A real-time state analysis system based on three-dimensional pipe network topology

Abstract

This invention discloses a real-time status analysis system based on a three-dimensional pipeline network topology, belonging to the field of computer technology. It includes a three-dimensional pipeline network topology construction module, a real-time data interface module, a pipeline network component mathematical model construction module, a pipeline network analysis module, and a data storage and analysis module. Specifically, the three-dimensional pipeline network topology construction module uses a domestically developed 3D engine to record the 3D coordinates and connection relationships of each endpoint of the pipeline network; the real-time data interface module obtains the real-time status and parameters of each component from the operating database; and the pipeline network analysis module uses the A* path search algorithm and Dijkstra's algorithm to calculate the medium flow direction, flow velocity, and theoretical operating values ​​of each instrument in each pipeline. Compared with existing technologies, this invention deeply integrates 3D spatial information with real-time operating data, dynamically adapts to changes in valve opening and closing, pump start-up and shutdown, etc., and completes advanced early warning through joint upstream and downstream topology verification, thereby achieving real-time status analysis and fault early warning for all elements at the pipeline network level.

Description

Technical Field

[0001] This invention relates to three-dimensional pipeline topology construction, mathematical simulation modeling, and pipeline path search technology, and belongs to the field of computer technology. Background Technology

[0002] Various pipe networks exist in production and living facilities across industrial and residential sectors, such as water supply networks, steam networks, gas networks, and chemical pipelines. Pipe networks typically consist of components such as tanks, pumps, fans, heat exchangers, straight pipe sections, pipe joints, orifice plates, valves, pressure gauges, flow meters, and temperature gauges. Pumps, fans, and elevation differences provide the power source for the pipe network; pipes, orifice plates, and valves generate resistance; and instruments are used to measure real-time operational data. During pipe network design, parameters of each component need to be calculated based on fluid medium characteristics, elevation differences between the beginning and end points of the network, and design flow and pressure, to select appropriate component models. Therefore, theoretically, the real-time state of a pipe network can be analyzed, and the operational status of each component can be modeled and simulated. If there is a deviation between the theoretical operating parameters of each component calculated through mathematical modeling and the actual measured values ​​of the instruments, and this deviation is getting larger and larger, it indicates that there are health problems in the pipeline network, such as: external leakage in the pipeline, internal leakage in the valve, wear of the orifice plate, performance degradation of pumps and fans, instrument malfunctions, etc.

[0003] Currently, fault early warning for equipment in the aforementioned fields has been developing for many years, and related technologies and products are relatively mature. However, fault early warning for pipeline networks is still in its early stages. Existing equipment fault early warning methods are mainly divided into three categories: methods based on mechanistic models, methods based on knowledge bases, and methods based on big data analysis.

[0004] (1) Mechanism-based methods

[0005] This is the earliest developed and most thoroughly researched method for fault early warning and condition monitoring. This type of method is closely integrated with control theory, primarily employing three specific methods—parameter estimation, state estimation, and equivalent space—to construct the difference sequence. Among these, the state estimation method is the most commonly used and can be implemented using an observer or a Kalman filter.

[0006] (2) Knowledge base-based approach

[0007] These methods primarily rely on the heuristic experience and knowledge of relevant experts and operators to qualitatively or quantitatively describe the connections between units and fault propagation patterns during the process. After abnormal signs appear in the equipment, they simulate the reasoning abilities of process experts in monitoring through inference and deduction, thereby automatically completing equipment fault early warning and equipment monitoring. While these methods do not require precise mathematical models, they heavily depend on expert knowledge. Commonly used methods include expert systems, fault decision trees, directed graphs, and fuzzy logic.

[0008] (3) Big data analysis-based methods

[0009] These methods establish mathematical models and express process states by mining the inherent information in process data, and then implement effective process monitoring based on the models. With the widespread application of intelligent instruments and computer storage technology, massive amounts of process data can be effectively monitored, collected, and stored. These methods are based on such massive amounts of data, and in terms of monitoring and early warning algorithms, they can be divided into five major categories: signal processing, rough sets, machine learning, information fusion, and multivariate statistics. Among them, machine learning algorithms are the most actively developing branch in both theory and practice, specifically including Bayesian classifiers, neural networks, support vector machines, k-nearest neighbor algorithms, clustering algorithms, principal component analysis, and other algorithms.

[0010] Compared with equipment failure early warning, the pipe network structures in the industrial and residential sectors vary greatly, and the status of each component in the same pipe network can change at any time (such as pumps and fans may start and stop at any time, and valves may open or close at any time). This makes dynamic analysis of the pipe network very difficult, so the development of failure early warning for pipe networks is relatively slow at present.

[0011] Pipeline networks comprise equipment, pipes, and instruments. While the performance of a pipeline network primarily depends on the performance of the equipment, pipes occupy the largest area, and instruments directly impact the control decisions of the control system. If only equipment fault warnings are provided, while fault warnings for pipes and instruments are neglected, the performance and safety of the entire pipeline system will also be compromised. For example, accidents caused by pipe ruptures can lead to a decline in the overall performance of the pipeline network, even resulting in production shutdowns, and in severe cases, personal injury accidents. Instrument malfunctions can reduce equipment performance to the point of causing the control system to issue erroneous commands, leading to shutdowns. Especially in high-temperature, high-pressure pipelines, gas pipelines, and chemical pipelines, accidents can cause secondary damage. Summary of the Invention

[0012] To address the aforementioned technical issues, this invention presents a real-time status analysis system based on a three-dimensional pipeline network topology. This system can calculate the theoretical operating flow rate, operating pressure, and operating temperature of each pipeline based on the pipeline network topology and pipeline path search algorithm. By comparing these values ​​with actual measured flow rate, temperature, and pressure, it can determine whether the three-dimensional pipeline network deviates from the theoretical operating conditions.

[0013] This invention provides a real-time dynamic analysis scheme for pipeline networks. It updates the network's operating status and data based on the real-time status of each component, ultimately constructing a real-time dynamic three-dimensional pipeline network topology. Then, a path search algorithm is used to analyze the flow velocity of the medium in each pipe of the current network, and mathematical models of each component are used to analyze the current network operating parameters, obtaining the theoretical operating parameters of each component. The deviation between the theoretical operating parameters and the actual measured values ​​is then calculated. Based on trend analysis of the deviation values, pipeline components with potential safety hazards are quickly identified, thereby achieving real-time early warning for the pipeline network.

[0014] The main technological innovations are as follows:

[0015] (1) A three-dimensional pipeline network topology is constructed using a domestic 3D engine to record the relationships between various components of the pipeline network (such as equipment, pipes, valves, orifice plates, measuring points, etc.) and to update the pipeline network status based on real-time data. Since a domestic 3D engine is used, the technical security is guaranteed, which is particularly important for the application of this invention in the domestic industrial and civilian sectors.

[0016] (2) The three-dimensional pipeline network topology is combined with real-time operating data. The elevation difference in the topology and the real-time operating data are substituted into Bernoulli's equation, heat exchanger calculation formula, etc., to perform real-time dynamic verification of the pipeline network. The pipeline network topology relationship is used to perform pre- and post-verification of pipeline measuring points to confirm whether the measuring point data is reasonable.

[0017] (3) For the real-time status of the three-dimensional pipeline network, the A* path search algorithm is used to analyze the medium status of each pipeline and calculate parameters such as the flow velocity of the pipeline medium. The A* path search algorithm is a widely used algorithm for path finding and graph traversal. The search process is guided by a heuristic function to find the shortest path from the starting point to the ending point. This function is usually expressed as g(n) + h(n). Where g(n) is the actual distance from the starting point to node n, and h(n) is the estimated distance from node n to the target.

[0018] (4) The theoretical pressure value at any point in the pipeline is calculated using the Dijkstra algorithm. Trend analysis is performed on the deviation between the theoretical value and the actual measured value to identify potential safety hazards. Depending on the actual situation, specific deviation analysis methods such as data sorting, statistical analysis, deviation measurement, and distribution pattern analysis can be used.

[0019] The technical solution adopted in this invention is a real-time analysis system based on a three-dimensional pipeline network topology, comprising:

[0020] The 3D pipeline topology construction module is used to construct a 3D pipeline topology structure by setting equipment, valves, tanks, orifice plates, etc. in the pipeline as endpoints according to the actual installation diagram of the industrial pipeline network, recording the endpoint codes, names, types, 3D coordinates and connection relationships, and recording pipeline attributes and instrument lists.

[0021] Real-time data interface module: used to obtain the status and operating parameters of each component in the pipeline network from the real-time operating database, and update all component data before each analysis cycle;

[0022] Pipeline Component Mathematical Model Construction Module: Used to build mathematical models of components such as pumps, fans, straight pipe sections, valves, and orifice plates, and calculate the theoretical working parameters of the components by combining elevation differences and real-time data;

[0023] Pipeline analysis module: Based on the three-dimensional topology and real-time data, it uses a path search algorithm to analyze the medium flow direction and velocity of each pipeline segment, and calculates the theoretical working values ​​of each instrument.

[0024] Data storage and analysis module: Used to store the theoretical and actual values ​​of each instrument in each analysis cycle, perform trend analysis on the deviation, and identify potential safety hazards in the pipeline network.

[0025] Furthermore, the three-dimensional pipeline topology is constructed using a domestic 3D engine, and the three-dimensional coordinates of each component are recorded for calculating pipeline length and potential energy.

[0026] Furthermore, the path search algorithm is the A* path search algorithm, which is used to analyze the flow direction and velocity of the medium in the pipeline. Its heuristic function is represented as g(n) + h(n), where g(n) is the actual distance from the starting point to node n, and h(n) is the estimated distance from node n to the target.

[0027] Furthermore, the pipeline analysis module uses the Dijkstra algorithm to calculate the theoretical pressure value at any point in the pipeline.

[0028] Furthermore, the trend analysis method for the deviation includes one or more of the following: data sorting, statistical analysis, deviation measurement, and distribution pattern analysis.

[0029] Furthermore, the data acquired by the real-time data interface module includes one or more of the following: valve on / off status, regulating valve opening degree, pump start / stop status, pump power, pipeline pressure, pipeline temperature, and pipeline flow rate.

[0030] Furthermore, in the pipeline component mathematical model construction module, the mathematical models of pumps and fans are derived from existing models, manufacturer data, or mathematical statistics of historical operating data.

[0031] Furthermore, the data storage and analysis module is also used to perform one or more of the following analyses:

[0032] Dynamic upper and lower limit analysis of parameters;

[0033] Parameter change rate alarm;

[0034] Pipeline topology analysis;

[0035] Add measuring points and verification for key equipment or pipelines.

[0036] Furthermore, the system is used to determine one or more of the following fault types: external leakage in pipelines, internal leakage in valves, orifice plate wear, degraded performance of pumps or fans, and instrument malfunction.

[0037] Furthermore, the system is applicable to pipeline networks containing liquid or gaseous fluids, including water supply networks, steam networks, gas networks, and chemical pipeline networks.

[0038] Furthermore, it also includes a set of data interaction interfaces, including:

[0039] First data interaction interface: Located between the three-dimensional pipeline topology construction module and the real-time data interface module, it is used to transmit component codes, endpoint connection relationships, and instrument list information to the real-time data interface module as a basis for obtaining real-time operating data;

[0040] The second data interaction interface is located between the real-time data interface module and the pipeline analysis module, and is used to transmit the updated component status and operating parameters to the pipeline analysis module.

[0041] The third data interaction interface is set between the three-dimensional pipeline topology construction module and the pipeline analysis module. It is used to provide the pipeline analysis module with topological relationships, endpoint coordinates, pipeline attributes and start and end point setting information for path search and elevation difference calculation.

[0042] The fourth data interaction interface is located between the pipeline component mathematical model construction module and the pipeline analysis module. It is used to provide the mathematical models of each component to the pipeline analysis module for use in the calculation of theoretical working parameters.

[0043] The fifth data interaction interface is located between the pipeline analysis module and the data storage and analysis module. It is used to transmit the theoretical working values ​​of each instrument in each analysis cycle to the data storage and analysis module for comparison, storage and trend analysis with the actual measured values.

[0044] The sixth data interaction interface is located between the data storage and analysis module and the three-dimensional pipeline topology construction module. It is used to feed back the information of identified components with potential safety hazards to the three-dimensional pipeline topology construction module so that abnormal components can be highlighted in the three-dimensional visualization interface.

[0045] Compared with existing technologies, this invention is the first to deeply integrate three-dimensional pipeline network topology with real-time operational data. Existing technologies (such as methods based on mechanistic models, knowledge bases, or big data analysis) generally neglect the influence of the spatial geometry of the pipeline network and elevation differences on the fluid state, and cannot dynamically adapt to real-time changes such as valve opening and closing, pump start-up and shutdown. This invention uses a domestically developed 3D engine to construct the pipeline network topology, records the three-dimensional coordinates of each endpoint, and directly substitutes the elevation differences into Bernoulli's equation, significantly improving the analytical accuracy of theoretical parameters such as pressure and flow rate.

[0046] This invention achieves early warning of pipeline network faults through path search algorithms and deviation trend analysis. Traditional methods typically analyze data from individual measuring points in isolation, lacking cross-verification between upstream and downstream parameters, and can only provide threshold alarms, failing to identify slowly developing faults. This invention uses the A* algorithm and Dijkstra's algorithm to jointly analyze the flow direction and pressure distribution in the pipeline, and through periodic comparisons and trend analysis (such as data sorting and deviation measurement) of theoretical and actual measurements, it can issue early warnings several days to several weeks before faults such as valve internal leakage and pump performance degradation occur.

[0047] The modular architecture and domestically developed technology approach of this invention offer excellent scalability and security. The modules are loosely coupled through standardized data interaction interfaces, supporting distributed deployment and hot model updates, and can flexibly adapt to various pipeline network scenarios such as steam, gas, chemical, and heating. Furthermore, the use of a domestically developed 3D engine fills the gap in existing technologies for real-time dynamic analysis and fault early warning at the pipeline network level. Attached Figure Description

[0048] Figure 1 This is a connection diagram of the system of the present invention. Detailed Implementation

[0049] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0050] This invention mainly includes the following modules: a three-dimensional pipeline network topology construction module, a real-time data interface module, a pipeline network component mathematical model construction module, a pipeline network analysis module, and a data storage and analysis module. The functions of each module are described below.

[0051] (1) Three-dimensional pipeline topology construction module.

[0052] The components of a pipeline network are interconnected. Tanks, pumps, fans, orifice plates, and valves are all connected by pipes. Instruments are installed on pipes or equipment, and the network formed by these components is an organic whole. This module, based on the actual installation diagram of the industrial pipeline network, sets equipment, valves, tanks, orifice plates, etc., as pipeline endpoints. It inputs the endpoint's code, name, endpoint type, and 3D coordinates (x, y, z), and records the list of other endpoints connected to it (ConnectPots), as well as the pipe attributes connecting each endpoint (pipe diameter, pipe material, pipe fluid, pipe design pressure, pipe design temperature, etc.). Then, it records the instrument list (MeasureList) in an external document, recording the code, name, instrument model, 3D coordinates of pipe openings, and 3D coordinates of instrument placement positions for all instruments. At this point, the complete 3D pipeline network topology is finished.

[0053] This module also requires setting the design flow direction of each pipeline, and the analytical start and end points of the pipeline network. During real-time dynamic analysis of the pipeline network, path searches will be performed for each start and end point to improve analysis efficiency. The three-dimensional coordinates at the pipeline network endpoints will be used to calculate pipeline length and potential energy, which is a built-in function of the three-dimensional topology structure.

[0054] (2) Real-time data interface module.

[0055] This module constructs the codes of each component in the module based on the three-dimensional pipeline network topology, and uses interface technology to obtain the status (valve opening and closing, regulating valve opening, pump start and stop) and operating parameters (pump power, pipeline pressure, pipeline temperature, pipeline flow, etc.) of each component in the pipeline network from the real-time running database, and updates all component data before the parsing cycle.

[0056] (3) Mathematical model construction module for pipeline components.

[0057] Pumps, fans, and elevation differences are the power sources for the pipeline network, with elevation differences directly obtainable from the network model. The mathematical models for pumps and fans can utilize existing commercially available models, information from relevant manufacturers, or model functions derived from historical operating data through statistical analysis. All variables involved in the mathematical models of each pipeline component should be related to known parameters at their respective network endpoints, avoiding situations requiring manual input or calculation. For example, the mathematical models of pumps and fans may require a specified input parameter, such as power or flow rate. This pre-defined input parameter value can be obtained through a real-time data interface and automatically substituted into the pump and fan mathematical models to calculate other unknown parameters.

[0058] Straight pipe sections, pipe joints, tees, orifice plates, and valves are the sources of resistance in the pipeline network. These can be solved using models provided by the manufacturer or by using Bernoulli's equation.

[0059] Pressure, flow, and temperature instruments serve as verifiers for real-time data analysis of the pipeline network. By analyzing the trend of the difference between the theoretical and actual values ​​of these instruments, the health status of the pipeline network can be analyzed. The focus is on determining whether there are external leaks in the pipes, internal leaks in the valves, orifice plate wear, assessing the degree of performance degradation of pumps and fans, and identifying instrument malfunctions. These malfunctions can all put the pipeline network in a sub-healthy state. If left unaddressed for a long time, this can lead to increased energy consumption or, in severe cases, accidents.

[0060] (4) Pipeline analysis module.

[0061] After obtaining the three-dimensional model of the pipeline network at a certain moment, it is necessary to analyze the flow direction and velocity of each pipeline segment, and then calculate the theoretical value of each instrument.

[0062] The pipeline network analysis module needs to be generalizable, capable of automatically analyzing different pipeline network models without requiring additional settings. This is a key challenge in pipeline network analysis. This is especially true for various parallel pipelines, where the dynamics and resistances in each branch are different, yet the parameters at the final merging point are unified.

[0063] (5) Data storage and analysis module.

[0064] During each analysis cycle, each instrument will have theoretical and actual values, and this data will be stored in the database. Trend analysis will be performed periodically on each component of the pipeline network, and conclusions will be drawn.

[0065] Based on trend analysis, early warning systems are issued for faults such as internal valve leakage, pipeline leakage, inaccurate instrument readings, and increased energy consumption in the pipeline network. The analysis includes the following aspects:

[0066] Parameter dynamic upper and lower limit analysis

[0067] For various parameters in the pipeline network, including temperature, pressure, liquid level, vibration, current, flow rate, and flow velocity, the upper and lower limits of the parameters are calculated based on the current operating status of the pipeline network, and the current parameters are diagnosed. When an abnormality occurs, an alarm is issued.

[0068] Parameter change rate alarm

[0069] The rate of change of various parameters in the pipeline network is also an important monitoring method. When parameters change rapidly, the rate of change should be intelligently monitored, and an alarm should be issued when an abnormality occurs.

[0070] Pipeline topology analysis

[0071] Typically, for critical parameters, a two-out-of-three or two-out-of-two mathematical algorithm is used to eliminate interference from erroneous instruments and calculate the final measured value. However, this redundant configuration involves very few measuring points, and configuring redundant instruments for all measuring points is impractical. Thermodynamic calculations can be performed using the pipeline topology and real-time system operating data, and trend analysis can be conducted on the difference between calculated and actual values ​​to achieve dynamic, instantaneous analysis of the entire pipeline network. This allows for cross-verification using parameters from upstream and downstream of the pipeline, overcoming the limitations of previous isolated parameter analyses.

[0072] For valve leaks, underground pipeline leaks, and the condition of high-altitude pipe networks, manual inspections are often insufficient. Furthermore, manual inspections cannot determine the accuracy of instrument readings, nor can they perform historical data trend analysis. For these critical, difficult-to-inspect (such as underground pipe networks) and frequently malfunctioning equipment or pipelines, increasing the number of measuring points and verification parameters can improve early warning capabilities.

[0073] The specific analysis process of the pipeline network analysis module includes: three-dimensional topology (endpoint coordinates, connection relationships, pipeline attributes); real-time operating data (valve opening, pump status, measured values ​​at each measuring point); and component mathematical models (pump characteristic curves, valve Cv values, orifice plate formulas, etc.).

[0074] Analysis steps:

[0075] Step 1: Determine the start and end points of the pipeline analysis;

[0076] Step 2: The A* algorithm searches for the medium flow path;

[0077] Step 3: Dijkstra's algorithm is used to calculate the pressure distribution;

[0078] Step 4: Solve for the flow velocity in each pipe section using Bernoulli's equation;

[0079] Step 5: Calculate the theoretical value of the instrument using the component mathematical model;

[0080] Step 6: Output the parsing results.

[0081] The detailed steps are described below:

[0082] Step 1: Determine the starting and ending points of the pipeline analysis.

[0083] Read the preset list of parsing start points and end points from the 3D pipeline topology construction module; the parsing start point is usually the power source outlet, such as pump outlet, boiler outlet, heat source station outlet, etc.; the parsing end point is usually the user access point, tank inlet, heat exchange station inlet, etc.

[0084] Step 2: The A* algorithm searches for the medium flow path.

[0085] Determine the possible flow paths of the medium from each starting point to each ending point, and identify the flow direction in the pipeline.

[0086] Each endpoint (tank, pump, valve, tee, etc.) in the three-dimensional topology is considered as a node in the graph.

[0087] The pipes connecting the endpoints are considered as edges in the diagram, and the weight of the edges can be set according to the pipe length or resistance coefficient.

[0088] The evaluation function of the A* algorithm:

[0089] f(n) = g(n) + h(n)

[0090] Where g(n): the actual pipe length (or cumulative resistance loss) from the starting point to the current node n; h(n): the estimated distance from the current node n to the end point, using Euclidean distance (calculated based on three-dimensional coordinates) or Manhattan distance.

[0091] Starting from the beginning, traverse all adjacent nodes; calculate the f-value of each candidate node, and prioritize expanding the node with the smallest f-value; repeat the above process until the search reaches the end point or all reachable nodes have been traversed.

[0092] Each path found is a possible flow path for the medium; the path direction is the actual flow direction of the medium in the pipe; for parallel pipes, the system will find multiple parallel paths and record them separately.

[0093] Step 2 output: The direction of medium flow for each pipe (from the start point to the end point); a complete list of paths between each start-end point pair.

[0094] Step 3: Dijkstra's algorithm is used to calculate the pressure distribution.

[0095] The endpoints of the pipeline network are taken as nodes, and the pipelines are taken as edges; the weight of the edges is set as the pipeline resistance coefficient (related to pipe diameter, length, roughness, and medium density).

[0096] Resistance loss in each section of the pipeline:

[0097] ΔP=λ·(L / D)·(ρv² / 2) where, λ: friction coefficient (look up or calculate based on Reynolds number and pipe roughness); L: pipe length (calculated based on the three-dimensional coordinates of the two endpoints); D: pipe diameter; ρ: medium density; v: medium velocity (a velocity can be assumed in the initial iteration and corrected in subsequent iterations).

[0098] With the starting point as the source point, the pressure value is known (such as the measured value of the pump outlet pressure).

[0099] Dijkstra's algorithm calculates the "path of least resistance" from the source to all other nodes;

[0100] By accumulating the resistance losses along this path, the theoretical pressure values ​​at each node are obtained:

[0101] P(node) = P(starting point) − ΣΔP(each segment of the pipeline in the path)

[0102] If the pipeline network has multiple power sources (such as multiple pumps in parallel), then each starting point is used as the source point for calculation.

[0103] At the confluence point, the final pressure is determined using either a flow-weighted average or the principle of minimum energy.

[0104] Step 3 outputs: Theoretical pressure value for each endpoint; theoretical pressure value for each pressure gauge installation point (interpolated based on the gauge's opening location on the pipeline).

[0105] Step 4: Solve for the flow velocity in each pipe section using Bernoulli's equation.

[0106] 1. Bernoulli's equation in form

[0107] For any two cross-sections in a pipe (such as the pipe inlet and outlet):

[0108] P1 / ρg+v1² / 2g+z1=P2 / ρg+v2² / 2g+z2+h_f

[0109] Where: P is pressure, v is flow velocity, z is elevation, and h_f is friction loss.

[0110] 2. Mass conservation constraint

[0111] At the node, the total inflow = the total outflow (for incompressible fluid); Q = A·v, where A is the cross-sectional area of ​​the pipe and v is the flow velocity.

[0112] 3. Solve simultaneously

[0113] Using the nodal pressures calculated in step 3 as known conditions, we substitute them into Bernoulli's equation to deduce the flow velocity v in each pipe segment.

[0114] If the calculated flow rate deviates from the assumed value by more than a threshold, iterative correction is performed (the drag loss is recalculated, and the process returns to step 3).

[0115] 4. Convergence Judgment

[0116] When the rate of change of flow velocity between two iterations is less than 1%, it is considered to have converged, and the final flow velocity is output.

[0117] Step 4 output: Medium velocity (m / s) for each pipe segment; Volumetric flow rate (m³ / s or m³ / h) for each pipe segment.

[0118] Step 5: Calculate the theoretical value of the instrument using the mathematical model of the component.

[0119] Using mathematical models of each component, the theoretical operating values ​​of instruments such as temperature, pressure, and flow rate are calculated.

[0120] The specific process is as follows:

[0121] 1. Theoretical value of pressure gauge

[0122] The theoretical pressure value at the measuring point location calculated in step 3 is used directly; if the measuring point is located in the middle of the pipeline, it is obtained by linear interpolation of the pressure at both ends.

[0123] 2. Theoretical value of flow meter

[0124] Use the volumetric flow rate of the pipe section calculated in step 4; for flow meters installed on branch pipes, take the calculated flow rate of the corresponding branch.

[0125] 3. Theoretical values ​​of temperature instruments

[0126] For non-heat exchange pipe networks: Assuming the temperature remains constant along the pipes (ignoring heat dissipation), the theoretical temperature equals the measured value of the starting temperature; For heat exchange pipe networks: Call the mathematical model of the heat exchanger to calculate the theoretical temperature after heat exchange, Q=U·A·ΔT_lm (heat transfer equation); Solve for the outlet temperature by combining the heat balance equation.

[0127] 4. Special component handling

[0128] Valves: Calculate the pressure drop across the valve based on the valve opening and Cv value to verify the rationality of the pressure distribution; Orifice Plate: Calculate the theoretical differential pressure based on the orifice plate diameter and flow rate, and compare it with the measured value of the differential pressure transmitter; Pump: Calculate the theoretical outlet pressure based on the pump inlet pressure and flow rate by calling the pump characteristic curve, and verify the measured outlet pressure.

[0129] Step 5 outputs: theoretical pressure value for each pressure gauge; theoretical flow rate value for each flow meter; theoretical temperature value for each temperature gauge.

[0130] Step 6: Output the parsing results.

[0131] All theoretical calculation results are packaged into structured data (such as JSON format) and transmitted to the data storage and analysis module through the fifth data interaction interface. At the same time, the following can be optionally displayed in real time in the 3D visualization interface: pipeline flow direction (arrow animation), flow velocity (color depth or width change), and comparison between theoretical and measured values ​​(dashboard format).

[0132] Example 1: Real-time Leakage Monitoring System Based on Steam Pipeline Network

[0133] Application scenarios

[0134] The steam pipeline network of an industrial park includes one boiler, three steam distribution cylinders, 15 main pipelines, more than 50 valves, 20 pressure measuring points, 15 temperature measuring points and 10 flow measuring points. The total length of the pipeline is about 3 kilometers, and some pipelines are buried underground.

[0135] Implementation steps

[0136] Step 1: Construction of 3D Topology

[0137] A domestic 3D engine was used to construct a 3D model of the steam pipeline network; the boiler outlet was set as the starting point for analysis and each user access point was set as the ending point; the 3D coordinates (x, y, z) of each endpoint were recorded, and the potential energy change caused by the pipeline elevation difference was calculated; the steam medium properties (temperature 220℃, pressure 1.0MPa, density, etc.) were set.

[0138] Step 2: Real-time data access

[0139] Real-time data is obtained from the DCS system via the OPCUA interface, with a refresh cycle of 2 seconds. The data includes: valve opening, pressure transmitter reading, temperature transmitter reading, and vortex flow meter reading.

[0140] Step 3: Mathematical Model Configuration

[0141] Pump / fan model: No pump is used here; boiler outlet pressure is used as the power source. Pipeline resistance model: Darcy-Weisbach formula is used, taking into account pipeline roughness, diameter, and length. Valve model: Local resistance is calculated using the Cv coefficient method. Orifice plate model: Differential pressure flow formula of ISO 5167 is used.

[0142] Step 4: Real-time analysis

[0143] The A* algorithm is used to search for the optimal path from the boiler to each user; the Dijkstra algorithm is used to calculate the theoretical pressure distribution at each point in the pipeline; the theoretical pressure is compared with the actual pressure measurement values ​​to generate a residual sequence.

[0144] Step 5: Deviation Trend Analysis

[0145] The theoretical and measured values ​​of the pressure difference before and after a valve were continuously monitored. It was found that the residual pressure difference continued to increase from the 5th day and exceeded the 3σ threshold on the 8th day. The system issued a "valve internal leakage warning" to prompt maintenance personnel to check the valve.

[0146] Example 2: Early warning of pump performance degradation based on chemical pipeline network

[0147] Application scenarios

[0148] A chemical plant's pipeline network for transporting benzene solution includes 2 centrifugal pumps (one in use and one on standby), 5 branch pipelines, 8 valves, 4 flow meters, and 6 pressure gauges. The medium temperature is ambient, and the design flow rate is 80 m³ / h.

[0149] Implementation steps

[0150] Step 1: Construction of the pump mathematical model

[0151] Obtain the performance curve from the pump manufacturer: H_theory = H0 - kQ² (H_theory is the theoretical head in meters; Q is the flow rate in m³ / h; H0 is the shut-off head in meters; k is the pump resistance coefficient); According to the data provided by the manufacturer, in this embodiment, we take: H0 = 50m, k = 0.002, then H_theory = 50 - 0.002Q²; Store the relevant model parameters in the mathematical model construction module.

[0152] Step 2: Real-time analysis

[0153] Real-time data obtained from the interface: pump outlet pressure 0.8MPa, pump inlet pressure 0.5MPa, flow rate 72m³ / h; calculated actual head: H_actual=(0.8-0.5)×102 / ρ≈34.8m (benzene density 0.88g / cm³).

[0154] Theoretical head: Substituting the actual flow rate of 72 m³ / h into the performance curve, we get H_theory = 50 - 0.002Q² = 39.632 m; Deviation: ΔH = 4.832 m, deviation rate 12.2%.

[0155] Step 3: Historical Trend Analysis

[0156] By tracing back the data from the past 30 days and plotting the deviation change curve, it was found that the deviation rate increased linearly from 2% to 11.6%, with the slope remaining positive. The system judged that the pump performance had deteriorated, and the impeller may have been worn or cavitated.

[0157] Step 4: Early Warning and Response

[0158] The system issued a "pump A performance degradation warning" and suggested switching to the standby pump. After switching, the theoretical value of the standby pump deviated from the actual value by only 1.2%, confirming the original pump failure. During inspection, slight cavitation of the impeller was found, and its performance was restored after repair.

[0159] Example 3: Rapid Identification of Instrument Faults Based on Gas Pipeline Network

[0160] Application scenarios

[0161] A gas pressure regulating station and its outgoing pipeline in a certain city, including a pressure regulator, flow meter, pressure transmitter (two redundant units), and temperature transmitter, supplies gas to 50,000 households.

[0162] Implementation steps

[0163] Step 1: Special processing of topology

[0164] Pressure measuring points P1 and P2 are set before and after the pressure regulator; two pressure transmitters, PT201A and PT201B (redundant), are installed at P2.

[0165] Step 2: Real-time analysis and verification

[0166] Using Bernoulli's equation, the theoretical pressure of P2 is calculated from the pressure, flow rate, and temperature of P1; at the same time, the measured values ​​of PT201A and PT201B are read.

[0167] Step 3: Deviation analysis to determine instrument malfunction

[0168] Data for a certain day: P1 measured: 0.4MPa; flow rate: 5000m³ / h; P2 theoretically calculated: 0.18MPa.

[0169] PT201A measured: 0.17MPa (deviation 5.5%); PT201B measured: 0.12MPa (deviation 33.3%).

[0170] Based on topology analysis: the data from the two measuring points should be basically consistent, and PT201B is obviously abnormal; the system determines that the PT201B instrument is faulty.

[0171] Step 4: Output Early Warning

[0172] The system issued an "Abnormal reading warning for instrument PT201B" and suggested verification or replacement; maintenance personnel conducted an on-site inspection and found that the instrument was drifting at zero point. After calibration, it returned to normal.

Claims

1. A real-time status analysis system based on a three-dimensional pipeline network topology, characterized in that, include: The 3D pipeline topology construction module is used to construct a 3D pipeline topology structure by setting equipment, valves, tanks, orifice plates, etc. in the pipeline as endpoints according to the actual installation diagram of the industrial pipeline network, recording the endpoint codes, names, types, 3D coordinates and connection relationships, and recording pipeline attributes and instrument lists. Real-time data interface module: used to obtain the status and operating parameters of each component in the pipeline network from the real-time operating database, and update all component data before each analysis cycle; Pipeline Component Mathematical Model Construction Module: Used to build mathematical models of components such as pumps, fans, straight pipe sections, valves, and orifice plates, and calculate the theoretical working parameters of the components by combining elevation differences and real-time data; Pipeline analysis module: Based on the three-dimensional topology and real-time data, it uses a path search algorithm to analyze the medium flow direction and velocity of each pipeline segment, and calculates the theoretical working values ​​of each instrument. Data storage and analysis module: Used to store the theoretical and actual values ​​of each instrument in each analysis cycle, perform trend analysis on the deviation, and identify potential safety hazards in the pipeline network.

2. The real-time status analysis system based on a three-dimensional pipeline network topology according to claim 1, characterized in that, The three-dimensional pipeline topology is constructed using a domestic 3D engine, and the three-dimensional coordinates of each component are recorded for calculating pipeline length and potential energy.

3. The real-time status analysis system based on a three-dimensional pipeline network topology according to claim 1, characterized in that, The path search algorithm is the A* path search algorithm, which is used to analyze the flow direction and velocity of the medium in the pipeline. Its heuristic function is represented as g(n) + h(n), where g(n) is the actual distance from the starting point to node n, and h(n) is the estimated distance from node n to the target.

4. The real-time status analysis system based on a three-dimensional pipeline network topology according to claim 1, characterized in that, The pipeline analysis module uses Dijkstra's algorithm to calculate the theoretical pressure value at any point in the pipeline.

5. The real-time status analysis system based on a three-dimensional pipeline network topology according to claim 1, characterized in that, The deviation trend analysis methods include one or more of the following: data sorting, statistical analysis, deviation measurement, and distribution pattern analysis.

6. The real-time status analysis system based on a three-dimensional pipeline network topology according to claim 1, characterized in that, The data acquired by the real-time data interface module includes one or more of the following: valve on / off status, regulating valve opening degree, pump start / stop status, pump power, pipeline pressure, pipeline temperature, and pipeline flow rate.

7. The real-time status analysis system based on a three-dimensional pipeline network topology according to claim 1, characterized in that, In the pipeline component mathematical model construction module, the mathematical models of pumps and fans are derived from existing models, manufacturer data, or mathematical statistics of historical operating data.

8. The real-time status analysis system based on a three-dimensional pipeline network topology according to claim 1, characterized in that, The data storage and analysis module is also used to perform one or more of the following analyses: Dynamic upper and lower limit analysis of parameters; Parameter change rate alarm; Pipeline topology analysis; Add measuring points and verification for key equipment or pipelines.

9. The real-time status analysis system based on a three-dimensional pipeline network topology according to claim 1, characterized in that, The system is used to determine one or more of the following fault types: external pipeline leakage, internal valve leakage, orifice plate wear, pump or fan performance degradation, and instrument failure.

10. The real-time status analysis system based on a three-dimensional pipeline network topology according to claim 1, characterized in that, It also includes a set of data interaction interfaces, which includes: First data interaction interface: Located between the three-dimensional pipeline topology construction module and the real-time data interface module, it is used to transmit component codes, endpoint connection relationships, and instrument list information to the real-time data interface module as a basis for obtaining real-time operating data; The second data interaction interface is located between the real-time data interface module and the pipeline analysis module, and is used to transmit the updated component status and operating parameters to the pipeline analysis module. The third data interaction interface is set between the three-dimensional pipeline topology construction module and the pipeline analysis module. It is used to provide the pipeline analysis module with topological relationships, endpoint coordinates, pipeline attributes and start and end point setting information for path search and elevation difference calculation. The fourth data interaction interface is located between the pipeline component mathematical model construction module and the pipeline analysis module. It is used to provide the mathematical models of each component to the pipeline analysis module for use in the calculation of theoretical working parameters. The fifth data interaction interface is located between the pipeline analysis module and the data storage and analysis module. It is used to transmit the theoretical working values ​​of each instrument in each analysis cycle to the data storage and analysis module for comparison, storage and trend analysis with the actual measured values. The sixth data interaction interface is located between the data storage and analysis module and the three-dimensional pipeline topology construction module. It is used to feed back the information of identified components with potential safety hazards to the three-dimensional pipeline topology construction module so that abnormal components can be highlighted in the three-dimensional visualization interface.

Images