (usually the calculation of one layer of neural network); this means that the representation (feature) of each node is only affected by its direct neighbor nodes; in the single-step GNN model, the information between nodes is only propagated once, which is usually suitable for simple graph structures or scenarios that require fast calculations. The multi-step graph neural network model refers to updating the representation of the node in multiple steps (multi-layer information transmission); in this model, the representation of each node is not only affected by its direct neighbor nodes, but also by distant nodes (through multiple hop paths); this enables the node to learn more levels of information in multiple propagation steps, which is particularly suitable for complex or long-link graph structures.

[0062] Exemplarily, the graph neural network assigns the corresponding boundary values ​​of the boundary nodes in the graph neural network as additional attribute parameters according to the preset future boundary values ​​of the gas source and user nodes, and takes the pressure, flow, and temperature of the nodes at time t as initial values. The graph neural network updates the initial values ​​of these nodes to obtain the predicted values ​​of the nodes.

[0063] Exemplarily, the update of nodes using the graph neural network model in this application can be implemented by the following formula:

[0064] V t+1 =Update(V t , V boundary , E t );

[0065] Among them, V t+1 Refers to the predicted value of the node. Update refers to the update function of the graph neural network model. V refers to the initial value of the node at time t. V boundary It refers to the boundary value of the node at time t. t It refers to the initial value of the edge at time t.

[0066] Optionally, based on a multi-step graph neural network model, with parameter dynamic data as initial values ​​and preset system data as boundary conditions, data prediction is performed on the network boundary nodes of the network boundary node layer within a first preset time period to obtain first node prediction data; the first preset time period refers to the time period from the current moment to the candidate future moment within the future preset time period; based on a single-step graph neural network model, with the first node prediction data as initial values, the node prediction value corresponding to the candidate future moment is predicted to obtain the first node prediction value, so as to correct the node prediction value corresponding to the candidate future moment in the first node prediction data; continue to execute data prediction operations based on the multi-step graph neural network model and correction operations based on the single-step graph neural network model until candidate prediction data for multiple nodes in the network boundary node layer within the future preset time period are obtained.

[0067] In this embodiment, the first preset time period refers to a future time interval, from the current moment to a certain time period of the candidate future moment; within this time period, the boundary node layer of the pipe network will be predicted, and the prediction is the state change of each node in the pipe network system within this interval; the first preset time period can be a time period within one hour from the current moment. The first node prediction data refers to the data obtained after predicting each node in the boundary node layer of the pipe network within the first preset time period based on the multi-step graph neural network model. Candidate future moments refer to certain moments in the future that the model needs to speculate during the prediction process. Usually these moments are from the current moment through a preset time period (such as 1 hour, 1 day, etc.); within this time period, the model will predict the node data of the pipe network system at different times. The node prediction value refers to the prediction value of the candidate future moment obtained after predicting each node of the boundary node layer of the pipe network through a single-step GNN model.

[0068] For example, the 3-minute frequency future dynamic simulation calculation can be divided into two levels: hourly level: call the multi-step GNN model to directly predict the node data of 20 time steps in the next hour; minute level, call the single-step GNN model to predict the node data of the next time step. Specifically, first use multi-step prediction to obtain the node data within 1 hour when the boundary conditions remain unchanged, and then use the node data as the initial value, and use single-step prediction to obtain the node data when the boundary conditions change from the first hour to the second hour. By repeatedly calling the two models, a complete prediction for each moment in the next multiple hours is obtained.

[0069] It is understandable that the combined prediction of the multi-step GNN model and the single-step GNN model avoids the error accumulation phenomenon caused by repeated calls to single-step predictions in traditional data models during predictions. With a prediction frequency of 3 minutes for 12 hours, the original 240 predictions are reduced to 12 calls to the multi-step prediction model and 12 calls to the single-step prediction model, a total of 24 predictions, which reduces the number of predictions by ten times, strictly controls the error accumulation, and greatly improves the calculation speed.

[0070] Exemplarily, the use of a single-step GNN model and a multi-step GNN model can be implemented by the following formula:

[0071]

[0072] Among them, V t+1 Refers to the predicted value of the node at time t+1. GNN s Refers to a single-step GNN model. GNN refers to a multi-step GNN model.

[0073] S240: Substitute the candidate prediction data as initial values ​​into the mechanism simulation model to correct the candidate node prediction data, and obtain the boundary node prediction data of the boundary node layer of the pipeline network in the target natural gas pipeline network.

[0074] Exemplarily, the predicted value of each time step obtained by the graph neural network GNN is used as the initial value and substituted into the mechanism model to perform a dynamic simulation. After a correction is completed, the predicted value of the coarse node level of the natural gas pipeline network combining data and mechanism is obtained.

[0075] It is understandable that since the predicted value obtained by the graph neural network GNN model is close to the true value, calling the mechanism simulation model for simulation at this time can avoid multiple iterative calculations of the mechanism model. Only one iterative calculation is required to complete the correction.

[0076] S250, based on the neural operator model and the mechanism simulation model, according to the parameter dynamic data and the candidate node prediction data, determine the pipe segment node prediction data of the internal node layer of the pipe segment.

[0077] Optionally, based on the neural operator model, with parameter dynamic data as initial value and candidate node prediction data as boundary condition, data prediction is performed on the pipe segment nodes of the internal node layer of the pipe segment to obtain intermediate node prediction data of multiple nodes of the internal node layer of the pipe segment within a preset time period in the future; the intermediate node prediction data is substituted into the mechanism simulation model as the initial value to correct the intermediate node prediction data to obtain the pipe segment node prediction data of the internal node layer of the pipe segment.

[0078] In this embodiment, the intermediate node prediction data refers to the node data of the internal node layer of the pipe segment obtained in the neural operator model prediction; these data are the future state prediction results of the internal nodes of the pipe segment calculated by the neural operator model under given initial values ​​and boundary conditions.

[0079] It should be noted that the neural operator model transforms the solution process of the partial differential equations contained in the flow process of the pipe section into a combination of basis functions of the operator network through a neural network. After giving the boundary conditions, initial conditions and pipe section parameters, the dynamic simulation prediction of each node in the pipeline is realized.

[0080] Exemplarily, the prediction of the neural operator model can be achieved by the following formula:

[0081]

[0082] in, It refers to the predicted value of the internal node n of the pipe segment at time t+1. F refers to the neural operator model. P in It refers to the pressure boundary value at the inlet of the pipe section. out It refers to the flow boundary value at the outlet of the pipe section. T refers to the pressure boundary value of the pipe section. It refers to the initial value of the pressure at the internal node n of the pipe segment. It refers to the initial flow value of node n inside the pipe section. It refers to the initial value of the temperature of the internal node n of the pipe segment.

[0083] S260: Summarize the boundary node prediction data and the pipe segment node prediction data to obtain the pipe network simulation prediction data of the target natural gas pipe network within a preset time period in the future.

[0084] The embodiment of the present application determines the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment based on the mechanism simulation model, according to the historical system data and the current system data; the pipe segment data, compressor data and valve data constitute the edges of the graph neural network, and the pipe network node data constitute the nodes of the graph neural network to obtain the target graph neural network; based on the graph neural network model and the target graph neural network, with the parameter dynamic data as the initial value and the preset system data as the boundary condition, the data of the pipe network boundary nodes in the pipe network boundary node layer are predicted to obtain candidate prediction data of multiple nodes in the pipe network boundary node layer; the candidate prediction data are substituted into the mechanism simulation model as the initial value to correct the candidate node prediction data to obtain the boundary node prediction data of the pipe network boundary node layer in the target natural gas pipeline network; based on the neural operator model and the mechanism simulation model, the pipe segment node prediction data of the pipe segment internal node layer is determined according to the parameter dynamic data and the candidate node prediction data; the boundary node prediction data and the pipe segment node prediction data are summarized to obtain the pipe network simulation prediction data of the target natural gas pipeline network in the preset time period in the future. The above technical solution, based on the combination of mechanism model and GNN (Graph Neural Network) data model, has the effect of improving the speed, accuracy and stability of simulation calculation. By integrating traditional mechanism model with modern data-driven technology, this method can significantly speed up the calculation process while ensuring prediction accuracy, thus meeting the needs of real-time and frequent online simulation.

[0085] Embodiment 3

[0086] Figure 3 This is a schematic diagram of the structure of a natural gas pipeline network simulation prediction device provided in Example 3 of the present application, which can be applied to the case of online simulation of hydraulic and thermal parameter data of the natural gas pipeline network. The natural gas pipeline network simulation prediction device can be implemented in the form of hardware and / or software, and the natural gas pipeline network simulation prediction device can be configured in a computer device, such as a server. Figure 3 As shown, the device comprises:

[0087] The data determination module 310 is used to determine the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment based on the mechanism simulation model, according to the historical system data and the current system data; the historical system data refers to the system data of the historical time period from the candidate historical moment to the current moment obtained from the intermediate database at a preset frequency; the current system data refers to the system data at the current moment obtained from the intermediate database at a preset frequency;

[0088] The boundary node simulation module 320 is used to determine the boundary node prediction data of the boundary node layer of the pipeline network in the target natural gas pipeline network based on the graph neural network model and the mechanism simulation model according to the current structure data, preset system data and parameter dynamic data of the target natural gas pipeline network; the preset system data refers to the preset values ​​of the gas source pressure and user flow in the target natural gas pipeline network within a preset time period in the future;

[0089] The pipe segment node simulation module 330 is used to determine the pipe segment node prediction data of the internal node layer of the pipe segment based on the neural operator model and the mechanism simulation model according to the parameter dynamic data and the candidate node prediction data; the candidate node prediction data is the prediction value of the nodes at both ends of each pipe segment in the target natural gas pipeline network screened from the boundary node prediction data;

[0090] The data aggregation module 340 is used to aggregate the boundary node prediction data and the pipe segment node prediction data to obtain the pipe network simulation prediction data of the target natural gas pipe network within a preset time period in the future.

[0091] The embodiment of the present application determines the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment based on the mechanism simulation model, according to the historical system data and the current system data; determines the boundary node prediction data of the pipeline boundary node layer in the target natural gas pipeline network based on the graph neural network model and the mechanism simulation model according to the current structure data, preset system data and parameter dynamic data of the target natural gas pipeline network; determines the segment node prediction data of the internal node layer of the pipeline segment based on the neural operator model and the mechanism simulation model according to the parameter dynamic data and the candidate node prediction data; summarizes the boundary node prediction data and the segment node prediction data to obtain the pipeline simulation prediction data of the target natural gas pipeline network within a preset time period in the future. The above-mentioned technical scheme, based on the combination of mechanism model and GNN (Graph Neural Network) data model, has the effect of improving the speed, accuracy and stability of simulation calculation. By integrating the traditional mechanism model with modern data-driven technology, this method can significantly speed up the calculation process while ensuring the prediction accuracy, and meet the real-time and frequent online simulation needs. In addition, the data model effectively improves the prediction ability of future states, while the mechanism model helps to correct the prediction deviation that may occur in the data model, ensuring the reliability and consistency of the simulation results. Through these technical means, the present invention realizes high-speed and stable online simulation of natural gas pipeline networks, which can meet the high-frequency computing needs of real-time operation and maintenance, and has good economy and practicality, and has broad application prospects.

[0092] Optionally, the current structure data includes pipe segment data, compressor data, valve data and pipe network node data; accordingly, the boundary node simulation module 320 includes:

[0093] A network construction unit, used to construct edges of the graph neural network with pipe section data, compressor data and valve data, and to construct nodes of the graph neural network with pipe network node data, so as to obtain a target graph neural network;

[0094] A data prediction unit is used to predict data of the pipe network boundary nodes in the pipe network boundary node layer based on the graph neural network model and the target graph neural network, taking the parameter dynamic data as the initial value and the preset system data as the boundary condition, and obtain candidate prediction data of multiple nodes in the pipe network boundary node layer;

[0095] The data correction unit is used to substitute the candidate prediction data as initial values ​​into the mechanism simulation model to correct the candidate node prediction data and obtain the boundary node prediction data of the boundary node layer of the pipeline network in the target natural gas pipeline network.

[0096] Optionally, the graph neural network model includes: a single-step graph neural network model and a multi-step graph neural network model; correspondingly, the data prediction unit is specifically used to:

[0097] Based on the multi-step graph neural network model, with the parameter dynamic data as the initial value and the preset system data as the boundary condition, the data of the pipe network boundary nodes of the pipe network boundary node layer within the first preset time period are predicted to obtain the first node prediction data; the first preset time period refers to the time period from the current moment to the candidate future moment within the future preset time period;

[0098] Based on the single-step graph neural network model, the node prediction value corresponding to the candidate future moment is predicted with the first node prediction data as the initial value to obtain the first node prediction value, so as to correct the node prediction value corresponding to the candidate future moment in the first node prediction data;

[0099] Continue to perform the data prediction operation based on the multi-step graph neural network model and the correction operation of the single-step graph neural network model until the candidate prediction data of multiple nodes in the boundary node layer of the pipeline network within a preset time period in the future are obtained.

[0100] Optionally, the pipe segment node simulation module 330 is specifically used for:

[0101] Based on the neural operator model, the parameter dynamic data is used as the initial value and the candidate node prediction data is used as the boundary condition to predict the data of the pipe segment nodes in the internal node layer of the pipe segment, and the intermediate node prediction data of multiple nodes in the internal node layer of the pipe segment in the future preset time period are obtained;

[0102] The intermediate node prediction data is substituted into the mechanism simulation model as the initial value to correct the intermediate node prediction data and obtain the pipe segment node prediction data of the internal node layer of the pipe segment.

[0103] Optionally, the data determination module 310 includes:

[0104] A steady-state data determination unit is used to perform steady-state simulation calculations on multiple nodes in the target natural gas pipeline network based on a mechanism simulation model and taking historical system data corresponding to the candidate historical moment as a boundary value, so as to obtain parameter steady-state data of multiple nodes in the target natural gas pipeline network at the candidate historical moment;

[0105] The dynamic data determination unit is used to determine the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment based on the mechanism simulation model, according to the candidate system data and the parameter steady-state data; the candidate system data refers to the historical system data corresponding to the time period from the next moment of the candidate historical moment to the current moment.

[0106] Optionally, the dynamic data determination unit is specifically used for:

[0107] Based on the mechanism simulation model, the candidate system data is used as the boundary value and the parameter steady-state data is used as the initial value to perform dynamic simulation calculations on multiple nodes in the target natural gas pipeline network to obtain the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment;

[0108] Among them, the parameter dynamic data includes the pressure, flow and temperature of multiple nodes in the target natural gas pipeline network.

[0109] The natural gas pipeline network simulation prediction device provided in the embodiment of the present application can execute the natural gas pipeline network simulation prediction method provided in any embodiment of the present application, and has the corresponding functional modules and beneficial effects for executing each natural gas pipeline network simulation prediction method.

[0110] According to an embodiment of the present application, the present application also provides an electronic device, a readable storage medium and a computer program product.

[0111] Embodiment 4

[0112] Figure 4 4 is a schematic diagram of the structure of an electronic device 410 for implementing the natural gas pipeline network simulation prediction method of an embodiment of the present application. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0113] like Figure 4As shown, the electronic device 410 includes at least one processor 411, and a memory connected to the at least one processor 411 in communication, such as a read-only memory (ROM) 412, a random access memory (RAM) 413, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 411 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 412 or the computer program loaded from the storage unit 418 to the random access memory (RAM) 413. In RAM413, various programs and data required for the operation of the electronic device 410 can also be stored. The processor 411, ROM412 and RAM413 are connected to each other via a bus 414. An input / output (I / O) interface 415 is also connected to the bus 414.

[0114] Multiple components in the electronic device 410 are connected to the I / O interface 415, including: an input unit 416, such as a keyboard, a mouse, etc.; an output unit 417, such as various types of displays, speakers, etc.; a storage unit 418, such as a disk, an optical disk, etc.; and a communication unit 419, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 419 allows the electronic device 410 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0115] The processor 411 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 411 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 411 executes the various methods and processes described above, such as a natural gas pipeline network simulation prediction method.

[0116] In some embodiments, the natural gas pipeline network simulation prediction method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 418. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 410 via the ROM 412 and / or the communication unit 419. When the computer program is loaded into the RAM 413 and executed by the processor 411, one or more steps of the natural gas pipeline network simulation prediction method described above may be performed. Alternatively, in other embodiments, the processor 411 may be configured as a natural gas pipeline network simulation prediction method in any other appropriate manner (e.g., by means of firmware).

[0117] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0118] The computer programs for implementing the methods of the present application may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable natural gas pipeline network simulation prediction device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0119] In the context of the present application, a computer readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, device or equipment. A computer readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer readable storage medium may be a machine readable signal medium. A more specific example of a machine readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0120] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0121] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0122] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0123] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this application can be executed in parallel, sequentially or in different orders, as long as the expected results of the technical solution of this application can be achieved, and this document is not limited here.

[0124] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A natural gas pipeline network simulation prediction method, characterized in that: include: Based on the mechanism simulation model, according to the historical system data and the current system data, the dynamic parameter data of multiple nodes in the target natural gas pipeline network at the current moment are determined; The historical system data refers to the system data of the historical time period from the candidate historical moment to the current moment obtained from the intermediate database at a preset frequency; the current system data refers to the system data of the current moment obtained from the intermediate database at a preset frequency; Based on the graph neural network model and the mechanism simulation model, the boundary node prediction data of the boundary node layer of the target natural gas pipeline network is determined according to the current structure data of the target natural gas pipeline network, the preset system data and the parameter dynamic data; the preset system data refers to the preset values ​​of the gas source pressure and user flow in the target natural gas pipeline network within a preset time period in the future; Based on the neural operator model and the mechanism simulation model, the pipe segment node prediction data of the internal node layer of the pipe segment is determined according to the parameter dynamic data and the candidate node prediction data; the candidate node prediction data is the prediction value of the nodes at both ends of each pipe segment in the target natural gas pipeline network screened from the boundary node prediction data; The boundary node prediction data and the pipe section node prediction data are aggregated to obtain the pipe network simulation prediction data of the target natural gas pipe network within the future preset time period.

2. The method according to claim 1, characterized in that The current structure data includes pipe section data, compressor data, valve data and pipeline network node data; accordingly, based on the graph neural network model and the mechanism simulation model, according to the current structure data of the target natural gas pipeline network, the preset system data and the parameter dynamic data, the boundary node prediction data of the pipeline network boundary node layer in the target natural gas pipeline network is determined, including: The pipe section data, the compressor data and the valve data constitute the edges of the graph neural network, and the pipe network node data constitute the nodes of the graph neural network, to obtain a target graph neural network; Based on the graph neural network model and the target graph neural network, taking the parameter dynamic data as the initial value and the preset system data as the boundary condition, data prediction is performed on the pipe network boundary nodes of the pipe network boundary node layer to obtain candidate prediction data of multiple nodes in the pipe network boundary node layer; Substituting the candidate prediction data as initial values ​​into the mechanism simulation model to correct the candidate node prediction data, thereby obtaining the boundary node prediction data of the boundary node layer of the pipeline network in the target natural gas pipeline network.

3. The method according to claim 2, characterized in that The graph neural network model includes: a single-step graph neural network model and a multi-step graph neural network model; accordingly, based on the graph neural network model and the target graph neural network, the parameter dynamic data is used as the initial value, and the preset system data is used as the boundary condition to perform data prediction on the pipe network boundary nodes of the pipe network boundary node layer, and obtain candidate prediction data of multiple nodes in the pipe network boundary node layer, including: Based on the multi-step graph neural network model, taking the parameter dynamic data as the initial value and the preset system data as the boundary condition, data prediction is performed on the pipe network boundary nodes of the pipe network boundary node layer within a first preset time period to obtain first node prediction data; the first preset time period refers to a time period from the current moment to the candidate future moment within a future preset time period; Based on the single-step graph neural network model, taking the first node prediction data as an initial value, predicting the node prediction value corresponding to the candidate future moment to obtain a first node prediction value, so as to correct the node prediction value corresponding to the candidate future moment in the first node prediction data; Continue to perform the data prediction operation based on the multi-step graph neural network model and the correction operation of the single-step graph neural network model until the candidate prediction data of multiple nodes in the pipeline network boundary node layer within the future preset time period are obtained.

4. The method according to claim 1, characterized in that Based on the neural operator model and the mechanism simulation model, according to the parameter dynamic data and the candidate node prediction data, the pipe segment node prediction data of the internal node layer of the pipe segment is determined, including: Based on the neural operator model, the parameter dynamic data is used as the initial value and the candidate node prediction data is used as the boundary condition to predict the data of the pipe segment nodes of the internal node layer of the pipe segment, and the intermediate node prediction data of multiple nodes of the internal node layer of the pipe segment in the future preset time period are obtained; Substitute the intermediate node prediction data as initial values ​​into the mechanism simulation model to correct the intermediate node prediction data and obtain the pipe segment node prediction data of the internal node layer of the pipe segment.

5. The method according to claim 1, characterized in that Based on the mechanism simulation model, according to the historical system data and the current system data, the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment are determined, including: Based on the mechanism simulation model, taking the historical system data corresponding to the candidate historical moment as the boundary value, a steady-state simulation calculation is performed on multiple nodes in the target natural gas pipeline network to obtain the parameter steady-state data of multiple nodes in the target natural gas pipeline network at the candidate historical moment; Based on the mechanism simulation model, according to the candidate system data and the parameter steady-state data, the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment are determined; the candidate system data refers to the historical system data corresponding to the time period from the next moment of the candidate historical moment to the current moment.

6. The method according to claim 5, characterized in that Based on the mechanism simulation model, according to the candidate system data and the parameter steady-state data, the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment are determined, including: Based on the mechanism simulation model, taking the candidate system data as the boundary value and the parameter steady-state data as the initial value, a dynamic simulation calculation is performed on multiple nodes in the target natural gas pipeline network to obtain the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment; The parameter dynamic data includes pressure, flow and temperature of multiple nodes in the target natural gas pipeline network.

7. A natural gas pipeline network simulation prediction device, characterized in that: include: A data determination module is used to determine the parameter dynamic data of multiple nodes in the target natural gas pipeline network at the current moment based on the mechanism simulation model, according to the historical system data and the current system data; The historical system data refers to the system data of the historical time period from the candidate historical moment to the current moment obtained from the intermediate database at a preset frequency; the current system data refers to the system data of the current moment obtained from the intermediate database at a preset frequency; A boundary node simulation module is used to determine the boundary node prediction data of the boundary node layer of the target natural gas pipeline network based on the graph neural network model and the mechanism simulation model according to the current structure data of the target natural gas pipeline network, the preset system data and the parameter dynamic data; the preset system data refers to the preset values ​​of the gas source pressure and user flow in the target natural gas pipeline network within a preset time period in the future; A pipe segment node simulation module, for determining the pipe segment node prediction data of the internal node layer of the pipe segment based on the neural operator model and the mechanism simulation model according to the parameter dynamic data and the candidate node prediction data; the candidate node prediction data is the prediction value of the nodes at both ends of each pipe segment in the target natural gas pipeline network screened from the boundary node prediction data; The data aggregation module is used to aggregate the boundary node prediction data and the pipe segment node prediction data to obtain the pipeline network simulation prediction data of the target natural gas pipeline network within the future preset time period.

8. An electronic device, characterized in that: include: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the natural gas pipeline network simulation prediction method as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the natural gas pipeline network simulation prediction method as described in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the natural gas pipeline network simulation prediction method according to any one of claims 1 to 6 is implemented.

Images