Automatic monitoring data processing method for pipe network pressure

Abstract

The present application relates to the technical field of pipe network data processing, and particularly relates to a kind of automatic monitoring data processing methods of pipe network pressure.The method comprises the following steps: obtaining pipe network structure data;Based on pipe network structure data, pipeline topological network construction is carried out, and pipeline topological network is generated;The main pipeline and branch pipe screening are carried out to the pipeline topological network, and pipe network flow data is obtained;Based on pipe network flow data, main pipeline and branch pipe are deployed with pipe wall microwave resonator, and the circumferential deformation harmonic of pipe is captured in real time through microwave resonator after deployment;The longitudinal wave propagation time delay of pipe network flow data is analyzed, and the pipe body structure dynamic resonance spectrum is constructed in combination with the circumferential deformation harmonic;The pipe network internal pressure and pipe wall equivalent stress of pipeline topological network are calculated through pipe body structure dynamic resonance spectrum.The present application improves the accuracy of pipe network pressure monitoring and identification by combining pipeline topological analysis, microwave resonator monitoring, dynamic resonance spectrum and bidirectional fluid-structure coupling analysis.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of pipe network data processing, and particularly relates to an automatic monitoring data processing method for pipe network pressure. BACKGROUND

[0002] In the early days, pipe network pressure monitoring mainly relied on manual inspection and traditional mechanical instruments, which had low work efficiency and was prone to human errors. This traditional method cannot monitor the running state of the pipe network in real time, and has limited data processing and analysis capabilities, so that potential safety hazards in the operation of the pipe network cannot be discovered in time. The emergence of intelligent sensors and data acquisition technology has gradually realized the automation of pipe network pressure monitoring. In the early stage, the system transmits sensor data to the centralized control center through wired connection, but this method still has problems of wiring difficulty and data transmission delay. Based on the application of big data analysis and artificial intelligence (AI) technology, the pipe network pressure monitoring system can not only monitor the pressure changes in real time, but also can perform predictive analysis to discover potential fault risks in advance, thereby improving the safety and reliability of the pipe network. However, the current traditional system is limited to the measurement of pressure, and cannot effectively calculate the stress distribution of the pipe wall and the pressure in the pipe, and the analysis of the interaction between the complex flow and pressure changes in the pipe network is also low, which leads to low accuracy of pipe network pressure monitoring and identification. SUMMARY

[0003] Therefore, it is necessary to provide an automatic monitoring data processing method for pipe network pressure to solve at least one of the above technical problems.

[0004] To achieve the above-mentioned purpose, an automatic monitoring data processing method for pipe network pressure is provided, which comprises the following steps:

[0005] Step S1: obtaining pipe network structure data; performing pipe topology network construction based on the pipe network structure data to generate a pipe topology network; performing main pipe and branch pipe screening on the pipe topology network to obtain pipe network flow data;

[0006] Step S2: deploying a pipe wall microwave resonator on the main pipe and the branch pipe based on the pipe network flow data, and capturing pipe circumferential deformation harmonics in real time through the deployed microwave resonator; analyzing the longitudinal wave propagation time delay of the pipe network flow data, and constructing a pipe body structure dynamic resonance spectrum in combination with the pipe circumferential deformation harmonics;

[0007] Step S3: performing pipe network internal pressure and pipe wall equivalent stress calculation on the pipe topology network through the pipe body structure dynamic resonance spectrum to obtain pipe internal pressure value and pipe wall equivalent stress value; performing pipe network abnormal pressure discrimination on the pipe network flow data based on the pipe internal pressure value and the pipe wall equivalent stress value to obtain pipe network abnormal pressure detection data;

[0008] Step S4: screening pipe fittings according to the pipe network abnormal pressure detection data to obtain pipe network pipe fitting data; analyzing the two-way fluid-structure coupling of the pipe network pipe fitting data to obtain the water hammer coupling pulse pressure peak value; and accurately positioning the abnormal pipe network pressure area in the pipe network topology network through the pipe network abnormal pressure detection data and the water hammer coupling pulse pressure peak value to realize automatic monitoring of the pipe network pressure.

[0009] The present application can capture the circumferential deformation harmonic of the pipeline in real time by using the microwave resonator deployment technology, realize high sensitivity sensing of the microstructure change of the pipe body, and improve the early warning ability of potential pipe network risks. Combining the longitudinal wave propagation time delay in the flow data with the deformation harmonic, a dynamic resonance spectrum is constructed, which helps to accurately reflect the pipeline operation state, so as to realize accurate calculation of the in-pipe pressure and pipe wall equivalent stress, and provide a high credible basis for abnormal pressure judgment. Combined with the pipe fitting screening result and the two-way fluid-structure coupling analysis, the pulse pressure peak value caused by the water hammer effect can be obtained, which significantly improves the abnormal pressure identification ability under complex working conditions. Combined analysis of the abnormal pressure detection data and the water hammer pulse peak value can accurately locate the pipe section area where the pressure anomaly occurs, avoid large-scale blind investigation, and improve the maintenance efficiency. Based on the complete data flow and the structural mechanics model, the whole process automation from data acquisition, state analysis to abnormal positioning is realized, which is beneficial to the construction of an intelligent urban pipe network management system. Through early identification of abnormal pressure sections, major accidents such as pipe network burst and leakage can be effectively prevented, and the cost of emergency repair and loss compensation can be reduced. Therefore, the present application improves the accuracy of pipe network pressure monitoring and identification by combining pipe topology analysis, microwave resonator monitoring, dynamic resonance spectrum and two-way fluid-structure coupling analysis.

[0010] Preferably, step S1 comprises the following steps:

[0011] Step S11: obtaining a pipe network original engineering drawing and extracting pipe network structure information in the pipe network original engineering drawing to obtain pipe network structure data;

[0012] Step S12: analyzing the node connectivity of the pipe network structure data to generate pipe network topology data;

[0013] Step S13: clustering the flow path of the pipe network topology data to generate candidate main path data;

[0014] Step S14: determining the path weight of the candidate main path data, and dividing the candidate main path data into main pipes and branch pipes according to the determination result to obtain main pipe and branch pipe division data;

[0015] Step S15: extracting the flow logic of the main pipe and branch pipe division data to generate pipe network flow data.

[0016] The application extracts pipe network structure information from original engineering drawings automatically (S11), saves a large amount of manual modeling work, reduces the risk of human error, and ensures the accuracy and integrity of the basic data. The node connectivity analysis (S12) combines the topology generation algorithm to quickly build a pipeline topology network reflecting the real pipeline connection relationship, providing a basic support for subsequent pressure analysis and simulation calculation. The flow path clustering technology (S13) can effectively identify the main flow path, and the path weight judgment (S14) is used for main pipe / branch pipe division, which helps to highlight the core conveying path and enhance the monitoring and management ability of key areas. By extracting the flow logic in the main pipe and branch pipe division result (S15), the structured "pipe network flow data" is generated, which provides an efficient input data source for subsequent pressure modeling, anomaly detection and resonance analysis. The clear division of the main pipe and the branch pipe enables subsequent hierarchical processing of pressure response analysis, water hammer wave propagation modeling, etc. according to different network levels, improving the resolution and response efficiency of the overall system modeling. The generation of pipe network flow data marks the completion of dual modeling of physical topology and flow logic, and provides a key support foundation for pipeline structure resonance analysis, equivalent stress calculation and abnormal area identification.

[0017] Preferably, step S14 comprises the following steps:

[0018] Step S141: performing path weight judgment on the candidate main path data to obtain a judgment result, wherein the path weight includes the geometric length of the path, the flow demand on the path, and the pressure demand of the path;

[0019] Step S142: when the following conditions occur simultaneously, the candidate main path data is judged as main pipe division data: the geometric length of the path is between 500 meters and 5000 meters, the flow demand on the path is between 50 cubic meters / hour and 1000 cubic meters / hour, and the pressure demand of the path is between 0.5 MPa and 5 MPa;

[0020] Step S143: when any of the following conditions occurs, the candidate main path data is judged as branch pipe division data: the geometric length of the path is less than 500 meters, the flow demand on the path is less than 50 cubic meters / hour, and the pressure demand of the path is less than 0.5 MPa or higher than 5 MPa;

[0021] Step S144: integrating the main pipe division data and the branch pipe division data into the main pipe and branch pipe division data.

[0022] The application considers the three key weight factors of geometric length, flow demand and pressure demand (S141) to construct a path judgment system based on real operating parameters, effectively improving the rationality and engineering applicability of main pipe / branch pipe division. By using numerical intervals (such as geometric length 500-5000 meters, flow 50-1000 m 3As a criterion, the pressure range of 0.5-5 MPa (S142-S143) is used to avoid subjective judgment and ensure the consistency, reproducibility and engineering operability of the division process. For large-scale urban pipe networks, the automatic classification of candidate paths is realized through programmed rule judgment (S142, S143), which greatly reduces the workload of manual identification and improves the efficiency of pipe network modeling. The differentiation of main pipes and branch pipes and the unified integration (S144) are helpful to build a clear and logical pipe network structure, which provides visual and controllable support for subsequent structure analysis, flow control and fault isolation. The accurate division of main pipes and branch pipes can provide accurate deployment areas for subsequent flow logic modeling (S15) and microwave resonator deployment (S2), which improves the diagnostic sensitivity and prediction ability of the whole system.

[0023] Preferably, the step S2 of deploying the pipe wall microwave resonator based on the pipe network flow data includes:

[0024] Screening the key nodes of the pipe network flow data;

[0025] Segmenting and mapping the key nodes to generate segmented deployment unit data;

[0026] Analyzing the waveguide adaptability of the segmented deployment unit data, and optimizing the segmented deployment unit data according to the waveguide adaptability to generate resonator configuration scheme data;

[0027] Calibrating the position of the resonator configuration scheme data to generate microwave resonator deployment parameter data;

[0028] Deploying the pipe wall microwave resonator according to the microwave resonator deployment parameter data, thereby obtaining the deployed microwave resonator.

[0029] The application intelligently screens key nodes from pipe network flow data, accurately locates the key area of resonator layout, avoids resource waste, and improves the coverage and response capability of the monitoring system in high-risk sections. The dry and branch segmentation mapping (reasonable partitioning of the main trunk and branch pipes) generates segmented deployment unit data, laying the foundation for subsequent structure matching and frequency planning, and improving the consistency and topological coordination of the deployment structure and the pipe network structure. The segmented unit is subjected to waveguide structure adaptability analysis, and based on the results, multi-point frequency distribution optimization is carried out, and resonator configuration scheme data that meet the resonance characteristics of different pipe sections are constructed, improving the monitoring sensitivity under different material, caliber and flow conditions. Through precise position calibration of the resonator configuration scheme, high-precision microwave resonator deployment parameter data is generated, ensuring the accurate fitting and signal coupling quality of the resonator on the physical pipe wall, and improving the stability and accuracy of signal acquisition. The finally generated deployment parameter data has a structured feature, which can be used to guide the intelligent deployment equipment or construction system to carry out automatic installation, reduce the risk of manual operation, and improve the construction efficiency. After deployment, the microwave resonator becomes the core device for pipe circumferential deformation monitoring, providing high-quality harmonic data source for subsequent resonance spectrum construction and pipe stress state evaluation.

[0030] Preferably, the step S2 of analyzing the longitudinal wave propagation time delay of the pipe network flow data comprises:

[0031] Segmented fitting of flow velocity data is performed on the pipe network flow data to generate in-segment flow velocity data;

[0032] The sound velocity of the in-segment flow velocity data is coupled to obtain medium sound velocity distribution data;

[0033] The medium sound velocity distribution data is topologically mapped according to the pipe network flow data to generate sound wave propagation path graph data;

[0034] The propagation time delay integral of the sound wave propagation path graph data is calculated to generate initial longitudinal wave propagation time delay data;

[0035] The drift anomaly of the longitudinal wave propagation time delay data is detected to obtain the longitudinal wave propagation time delay of the pipe network flow data.

[0036] The application generates high-resolution segment flow rate data by segment fitting of pipe network flow data (such as according to pipe segments, flow states or flow intervals), significantly improves the local accuracy of sound velocity calculation, and provides a solid foundation for modeling the sound wave propagation path. Based on the segmented flow rate data, the sound velocity coupling calculation is performed, the effects of temperature, pressure, medium composition and other factors on the sound velocity are fully considered, and more adaptive medium sound velocity distribution data are generated, which improves the physical authenticity of the propagation time delay calculation. The sound velocity distribution is mapped to the path using the pipe network topology to generate sound wave propagation path data that truly reflect the sound wave propagation path, supporting accurate modeling and visual analysis of the multi-path propagation phenomenon in complex networks. The propagation time delay integration method is used to accumulate and sum the propagation time of each segment in the path graph, avoiding errors caused by simple averaging, and generating high-precision initial longitudinal wave propagation time delay data, which can be used for comparison and detection and model verification. Through the drift anomaly detection of the propagation time delay data, the abnormal working conditions such as local blockage, pipe wall structure change or atypical flow state in the pipe network are identified, and a pre-warning mechanism for pipe network maintenance is provided. The accurately obtained longitudinal wave propagation time delay is one of the dynamic response characteristics, which can be combined with microwave harmonic data for joint analysis, used to construct a pipe body structure dynamic resonance spectrum that integrates time domain and frequency domain features, and enhance the global structure perception ability of the system.

[0037] Preferably, the step S2 of constructing the pipe body structure dynamic resonance spectrum in combination with the pipe circumferential deformation harmonic includes:

[0038] Synchronously calibrating the longitudinal wave propagation time delay data and the pipe circumferential deformation harmonic to generate time-frequency joint reference data;

[0039] Extracting resonance response features of the time-frequency joint reference data;

[0040] Mapping the pipe segments by the resonance response features to generate structure resonance distribution data;

[0041] Performing dynamic energy clustering analysis on the structure resonance distribution data to generate dynamic resonance clustering data;

[0042] Constructing the pipe body structure dynamic resonance spectrum based on the structure resonance distribution data, the dynamic resonance clustering data and the resonance response features.

[0043] The application effectively eliminates asynchronous sampling errors by synchronously calibrating the longitudinal wave propagation time delay data and the circumferential deformation harmonic of the pipeline to form time-frequency joint reference data in a unified time-frequency scale, and provides accurate reference for subsequent feature extraction and atlas construction. Based on the joint reference data, resonance response features can be extracted to capture the response modes of the structure under different frequencies and loading conditions, identify potential structural weaknesses and frequency coupling abnormalities, and improve the reliability of structural health assessment. Mapping the resonance response features back to the actual pipe segment location forms structural resonance distribution data, which can be used to locate the resonance abnormal concentration area and support pipe segment level structural performance analysis and maintenance decision. Dynamic energy clustering analysis is performed on the structural resonance distribution to identify pipe segments with similar dynamic resonance characteristics, and dynamic resonance clustering data is obtained to assist in constructing a "resonance risk layer" or "dynamic resonance mode spectrum". Based on the structural distribution data, clustering results and response features, a pipe body structure dynamic resonance atlas is constructed to realize time-frequency-space multidimensional fusion expression and comprehensively reflect the dynamic operation characteristics of the pipeline structure, with high interpretability and prediction ability. The dynamic resonance atlas can be used to compare and identify abnormal pipeline operation states, structural relaxation, stress concentration areas and other risk positions, and is the core basic data support for intelligent operation and maintenance, automatic monitoring and fault warning.

[0044] Preferably, step S3 comprises the following steps:

[0045] Step S31: frequency inversion is performed on the pipe body structure dynamic resonance atlas to generate resonance frequency-structure coupling data;

[0046] Step S32: stress of the resonance frequency-structure coupling data is calculated to obtain a pipe wall equivalent stress value;

[0047] Step S33: stress distribution mapping is performed on the pipeline topology network according to the pipe wall equivalent stress value to generate pipeline node stress map data;

[0048] Step S34: internal pressure inversion calculation is performed on the pipeline node stress map data in combination with the pipe network flow data to generate a pipeline internal pressure value;

[0049] Step S35: time sequence comparison analysis is performed on the pipeline internal pressure value data and the pipe network flow data to generate pressure time sequence deviation data; joint threshold value discrimination processing is performed on the pressure time sequence deviation data and the pipe wall equivalent stress value to generate pipe network abnormal pressure detection data.

[0050] The application establishes the coupling relationship between the frequency characteristics and the structure attributes by frequency inversion of the pipe body structure dynamic resonance spectrum, generates resonance frequency-structure coupling data, realizes nondestructive sensing and precise modeling of the structure health state. According to the resonance frequency-structure coupling relationship, equivalent stress calculation is carried out, and the pipe wall equivalent stress value obtained can comprehensively reflect the stress response of the structure in the running state, providing a basis for subsequent safety evaluation and residual life prediction. The pipe wall equivalent stress value is mapped to a topological network to generate pipe node stress map data, which can be used to identify stress abnormal concentration areas and realize precise positioning of the structural safety risk of the pipe network. Based on the stress map data and the pipe network flow data, internal pressure inversion calculation is carried out to obtain the pipe internal pressure value, which makes up for the dynamic change characteristics of the running pressure that cannot be reflected by relying on flow or structure observation alone. By comparing and analyzing the internal pressure value and the historical flow data in time sequence, the pressure change trend is identified, the pressure time sequence deviation data is formed, and the atypical flow mode caused by blockage, leakage or gas blockage is found in time. The pressure time sequence deviation data and the pipe wall equivalent stress value are jointly analyzed to construct a multi-factor joint discrimination model, and the pipe network abnormal pressure detection data is output, which enhances the response ability of the system to pressure mutation and fatigue hidden danger. This step realizes the closed-loop process of "structure response→stress calculation→pressure inversion→abnormal identification", and provides theoretical and data support for realizing intelligent diagnosis and online monitoring of the health state of the pipe network.

[0051] Preferably, step S33 comprises the following steps:

[0052] Step S331: grid reconstruction is performed on the pipe wall equivalent stress data to generate pipe wall grid stress data;

[0053] Step S332: spatial projection is performed on the pipe wall grid stress data to generate stress space mapping data;

[0054] Step S333: node fusion is performed on the stress space mapping data and the pipe topology network to generate initial node stress data; direction tensor decomposition is performed on the initial node stress data to generate node principal stress vector data;

[0055] Step S334: full map fusion and connectivity weighting are performed on the node principal stress vector data to generate pipe node stress map data.

[0056] The application reconstructs the grid of the equivalent stress data of the pipe wall, and the generated grid stress data converts the original irregularly distributed stress information into high-resolution data structure with geometric consistency, facilitating subsequent calculation and visualization. The grid stress data is projected in space to obtain stress space mapping data, ensuring the spatial consistency between the stress distribution and the actual geometric shape and layout environment of the pipeline, which helps to accurately restore the physical stress field distribution. The initial node stress data generated by the fusion of the stress space mapping data and the pipeline topology network realizes the direct coupling between the physical stress field and the topology structure nodes, facilitating the formation of the subsequent node-level structure response model. The node principal stress vector data obtained by the directional tensor decomposition of the initial node stress data contains the principal stress size and direction information, which significantly enhances the judgment ability of the stress directionality and potential instability risk of the pipeline. On the basis of the principal stress vector, the whole graph fusion and connectivity weighting are carried out, and the generated pipeline node stress map data not only maintains the local stress change characteristics, but also reflects the stress conduction link of the overall network, which helps to identify the structural weak points or hidden danger paths caused by stress concentration. The finally output node stress map provides a basic input for subsequent internal pressure inversion, fatigue analysis and pipeline network anomaly diagnosis, and has the advantages of high structure precision, high spatial resolution and high physical consistency.

[0057] Preferably, the joint threshold value discrimination processing of the pressure time sequence deviation data and the pipe wall equivalent stress value in step S35 comprises:

[0058] When any of the following conditions occurs, it is determined that there is a sudden pressure fluctuation anomaly, and sudden pressure fluctuation anomaly data is obtained: the pressure change rate within any 5 minutes exceeds 0.5 MPa / min; the single-point pressure value increases or decreases by more than 1.2 MPa within 10 minutes; the pressure fluctuation frequency is greater than 0.8 Hz, and the duration is more than 15 minutes;

[0059] When the following conditions occur simultaneously, it is determined that there is a periodic abnormal pressure fluctuation, and periodic pressure fluctuation anomaly data is obtained: the pressure peak deviation in adjacent monitoring periods exceeds ±20%; the wave peak / trough time interval change rate in three consecutive periods exceeds ±25%; after comparison with the historical baseline waveform, the dynamic time warping deviation exceeds 10%;

[0060] When the following conditions are met simultaneously, it is determined that there is a pipe wall equivalent stress overrun anomaly, and equivalent stress overrun anomaly data is obtained: the pipe wall equivalent stress value exceeds 80% of the design material yield limit, i.e. exceeds 160 MPa, wherein the design material yield limit is 200 MPa; the stress concentration coefficient is higher than 2.5; the equivalent stress sudden increase rate exceeds 15 MPa / 10 min within any 30-minute window, and the abnormal section length exceeds 20 meters;

[0061] The integrated burst pressure fluctuation anomaly data, periodic pressure fluctuation anomaly data and equivalent stress overrun anomaly data are aligned in space position and time stamp to generate pipe network anomaly pressure detection data.

[0062] The present application can accurately identify the burst pressure fluctuation anomaly by monitoring the indicators such as the rate of pressure change, the sharp fluctuation of single-point pressure value and the frequency of pressure fluctuation. The rapid capture of the burst pressure fluctuation anomaly can effectively warn the risk of sudden accidents in the system, such as pipe rupture or burst, and provide timely response at critical moments. By analyzing the periodic change of pressure fluctuation and monitoring indicators such as pressure peak deviation, wave crest / wave trough interval change rate and dynamic time warping deviation, the system can timely alarm when periodic pressure fluctuation occurs, effectively detecting regular fluctuation anomalies caused by pipe fatigue and equipment aging. This provides strong support for hidden danger diagnosis in long-term system operation. By comparing the equivalent stress of the pipe wall with the yield limit, potential damage risks caused by excessive stress on the pipe can be found in time, especially in positions with high stress concentration (such as elbows, joints, etc.) and sudden changes in stress increase rate, which can effectively prevent accidents such as pipe rupture and leakage caused by overload. Through the analysis of the equivalent stress increase rate and stress overrun condition, the overload state in the pipe can be accurately identified. By integrating the burst pressure fluctuation anomaly data, periodic pressure fluctuation anomaly data and pipe wall equivalent stress overrun anomaly data, and aligning them in space position and time stamp, a complete anomaly data atlas of the pipe network system at different time periods and spatial positions can be provided, providing data support for system operation optimization, maintenance decision-making and equipment replacement. For complex pipe network systems, early identification of potential pressure and stress anomalies can provide clear decision support for management personnel, especially before a sudden failure occurs, effectively reducing the risk of pipe network sudden failure and achieving more intelligent preventive maintenance. Joint threshold discrimination processing can classify different types of anomalies, reduce false positives and omissions, improve the accuracy and efficiency of pipe network monitoring, make the anomaly judgment more detailed, and ensure the long-term stable operation of the pipe network.

[0063] Preferably, step S4 comprises the following steps:

[0064] Step S41: analyze the spatial coupling of the pipe network anomaly pressure detection data to generate anomaly node screening data; and perform topological path extraction on the anomaly node screening data;

[0065] Step S42: component recognition of the pipe topological network according to the extracted topological path to generate pipe network component data; and bidirectional fluid-structure coupling modeling of the pipe network component data to generate water hammer effect structure response data;

[0066] Step S43: Extract the time-domain features of the structural response data of water hammer effect to obtain the peak value of water hammer coupled pulse pressure; perform joint localization processing on the peak value of water hammer coupled pulse pressure and the abnormal node screening data to generate abnormal pressure area data of the pipeline network;

[0067] Step S44: Perform adaptive monitoring deployment processing based on the abnormal pressure area data of the pipeline network to realize automatic monitoring of pipeline network pressure.

[0068] This invention accurately identifies anomalous nodes by analyzing the spatial coupling of abnormal pressure detection data in pipeline networks. This process effectively locates the source of pressure anomalies, providing precise target areas for subsequent pipeline maintenance and repair, thereby reducing the risk of pipeline damage. Through topology path extraction and pipeline component identification, key components in the pipeline system can be systematically identified and modeled in detail. This process helps the system comprehensively understand the health status of the pipeline structure and provides support for accurate diagnosis. Through bidirectional fluid-structure interaction modeling, the impact of water hammer on the pipeline structure can be comprehensively analyzed, thereby obtaining water hammer effect structural response data. This analysis plays a crucial role in water hammer effects caused by sudden flow changes, valve operations, etc., in pipeline networks, and can effectively predict and control the structural damage risks caused by water hammer. By extracting the pulse pressure peak of the water hammer effect and jointly locating it with the anomaly node screening data, abnormal pressure areas in the pipeline network can be accurately identified. This process can monitor the abnormal state of the pipeline system in real time and provide timely feedback to maintenance personnel, preventing the escalation of accidents. Based on the analysis of abnormal pressure area data in the pipeline network, the system can perform adaptive monitoring deployment and adjust monitoring strategies according to actual conditions. This flexible monitoring method can automatically adjust the monitoring focus according to the pipeline's operating conditions, ensuring the safety of the pipeline network at each stage of operation. The implementation of automated monitoring significantly reduces the pressure and workload of manual monitoring while ensuring the accuracy of real-time monitoring. Especially when there are significant changes in dynamic pipeline network pressure, automated monitoring can respond quickly and take appropriate action, improving the reliability of pipeline network operation. Attached Figure Description

[0069] Figure 1 A flowchart illustrating the steps of an automatic monitoring and data processing method for pipeline pressure;

[0070] Figure 2 for Figure 1 A detailed flowchart illustrating the implementation steps of step S3.

[0071] Figure 3 for Figure 1 A detailed flowchart illustrating the implementation steps of step S4.

[0072] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0073] The technical method of the present application will be described clearly and completely below in conjunction with the drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0074] In addition, the drawings are only schematic illustrations of the present application and are not necessarily drawn to scale. Identical reference numerals in the drawings represent identical or similar parts, and thus repeated descriptions thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities, which do not necessarily have to correspond to physically or logically independent entities. The functional entities can be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0075] It should be understood that although the terms "first", "second", etc. can be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, without departing from the scope of the exemplary embodiments, a first element can be called a second element, and similarly a second element can be called a first element. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0076] To achieve the above-mentioned purpose, please refer to Figures 1 to 3 An automatic monitoring data processing method of pipe network pressure, the method comprising the following steps:

[0077] Step S1: acquiring pipe network structure data; constructing a pipe topology network based on the pipe network structure data; filtering the main pipe and branch pipe of the pipe topology network to obtain pipe network flow data;

[0078] Step S2: deploying a pipe wall microwave resonator on the main pipe and branch pipe based on the pipe network flow data, and capturing a pipe circumferential deformation harmonic in real time through the deployed microwave resonator; analyzing the longitudinal wave propagation time delay of the pipe network flow data, and constructing a pipe body structure dynamic resonance spectrum in combination with the pipe circumferential deformation harmonic;

[0079] Step S3: calculating the pipe network internal pressure and pipe wall equivalent stress of the pipe topology network through the pipe body structure dynamic resonance spectrum to obtain the pipe internal pressure value and pipe wall equivalent stress value; discriminating the pipe network flow data based on the pipe internal pressure value and pipe wall equivalent stress value to obtain pipe network abnormal pressure detection data;

[0080] Step S4: According to the pipe network abnormal pressure detection data, the pipe fittings of the pipe network topology are screened to obtain pipe network pipe fitting data; the two-way fluid-solid coupling of the pipe network pipe fitting data is analyzed to obtain the water hammer coupling pulse pressure peak value; and the abnormal pipe network pressure area in the pipe network topology is accurately located through the pipe network abnormal pressure detection data and the water hammer coupling pulse pressure peak value, so as to realize automatic monitoring of the pipe network pressure.

[0081] By using the microwave resonator deployment technology, the circumferential deformation harmonic of the pipeline can be captured in real time, the high sensitivity sensing of the microstructure change of the pipe body is realized, and the early warning ability of the potential pipe network risk is improved. The longitudinal wave propagation time delay in the flow data is combined with the deformation harmonic to construct a dynamic resonance spectrum, which helps to accurately reflect the pipeline operation state, so as to realize accurate calculation of the in-pipe pressure and the equivalent stress of the pipe wall, and provide a high credible basis for abnormal pressure judgment. Combined with the pipe fitting screening result and the two-way fluid-solid coupling analysis, the pulse pressure peak value caused by the water hammer effect can be obtained, and the abnormal pressure identification ability under complex working conditions is significantly improved. The abnormal pressure detection data and the water hammer pulse peak value are combined and analyzed to accurately locate the pipe section area where the pressure anomaly occurs, avoid blind investigation in a large range, and improve the maintenance efficiency. Based on the complete data flow and the structural mechanics model, the whole process automation from data acquisition, state analysis to abnormal positioning is realized, which is beneficial to the construction of an intelligent urban pipe network management system. Through early identification of the abnormal pressure section, the occurrence of major accidents such as pipe network burst and leakage can be effectively prevented, and the emergency repair and loss compensation cost can be reduced. Therefore, by combining the pipe network topology analysis, the microwave resonator monitoring, the dynamic resonance spectrum and the two-way fluid-solid coupling analysis, the accuracy of the pipe network pressure monitoring and identification is improved.

[0082] In the embodiment of the present application, as shown in the reference Figure 1 The pipe network pressure automatic monitoring data processing method includes the following steps:

[0083] Step S1: Obtain pipe network structure data; based on the pipe network structure data, a pipe network topology is constructed to generate a pipe network topology; the main pipe and branch pipe of the pipe network topology are screened to obtain pipe network flow data;

[0084] Step S2: Based on the pipe network flow data, a pipe wall microwave resonator is deployed on the main pipe and branch pipe, and the circumferential deformation harmonic of the pipe is captured in real time through the deployed microwave resonator; the longitudinal wave propagation time delay of the pipe network flow data is analyzed, and a pipe body structure dynamic resonance spectrum is constructed in combination with the circumferential deformation harmonic;

[0085] Step S3: The pipe network internal pressure and pipe wall equivalent stress are calculated by the pipe body structure dynamic resonance spectrum, and the pipe network internal pressure value and pipe wall equivalent stress value are obtained; the pipe network flow data is judged based on the pipe network internal pressure value and pipe wall equivalent stress value, and the pipe network abnormal pressure detection data is obtained;

[0086] Step S4: The pipe fittings of the pipe network topology are screened according to the pipe network abnormal pressure detection data, and the pipe network fitting data is obtained; the two-way fluid-solid coupling of the pipe network fitting data is analyzed, and the water hammer coupling pulse pressure peak value is obtained; the abnormal pipe network pressure area in the pipe network topology is accurately positioned through the pipe network abnormal pressure detection data and the water hammer coupling pulse pressure peak value, so as to realize the automatic monitoring operation of the pipe network pressure.

[0087] In the embodiments of the present application, complete pipe network structure data is obtained through a SCADA system, a BIM model or a GIS platform, including pipe diameter, pipe length, pipe material, installation year, interface type, pressure grade and the like, and historical operation data is obtained from flow meters, pressure sensors and the like. Data preprocessing is performed using Python or MATLAB to clean up missing values and abnormal values. Then, a pipe topology network is constructed based on pipe network nodes and pipe segment information using a graph theory library such as NetworkX, each pipe segment being taken as an edge of a directed graph and a node being taken as a vertex of the graph, and being given attribute weights such as flow and pipe diameter. Then, the network is divided by flow through a maximum flow minimum cut algorithm, and a K-means clustering method is combined to classify pipe segments with dense flow and wide connection range as main pipes and the rest as branch pipes, so as to obtain pipe network flow data reflecting the real flow condition of the pipe network. Microwave resonators are uniformly arranged on the outer walls of the main pipes and key branch pipes selected, specifically: high-sensitivity microwave resonators with a working frequency of 5-10 GHz are selected, and one is arranged every 5-20 meters, and the resonators are connected to a monitoring center through wireless means. The resonators detect the frequency shift of reflected waves by emitting microwave signals and capturing harmonic changes caused by pipe wall deformation. At the same time, a reference resonator is used to collect environmental noise to eliminate noise from the data of other nodes. The collected data is subjected to fast Fourier transform (FFT) to extract the main harmonic components. The longitudinal wave propagation time delay in the flow data is calculated by the time difference of water hammer signal propagation between nodes, and the cross-correlation algorithm is used to improve the time delay measurement accuracy. The circumferential deformation data and the longitudinal wave propagation time delay are aligned on the time axis, and empirical mode decomposition (EMD) is used to eliminate seasonal and periodic interference, and finally agile spectrum analysis, Hilbert-Huang transform and the like are used to construct a highly dynamic pipe structure resonance spectrum. The internal pressure and equivalent stress are inverted from the resonance spectrum. Specifically: a resonance frequency-stress-pressure parameter equation based on material mechanics is established, the basic parameters of the pipe, such as material properties, thickness and diameter, are input, the frequency shift is substituted into the model, and the internal pressure is solved. Then, a finite element simulation software (such as ANSYS) is used to discretize the pipe segment and apply loads, and the equivalent stress field distribution of each monitored pipe segment is solved by multi-physical field coupling. The pressure and stress values form time series data, and based on an ARIMA prediction model, an LSTM neural network and the like, the expected pressure prediction value is compared with the current value to identify abnormal fluctuations to form abnormal pressure detection data, and a warning threshold is set for immediate warning. Using the abnormal pressure detection data, the pipe fittings in the abnormal area are selected by comparing the pipe network topology to form a pipe network pipe fitting list, and valves, elbows, tees and supports are classified and marked.Based on these abnormal areas, a two-way fluid-structure coupling model containing fluid domain and solid domain is constructed, CFD software such as Fluent is used to simulate fluid impact, ANSYS Motion is used to simulate pipe wall response, and the pulse pressure peak caused by water hammer effect and its propagation rate and attenuation in the pipe wall are calculated. Using the reverse propagation algorithm of pulse pressure, the abnormal point is traced back to the source, so as to accurately locate the abnormal pressure source. Finally, the specific location and abnormal type are pushed to the pipe network monitoring system to realize automatic monitoring and maintenance suggestion.

[0088] Preferably, step S1 comprises the following steps:

[0089] Step S11: Obtain the original engineering drawing of the pipe network and extract the pipe network structure information in the original engineering drawing of the pipe network to obtain pipe network structure data;

[0090] Step S12: Perform node connectivity analysis on the pipe network structure data to generate pipe topology network data;

[0091] Step S13: Perform flow path clustering on the pipe topology network data to generate candidate main path data;

[0092] Step S14: Perform path weight determination on the candidate main path data, and divide the candidate main path data into main pipes and branch pipes according to the determination result to obtain main pipe and branch pipe division data;

[0093] Step S15: Extract the flow logic of the main pipe and branch pipe division data to generate pipe network flow data.

[0094] In the embodiment of the present application, the original engineering drawings of the pipe network are obtained, which includes CAD drawings, PDF format pipe layout drawings, BIM models or GIS data. Image recognition algorithms (such as OpenCV combined with deep learning models) are used to structure the extraction of key elements in the drawings, such as pipe segments, nodes, manholes, valves, etc. For CAD drawings, line segment and text object information can be obtained through DWG / DXF analysis tools, and parsed into node-pipe segment models; for BIM data, the corresponding pipe segment component information is directly extracted from the IFC structure. After processing, structured pipe network structure data is formed, including node number, pipe segment number, pipe diameter, length, material, node type and other fields. Based on the above pipe segment and node data, node connectivity analysis is performed to construct a directed graph structure. Each pipe segment is regarded as an edge, and the node is regarded as the vertex of the graph. According to the flow direction information, the edge is assigned a direction attribute. Graph traversal algorithms (such as depth-first search DFS or breadth-first search BFS) are used to identify isolated nodes and ring structures, and automatically repair broken segments. Finally, the pipe topology network data with spatial and directional attributes is formed, which can be visualized and analyzed using NetworkX or Gephi graph analysis tools. The constructed topology network is clustered for flow path. First, real-time or historical flow data is obtained, and the average flow per unit time is calculated for each edge and normalized. Based on the shortest path algorithm on the graph and the multi-source point flow weight, the density-based clustering algorithm (such as DBSCAN or SpectralClustering) is used to cluster the high-flow and continuous path segments into candidate trunk path data. This process can be combined with historical water supply, supply area planning and other information for auxiliary correction. The path weight of the candidate trunk path data is determined. The path weight calculation formula is defined, for example: Wi=α·Fi+β·Li+γ·Di; Wherein, Wi is the path weight, Fi is the total flow of the path, Li is the total length of the path, Di is the average pipe diameter, and α, β, γ are adjustable empirical weight coefficients. By setting a threshold, paths with a weight value higher than the threshold are selected as the main trunk, and the rest are branch pipes, and finally the main and branch pipe division data is generated, and each pipe segment is labeled with main / branch attributes. The flow logic relationship in the main and branch pipe division data is extracted, that is, the start point, end point, passing node, whether connected to other main / branch pipes, and other topological logic of each path are recorded. Combined with the directional attribute, historical hydraulic data and dispatching strategy, a directed flow graph model is constructed, and output as standard format pipe network flow data, supporting subsequent hydraulic simulation and dynamic monitoring.

[0095] Preferably, step S14 comprises the following steps:

[0096] Step S141: path weight determination is performed on the candidate trunk path data to obtain a determination result, wherein the path weight includes the geometric length of the path, the flow demand on the path, and the pressure demand of the path;

[0097] Step S142: When the following conditions occur simultaneously, the candidate trunk path data is determined as the trunk pipe division data: the geometric length of the path is between 500 meters and 5000 meters, the flow demand on the path is between 50 cubic meters / hour and 1000 cubic meters / hour, and the pressure demand of the path is between 0.5 MPa and 5 MPa;

[0098] Step S143: When any of the following conditions occurs, the candidate trunk path data is determined as the branch pipe division data: the geometric length of the path is less than 500 meters, the flow demand on the path is lower than 50 cubic meters / hour, and the pressure demand of the path is lower than 0.5 MPa or higher than 5 MPa;

[0099] Step S144: The trunk pipe division data and the branch pipe division data are integrated into the trunk and branch pipe division data.

[0100] In the embodiment of the present application, the path weight parameter of each path in the candidate trunk path data is calculated, including: using the geographic coordinates between nodes, the total length of the path is calculated by the geodesic distance formula or GIS measurement function (such as PostGIS ST_Length); in combination with the real-time flow monitoring data or historical flow records of the SCADA system, the average flow of each pipe section in the path is counted, and the total demand value is calculated; based on the hydraulic model simulation (such as EPANET or self-developed hydraulic calculation module) or the measured pressure data, the pressure fluctuation range of the path in different operation periods is analyzed, and the average or peak pressure required for operation is extracted. A set of multi-condition joint discrimination criteria is set. The system retrieves the candidate path data in turn and judges whether it meets the following three conditions at the same time: geometric length ∈ [500 meters, 5000 meters]: reflecting that the path has certain scale and cross-regional transmission capacity; flow demand ∈ [50 cubic meters / hour, 1000 cubic meters / hour]: indicating that the path is in the medium and high flow level and has the function of conveying main force; pressure demand ∈ [0.5MPa, 5MPa]: belonging to the normal water supply pressure range, meeting the requirements of main stable pressure supply. If the three conditions are met at the same time, the system will classify the path as main pipeline division data and label it as "main pipeline", and record the judgment reason and key parameter value for subsequent tracing. The system judges whether any of the following conditions is true: geometric length < 500 meters: usually represents local connection or end branch; flow demand < 50 cubic meters / hour: indicates that the daily flow is small and cannot carry large-scale transmission; pressure demand < 0.5MPa or > 5MPa: indicating that the pressure is abnormal and is not suitable for stable operation as a main pipe. If any of the above conditions is met, the path is directly identified as branch pipe division data and labeled as "branch pipe". The "main pipeline" path and "branch pipe" path are integrated and merged by the system to form the final main pipe and branch pipe division data. Each path is attached with a judgment label, original weight parameter, classification basis and judgment confidence index for reference and verification by the upstream module or operation and maintenance personnel.

[0101] Preferably, the step S2 of deploying the microwave resonator on the main pipeline and the branch pipe based on the pipe network flow data comprises:

[0102] Screening key nodes of the pipe network flow data;

[0103] Segmenting and mapping the key nodes into trunk and branch segments to generate segment deployment unit data;

[0104] Analyzing the waveguide adaptability of the segment deployment unit data, and optimizing the multi-point frequency distribution of the segment deployment unit data according to the waveguide adaptability to generate resonator configuration scheme data;

[0105] Calibrating the position of the resonator configuration scheme data to generate microwave resonator deployment parameter data;

[0106] A pipe wall microwave resonator is deployed according to the microwave resonator deployment parameter data, thereby obtaining a deployed microwave resonator.

[0107] In the embodiment of the present application, by analyzing the pipe network topology and flow data, nodes with strong flow fluctuation, flow rate mutation, frequent pipe diameter change and historical abnormal point concentration are extracted. The sliding window technology is combined with the coefficient of variation (CV) and the volatility rate index to model the pressure and flow rate change trend of each node, thereby identifying the key node set. In addition, graph neural network (GNN) can be introduced to enhance the node importance evaluation, taking into account the structure topology and physical changes. For the key nodes, upstream and downstream splitting is performed along the pipeline path, and the entire main pipe or branch pipe is divided into multiple segmented deployment units. Each unit is composed of two key nodes or a path between a key node and an end node, and records its pipe length, pipe diameter, material, direction, surrounding environment and other attributes. This process can be completed on the GIS platform through depth-first traversal combined with buffer analysis. In each deployment unit, a simplified propagation model of the resonant cavity is established, and whether the waveguide boundary conditions meet the microwave resonance propagation requirements is calculated. The waveguide adaptability index (such as Q value, coupling efficiency, reflection loss) of each segment of pipe wall is analyzed through numerical simulation (such as COMSOL or HFSS). According to the waveguide adaptation result, the microwave frequency distribution is optimized to a non-equidistant deployment structure, and a multi-objective optimization method based on genetic algorithm (GA) or particle swarm optimization (PSO) is used to realize the optimal matching of deployment frequency, interval and direction, thereby generating resonator configuration scheme data, avoiding frequency overlap and harmonic interference. The ideal deployment points in the configuration scheme are mapped to the actual construction points, and the space around the deployment points is modeled in combination with three-dimensional laser scanning or BIM model, and factors such as construction accessibility, terrain obstruction, water level fluctuation, maintenance safety, etc. are considered. The three-dimensional space repositioning of the deployment points is realized by using local adjustment algorithm (such as ICP algorithm combined with spatial grid calibration). Finally, the deployment parameter data of each resonator is formed, including specific position coordinates (X, Y, Z), deployment direction, installation bracket specification, cable wiring suggestion, etc. The deployment parameters are input into the construction platform, and high-precision RTK or total station is used for field setting out. The microwave resonator unit is installed on site by using magnetic type, buckle type or bracket fixed type, and the frequency response of each resonator is tested by using a portable network analyzer to ensure that the resonant frequency is consistent with the simulation value. After installation, each resonator data node is connected to the data acquisition system through wireless or LoRa networking to form a deployed microwave resonator network that can be used for subsequent deformation detection.

[0108] Preferably, the step S2 of analyzing the longitudinal wave propagation time delay of the pipe network flow data comprises:

[0109] The flow velocity segmentation fitting is performed on the pipe network flow data to generate the in-segment flow velocity data;

[0110] The sound velocity of the in-segment flow velocity data is coupled to obtain the medium sound velocity distribution data;

[0111] The medium sound velocity distribution data is topologically mapped according to the pipe network flow data to generate the sound wave propagation path graph data;

[0112] The propagation time delay integration of the sound wave propagation path graph data is calculated to generate the initial longitudinal wave propagation time delay data;

[0113] The drift anomaly of the longitudinal wave propagation time delay data is detected to obtain the longitudinal wave propagation time delay of the pipe network flow data.

[0114] In the embodiment of the application, the flow data (pressure, flow velocity, temperature) of the entire pipeline is divided according to the topological structure and nodes, and the continuous flow segment is extracted for flow velocity modeling. The segmented polynomial regression model (such as third-order spline fitting or segmented Bezier function) is used to fit the flow velocity data of each segment to overcome the non-stationarity and local anomaly of the actual data. The in-segment flow velocity function vi(t,x) under the time-space distribution of each segment is output to form the in-segment flow velocity data set. Based on the flow velocity data of each segment and the thermal parameters such as temperature and pressure, the local medium sound velocity c i is calculated by combining the sound velocity derivation model in the ideal gas state equation or Navier-Stokes equation. Wherein γ is the specific heat ratio, P i is the pressure, and ρ i is the density, which can be inversely calculated from the flow velocity and temperature. For non-ideal working conditions, a correction coefficient or numerical simulation can be used to solve the coupling model to output the medium sound velocity distribution data to reflect the wave propagation ability of different pipe segments. In combination with the pipeline topological graph, the directed weighted graph model G(V,E) is constructed according to the flow direction and node connection relationship, wherein the weight of each edge e ij is the inverse of the medium sound velocity function on the corresponding pipe segment, i.e. 1 / c i (x), which is used as the propagation delay factor. Through topological traversal (such as Dijkstra algorithm or dynamic path planning), the paths from all starting nodes to terminal nodes are mapped to generate the sound wave propagation path graph data, i.e. the effective propagation path and the sound velocity distribution sequence between each pair of nodes. The integral delay estimation is performed on the pipe segments on each path. The propagation time delay can be defined by the following formula: In combination with the actual sound velocity function interpolation, the numerical integration method (such as Simpson integration or Romberg integration) is used to solve the propagation time of each path in the propagation path graph to generate the initial longitudinal wave propagation time delay data matrix τ ij, which represents the P-wave propagation time from node i to j. Time series analysis is performed on the initial propagation delay data matrix, and the delay residual sequence is compared with the reference standard model. By applying methods such as Z-score detection, CUSUM (cumulative sum control chart), or LSTM prediction residual model, drift abnormal points in the delay data are identified, especially focusing on mutations, trend drifts, or non-periodic jumps. After detecting the abnormality, the drift part is filtered and corrected (such as Kalman filtering or exponential smoothing), and finally the longitudinal wave propagation delay data with reasonable structure and controlled error is output as the core input parameter for subsequent dynamic resonance modeling.

[0115] Preferably, the step S2 of constructing the pipe body structure dynamic resonance spectrum in combination with the pipe circumferential deformation harmonic includes:

[0116] Synchronously calibrate the longitudinal wave propagation delay data and the pipe circumferential deformation harmonic to generate time-frequency joint reference data;

[0117] Extract the resonance response characteristics of the time-frequency joint reference data;

[0118] Map the pipe segments by the resonance response characteristics to generate structure resonance distribution data;

[0119] Perform dynamic energy clustering analysis on the structure resonance distribution data to generate dynamic resonance clustering data;

[0120] Construct the pipe body structure dynamic resonance spectrum based on the structure resonance distribution data, the dynamic resonance clustering data, and the resonance response characteristics.

[0121] In the embodiments of the present application, the circumferential deformation harmonics of the pipeline are obtained from the raw data collected in real time by the microwave resonator, and the frequency variation thereof reflects the slight deformation behavior of the pipe wall under fluid disturbance and internal pressure fluctuation. At the same time, the longitudinal wave propagation time delay data are obtained by the high-precision sensor array arranged longitudinally to accurately depict the space-time characteristics of fluid disturbance propagation at different positions of the pipeline. The time alignment of the longitudinal wave propagation time delay data and the deformation harmonic signal is performed using a high-precision clock synchronization module (such as a PTP protocol device). In order to avoid the error introduced by the time difference, the time characteristic points (such as the wave peak and the harmonic main frequency change point) of the two signal sources are calibrated using the cross-correlation algorithm; in the frequency domain, the two signals are converted into a unified time-frequency two-dimensional matrix using the short-time Fourier transform (STFT), and interpolation processing is performed to form a time-frequency joint reference data with unified sampling rate and resolution. The data generated in this step retains the time correlation and frequency response characteristics of the circumferential harmonics and the longitudinal wave propagation, and is the core basic data for resonance analysis. The empirical mode decomposition (EMD) and Hilbert transform are performed on the above joint data to extract each order intrinsic mode function (IMF), and identify the characteristic parameters such as the natural resonance frequency, the excitation frequency, the attenuation rate, and the quality factor (Q value) in the signal. The principal component analysis (PCA) or independent component analysis (ICA) method is used for dimension noise reduction to retain the main resonance frequency variation and energy distribution characteristics. Then, the frequency drift trajectory is extracted using the local maximum value tracking algorithm to obtain the resonance response spectrum, which is used as the response index for subsequent structure mapping. The spatial geometric coordinates recorded in the GIS or BIM platform are used to bind each group of harmonic response characteristics to its physical space position. Through the response-position mapping matrix, the resonance characteristics are projected onto the actual pipe segment to form a spatialized structure resonance distribution map. Combined with the material properties (such as elastic modulus, wall thickness, and density) and historical operation data, the finite element assisted mapping is used to construct a structure resonance distribution data set containing indicators such as resonance frequency, energy density, and wave propagation speed. The data is organized in the form of a spatial grid and can be regarded as a multi-layer data cube. The time series in the structure resonance distribution data are divided into sliding windows (for example, 10-second sliding windows), and the local harmonic energy density, energy change rate, and frequency jitter amplitude are calculated in each window. The time series clustering algorithm (such as K-shape, DTW-KMeans) or spectral clustering method is used to classify these local structures to identify the position groups in which frequent resonance or abnormal coupling occurs in the pipe segment. After the clustering results are expressed in the form of heat map or flow chart, the dynamic resonance clustering data are formed, which clearly indicate which areas have dynamic coupling strengthening phenomenon or long-term vibration trend. The three types of key data, (1) structure resonance distribution data, (2) dynamic resonance clustering data, and (3) resonance response characteristics, are used to model the dynamic structure response atlas of the entire pipeline using the graph neural network (GNN) framework.Each pipe segment node is a point in the graph, and the connecting pipe segment is an edge. The node attributes include frequency, stress state, energy concentration, etc. The edge attributes include time delay, resonance propagation path, etc. The potential structural coupling hotspots and implicit fatigue sections are inferred by graph generation algorithms (such as GCN or GraphSAGE), and the resonance state is dynamically refreshed.

[0122] Especially important is that the dynamic energy clustering analysis of the structural resonance distribution data also includes:

[0123] Performing frequency spectrum analysis on the structural resonance distribution data to generate resonance spectrum data;

[0124] Calculating the energy density of the resonance spectrum data;

[0125] Performing dynamic time clustering on the resonance spectrum data according to the energy density to generate dynamic time-frequency clustering data;

[0126] Extracting clustering features of the dynamic time-frequency clustering data, wherein the clustering features include peak frequency and amplitude;

[0127] Extracting energy distribution of the structural resonance distribution data through the clustering features to generate dynamic resonance clustering data.

[0128] In the embodiments of the present application, the structural resonance data is obtained, usually from sensor measurements, containing structural response data under different time periods and different vibration modes. The resonance distribution data is denoised, filtered and standardized to eliminate external noise and environmental interference, ensuring the accuracy of the analysis results. Fast Fourier Transform (FFT) or other frequency spectrum analysis methods are applied to convert the time-domain resonance distribution data into frequency-domain data. FFT can convert signals from the time domain to the frequency domain, revealing the frequency components contained in the vibration signal. Through frequency spectrum analysis, a spectrum diagram containing frequency and amplitude is obtained. The spectrum diagram reflects the resonance response of the structure under different frequencies. The energy density corresponding to each frequency point is calculated using the power spectral density method. PSD reflects the energy distribution of different frequency components, which can help understand the vibration energy of the structure under different frequencies. The energy density formula is: where E(f) is the energy density at frequency f, X(f) is the Fourier transform amplitude at that frequency, and Δf is the frequency resolution. From the calculated energy density data, an energy density map is generated. This map shows the energy distribution across different frequency bands of the structure, which aids further analysis and clustering. The energy density data is time-clustered using the Dynamic Time Warping (DTW) algorithm. DTW can handle the alignment of similar patterns in time series, and is particularly suitable for signals with time variations and phase differences. The energy density-based spectral data is dynamically time-clustered using clustering algorithms such as K-means, DBSCAN, hierarchical clustering, etc. By comparing the resonance responses across different time periods, similar spectra are grouped together, generating different dynamic time-frequency clustering data. The clustering process classifies the resonance spectral data of the structure across different time periods into multiple clusters, with each cluster representing a resonance mode with similar dynamic characteristics. The output clustering data includes the clustering labels and related time-frequency features for each time period. For each cluster, the peak frequency in its spectrum is extracted. The peak frequency typically represents the main resonance frequency of the structure under that cluster mode. The peak amplitude of each cluster is extracted, reflecting the vibration intensity at that frequency. The larger the amplitude, the stronger the resonance response at that frequency band. A feature vector is generated for each cluster, containing information such as peak frequency and amplitude. These features can be used for subsequent analysis or model training. Based on the extracted clustering features (such as peak frequency, amplitude, etc.), the energy distribution of each cluster is extracted. By calculating the energy density and frequency response of each cluster, the energy distribution of the structure under different modes can be obtained. Combining the energy features of each cluster with the frequency response, a description data containing the energy distribution of each resonance mode is generated. Combining the energy distribution with the time-frequency features, complete dynamic resonance clustering data is formed. These data will help analyze the resonance modes of the structure under different working conditions and the corresponding energy concentration areas.

[0129] As an example of the present application, reference is made to Fig. 1, which shows a schematic diagram of a pipe network topology, and Fig. 2, which shows a schematic diagram of a pipe network topology, in which the step S3 in the present example comprises: Figure 2

[0130] Step S31: Frequency inversion of the pipe body structure dynamic resonance spectrum is performed to generate resonance frequency-structure coupling data;

[0131] Step S32: Stress of the resonance frequency-structure coupling data is calculated to obtain pipe wall equivalent stress values;

[0132] Step S33: Stress distribution mapping of the pipe topology network is performed according to the pipe wall equivalent stress values to generate pipe node stress map data;

[0133] Step S34: Internal pressure inversion calculation is performed on the pipe node stress map data in combination with the pipe network flow data to generate pipe internal pressure values; ​

[0134] Step S35: Time series comparison analysis is performed on the pipe internal pressure value data and the pipe network flow data to generate pressure time series deviation data; joint threshold discrimination processing is performed on the pressure time series deviation data and the pipe wall equivalent stress value to generate pipe network abnormal pressure detection data.

[0135] In the embodiment of the present application, the dynamic response data of the pipe body structure is collected at different positions of the pipeline by vibration sensors or accelerometers. The sensor positions are ensured to cover the key points of the pipeline, especially the high-risk areas. The collected dynamic response signals are subjected to frequency spectrum analysis (such as Fourier transform) to generate the dynamic resonance spectrum of the pipe body. This step reveals the response characteristics of the pipeline at different frequencies. The resonance spectrum is subjected to frequency inversion using the structure vibration theory and inverse problem solving method. The resonance frequency of the pipeline structure and its coupling relationship with the structure are determined through inversion analysis to generate “resonance frequency-structure coupling data”. According to the physical characteristics (such as material characteristics, geometric shape, wall thickness, etc.) of the pipeline and the frequency inversion results, a dynamic structure coupling model of the pipeline is established. The stress distribution of the pipeline at different frequencies is calculated through finite element analysis (FEA) or analytical mechanics method using the coupling relationship between the resonance frequency and the pipeline structure. The maximum stress of the pipe wall is extracted from the stress distribution, and the equivalent stress value of the pipe wall is obtained by weighting the stress generated by different frequencies. This data reflects the stress response of the pipeline under dynamic load. A topological network model of the pipeline is constructed to identify the key nodes and branches of the pipeline. The equivalent stress value of the pipe wall is mapped to each node in the pipeline topological network. Through mechanical analysis, the stress value of each node is calculated with the stress transmission relationship of the surrounding pipeline to obtain the distribution of the pipeline node stress. The pipeline node stress map data is generated using a visualization tool to clearly display the stress intensity of the pipeline at each node, providing a basis for subsequent analysis. The fluid flow, flow rate, temperature, and other data of the pipe network are collected. These data are helpful for inferring the fluid state inside the pipeline. An internal pressure inversion model is established to combine the pipeline node stress map data and the pipe network flow data to solve the internal pressure of the pipeline through inversion calculation. The actual pressure inside the pipeline is estimated through inversion method, considering the fluid dynamics and the interaction between the fluid and the pipeline wall, to obtain the distribution data of the internal pressure of the pipeline. The internal pressure data of the pipeline and the pipe network flow data are compared in time series to analyze the difference between the pressure change trend and the flow, to generate pressure time series deviation data. This step reveals the abnormal fluctuation of the pipeline pressure with time. Joint threshold discrimination processing is performed on the pressure time series deviation data and the pipe wall equivalent stress value. By setting a threshold, the relationship between the pressure deviation and the stress is determined to identify potential pipeline abnormalities. According to the discrimination results, pipe network abnormal pressure detection data is generated to identify the problematic pipeline areas (such as local pressure abnormalities, pipeline damage, etc.), providing decision basis for further pipeline maintenance and monitoring.

[0136] Preferably, step S33 comprises the following steps:

[0137] Step S331: grid reconstruction is performed on the pipe wall equivalent stress data to generate pipe wall gridded stress data;

[0138] Step S332: spatial projection is performed on the pipe wall gridded stress data to generate stress spatial mapping data;

[0139] Step S333: node fusion is performed on the stress spatial mapping data and the pipeline topology network to generate initial node stress data; directional tensor decomposition is performed on the initial node stress data to generate node principal stress vector data;

[0140] Step S334: full map fusion and connectivity weighting are performed on the node principal stress vector data to generate pipeline node stress map data.

[0141] In the embodiments of the present application, the equivalent stress data of the pipe wall is reconstructed by selecting an appropriate meshing technique (for example, finite element method, triangular meshing method or hexahedral meshing method). This step needs to ensure that the mesh is fine enough to ensure the accuracy of the stress data at each node. According to the geometry of the pipeline and the actual situation of the pipe wall, the equivalent stress data is divided into grid nodes using the meshing algorithm. Each grid node will obtain a stress value, which is obtained by interpolation and reconstruction from the original equivalent stress data of the pipe wall. Through the stress value of each grid node obtained after mesh reconstruction, the pipe wall meshing stress data is generated, which prepares the data for subsequent spatial projection and node fusion. A spatial projection model of the pipeline is established. Considering the three-dimensional shape of the pipeline, a suitable spatial mapping method needs to be selected to map the meshing stress data to the actual three-dimensional space of the pipeline. The projection method (such as polynomial fitting, principal component analysis, etc.) is used to map the meshing stress data to the three-dimensional space of the pipeline. In this process, the spatial distribution characteristics of the stress data need to be maintained to reflect the true physical phenomenon. The data after spatial projection represents the stress distribution of different spatial points in the pipeline, forming stress spatial mapping data, which provides necessary input for subsequent node fusion. According to the topology of the pipeline, the stress spatial mapping data is fused with the topology network of the pipeline. Through the node correspondence relationship, the spatial stress data is mapped to each node of the pipeline to generate the initial node stress data. This step needs to correctly associate the stress data with the nodes according to the connection relationship of the pipeline (such as the branches and connection points of the pipeline). The initial node stress data is decomposed into a direction tensor. This can be done by eigenvalue decomposition of the stress tensor or principal stress calculation method to extract the principal stress direction and size of the node. The result of tensor decomposition is the node principal stress vector data, which reflects the stress size and direction of each node in the pipeline in different directions. Each node will obtain a principal stress vector data, which represents the stress response of the node in space along different directions. This data provides directional basis for further analysis of the stress state of the pipeline nodes. The principal stress vector data of all nodes is fused into a full graph. By merging all node data in the pipeline topology network, a global optimization method (such as least squares method, weighted average method, etc.) is used to combine the stress data of each node to obtain a global node stress distribution model. In the graph fusion process, the connectivity of the pipeline network is considered. For nodes connected to each other, the strength of the connection should be weighted to reflect the degree of stress transmission between nodes. This can be achieved by using the weighted edge algorithm in graph theory, which gives different weights to the connection according to the relative position and stress transmission relationship between nodes. Through full graph fusion and connectivity weighting, the node stress graph data of the pipeline is finally generated. This graph data can intuitively show the stress state of each node in the pipeline network, helping to identify the high stress sections and damage risks in the pipeline.

[0142] Preferably, the joint threshold discrimination processing of the pressure time series deviation data and the pipe wall equivalent stress value in step S35 includes:

[0143] When any of the following conditions occurs, it is determined that there is an abnormal sudden pressure fluctuation, and abnormal sudden pressure fluctuation data is obtained: the pressure change rate within any 5 minutes exceeds 0.5 MPa / min; the single-point pressure value increases or decreases by more than 1.2 MPa within 10 minutes; the pressure fluctuation frequency is greater than 0.8 Hz, and the duration is more than 15 minutes;

[0144] When the following conditions occur simultaneously, it is determined that there is an abnormal periodic pressure fluctuation, and abnormal periodic pressure fluctuation data is obtained: the pressure peak deviation in adjacent monitoring periods exceeds ±20%; the wave peak / trough time interval change rate in three consecutive periods exceeds ±25%; after comparison with the historical baseline waveform, the dynamic time warping deviation exceeds 10%;

[0145] When the following conditions are met simultaneously, it is determined that there is an abnormal pipe wall equivalent stress overrun, and equivalent stress overrun abnormal data is obtained: the pipe wall equivalent stress value exceeds 80% of the design material yield limit, i.e., exceeds 160 MPa, where the design material yield limit is 200 MPa; the stress concentration coefficient is higher than 2.5; the equivalent stress sudden increase rate exceeds 15 MPa / 10 min within any 30-minute window, and the abnormal section length exceeds 20 meters;

[0146] Integrate the abnormal sudden pressure fluctuation data, the abnormal periodic pressure fluctuation data, and the equivalent stress overrun abnormal data, align the spatial positions and time stamps, and generate pipe network abnormal pressure detection data.

[0147] As an example of the present application, reference is made to Figure 3 In this example, the step S4 includes:

[0148] Step S41: analyze the spatial coupling of the pipe network abnormal pressure detection data to generate abnormal node screening data; perform topological path extraction on the abnormal node screening data;

[0149] Step S42: component recognition of the pipe topological network according to the extracted topological path to generate pipe network component data; bidirectional fluid-structure coupling modeling of the pipe network component data to generate water hammer effect structure response data;

[0150] Step S43: extract the time domain features of the water hammer effect structure response data to obtain water hammer coupled pulse pressure peaks; joint positioning processing of the water hammer coupled pulse pressure peaks and the abnormal node screening data to generate pipe network abnormal pressure region data;

[0151] Step S44: adaptive monitoring deployment processing is performed based on the pipe network abnormal pressure area data, so as to realize automatic monitoring operation of the pipe network pressure.

[0152] In the embodiments of the present application, a spatial coupling model of abnormal pressure data in the pipe network is established based on the topology of the pipe network and the spatial distribution of pressure data. The model considers the mutual influence of each node and connecting pipeline in the pipe network, and analyzes the propagation characteristics of abnormal pressure data in the pipe network. By analyzing the relationship between abnormal pressure and adjacent nodes, the coupling coefficients between nodes are calculated to reflect the transmission strength and expansion range of abnormal pressure. Common methods include correlation analysis and covariance matrix calculation. According to the results of spatial coupling analysis, nodes with greater abnormal pressure fluctuations in the coupling relationship are selected to generate abnormal node screening data. These nodes are usually the areas with the most significant pressure abnormalities in the pipe network. In the topology network of the pipeline, the path from one abnormal node to other nodes is defined, and the connected path between abnormal nodes is focused on. The shortest path algorithm (such as Dijkstra algorithm) in graph theory is used to extract the topological path, ensuring that the extracted path reflects the propagation trend of abnormal pressure. The connection relationship of the pipeline and the distance between the nodes need to be considered when extracting the path. The extracted topological path is analyzed to evaluate the propagation path and influence range of abnormal pressure, providing data support for subsequent component identification and water hammer effect modeling. According to the topological path, the key components in the pipeline system are identified, including pipeline sections, valves, joints, supports, etc. These components are the main parts of the pipeline system and directly affect the transmission of water flow and pressure fluctuations. Each component in the pipe network is classified, and pipe network component data is generated. These data include the geometric characteristics, material properties, connection relationships, etc. of the components, which are helpful for subsequent fluid-structure coupling analysis and water hammer effect modeling. Based on the pipe network component data, a bidirectional coupling model of fluid and structure is established. The model considers the interaction between fluid pressure fluctuations and pipeline structure, especially when water hammer effect occurs, the influence of rapid changes of fluid on pipeline structure. In the model, the mutation of water flow (such as water hammer phenomenon caused by valve closing or pump station starting) is simulated, and the influence of water hammer effect on pipeline structure is calculated. In the process of fluid-structure coupling modeling, CFD (Computational Fluid Dynamics) and FEA (Finite Element Analysis) are used for joint simulation. According to the simulation results of the bidirectional fluid-structure coupling model, the response data of the pipeline structure under water hammer effect are obtained, including the deformation, vibration, stress change, etc. of the pipeline. This data provides detailed structural response information for subsequent water hammer effect analysis. Time domain features are extracted from the water hammer effect structural response data, focusing on the fluctuation of pressure, displacement, acceleration, etc. Common methods include short-time Fourier transform (STFT), wavelet transform, etc. Based on the time domain features, the pulse pressure peak in water hammer effect is identified. This pressure peak usually appears at the moment of flow rate mutation or pipeline sudden stop, and is the main influence of water hammer effect on pipeline structure. Through time domain analysis, the peak value of water hammer coupling pulse pressure is calculated, which is an important indicator for water hammer effect analysis and is further used for positioning the abnormal pressure area in the pipe network.The water hammer coupled pulse pressure peak value is combined with the abnormal node screening data for joint positioning analysis to identify the spatial relationship between the pipe network region where the pulse pressure peak value occurs and the abnormal node. Through joint analysis, the region in the pipe network affected by the water hammer effect and having abnormal pressure is located. The region is a potential failure point due to structural weakness, improper connection and other factors. According to the results of the joint positioning processing, pipe network abnormal pressure region data is generated. These data contain the coordinates of the region where the abnormal pressure occurs, the influence range and the pressure fluctuation degree, providing a basis for subsequent monitoring and maintenance. According to the pipe network abnormal pressure region data, an adaptive monitoring point deployment scheme is designed. The monitoring point should cover the abnormal pressure region and be able to collect real-time data such as pressure, flow, temperature, etc. A dynamic adjustment mechanism for the monitoring point is established to automatically adjust the position and collection frequency of the monitoring point according to real-time monitoring data, ensuring that the monitoring effect of important regions and potential risk points is maximized. Real-time pressure data of the pipe network is collected through the installed sensors, and real-time analysis is performed through the data processing and analysis system to identify whether there is abnormal pressure fluctuation. When the system detects abnormal pressure values or water hammer effect, the early warning and response mechanism is automatically triggered. This includes notifying the pipeline maintenance personnel, starting the emergency handling program, etc. When an anomaly occurs, the system can automatically adjust the pressure and flow of the pipe network according to the preset rules to reduce the impact of water hammer effect on the pipeline. Through the adaptive monitoring system, the pressure control strategy of the pipe network is continuously optimized to avoid excessive pressure on the pipeline structure or unnecessary impact of water hammer effect.

[0153] Especially important is that the joint positioning processing of the water hammer coupled pulse pressure peak value and the abnormal node screening data in step S43 also includes:

[0154] The water hammer coupled pulse pressure peak value data is subjected to spatiotemporal distribution reconstruction to generate pressure peak spatiotemporal graph data;

[0155] The abnormal node screening data is subjected to topological neighborhood expansion to generate abnormal node expansion region data;

[0156] The pressure peak spatiotemporal graph data and the abnormal node expansion region data are subjected to superimposition analysis to generate pressure fluctuation region data;

[0157] The pressure fluctuation region data and the pipe topology network are subjected to intersection analysis to generate pipe network abnormal pressure region data.

[0158] In the embodiments of the present application, by collecting the data of water hammer coupled pulse pressure peaks, the data includes time stamp, pressure peak, node position, etc. Using space-time interpolation techniques such as bilinear interpolation, Kriging interpolation, etc., the pressure peak data is reconstructed to generate a space-time graph containing time and space dimensions. Through spatial interpolation, the areas where sensor data is missing are filled in; through time interpolation, the distribution of pressure peaks at different time points is accurately reconstructed. The reconstructed data forms a space-time graph, where the horizontal coordinate represents time, the vertical coordinate represents the spatial position (node or pipe segment), and the data point represents the intensity of the pressure peak. This space-time graph can visually show the change of pressure peak with time and space. According to the topology of the pipeline, the neighborhood range of each node is defined. The neighborhood range can be defined based on distance, connectivity or flow path. By setting the expansion range, the influence of the abnormal node on its neighborhood nodes is determined. The influence range can be a fixed distance (for example, the length of the pipe segment), or it can be dynamically adjusted according to the strength of the pressure fluctuation and the coupling relationship. The neighborhood expansion rule is applied to the data of the abnormal node, and the expanded area is the area affected by the abnormal pressure. These areas include the abnormal node and the surrounding affected areas. Based on the topology neighborhood expansion analysis, the abnormal node expansion area data is generated. These data include the neighborhood range of each abnormal node, the boundary of the expansion area, and all the affected nodes or pipe segments in the area. The generated abnormal node expansion area data is stored in the database, and the dynamic changes of each expansion area are recorded in combination with the time information for subsequent analysis. The rules of space-time superposition are defined, mainly combining the pressure peak space-time graph data and the abnormal node expansion area data according to time and space position. The goal of superposition analysis is to find those areas that have both pressure fluctuations and abnormal node expansions. By calculating the overlap of the pressure peak space-time graph and the abnormal node expansion area in space-time, the areas that have significant pressure fluctuations and have an impact within the topology neighborhood are screened out. These areas are the pressure fluctuation areas of the pipeline network. After the superposition analysis is completed, the pressure fluctuation area data is identified. These areas are usually places in the pipeline network where there are large pressure fluctuations due to water hammer effect or other factors. The identified pressure fluctuation area data is stored, including the spatial position of the fluctuation area, the influence range, the pressure fluctuation intensity, etc. Through the intersection analysis of the pressure fluctuation area data and the pipeline topology network, it is identified which pipe segments, nodes or areas are simultaneously in the pressure fluctuation area and the pipeline topology network. The intersection analysis ensures that only those areas affected by the water hammer effect are considered. Using geometric calculations or intersection algorithms in graph theory, the intersection part of the pressure fluctuation area and the pipeline topology structure is determined. These parts represent the potential fault areas in the pipeline network affected by the water hammer effect. Based on the results of the intersection analysis, the areas in the pipeline network affected by abnormal pressure are identified and marked. These areas are usually pipe segments or nodes with large pressure fluctuations caused by water hammer effect.The generated abnormal pressure area data of the pipe network is stored as data for visualization or analysis, including information such as pipe section number, node number, pressure fluctuation degree, and the like.

[0159] Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting, the scope of the application being defined by the appended claims and not by the above description, and all changes falling within the meaning and range of equivalency of the elements of the patent file are therefore intended to be embraced within the present application.

[0160] The foregoing is considered as illustrative only of the principles of the application. Numerous modifications and adaptations will be apparent to those skilled in the art in view of the foregoing description, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the invention. Therefore, this application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method of processing data for automatic monitoring of pressure in a pipe network, characterized in that, The method comprises the following steps: Step S1: obtaining pipe network structure data; constructing a pipe topology network based on the pipe network structure data to generate a pipe topology network; and filtering main pipes and branch pipes to obtain pipe network flow data; Step S2: deploying a pipe wall microwave resonator on the main pipes and branch pipes based on the pipe network flow data, and capturing pipe circumferential deformation harmonics in real time through the deployed microwave resonator; analyzing the longitudinal wave propagation time delay of the pipe network flow data, and constructing a pipe body structure dynamic resonance spectrum in combination with the pipe circumferential deformation harmonics; wherein the analysis of the longitudinal wave propagation time delay of the pipe network flow data in step S2 comprises: segmenting and fitting the flow velocity data to generate in-segment flow velocity data; coupling the sound velocity of the in-segment flow velocity data to obtain medium sound velocity distribution data; topologically mapping the medium sound velocity distribution data based on the pipe network flow data to generate sound wave propagation path graph data; calculating the propagation time delay integral of the sound wave propagation path graph data to generate initial longitudinal wave propagation time delay data; detecting drift anomalies of the longitudinal wave propagation time delay data to obtain the longitudinal wave propagation time delay of the pipe network flow data; Step S3: calculating the pipe network internal pressure and pipe wall equivalent stress of the pipe topology network through the pipe body structure dynamic resonance spectrum to obtain the pipe internal pressure value and the pipe wall equivalent stress value; and discriminating the pipe network flow data based on the pipe internal pressure value and the pipe wall equivalent stress value to obtain pipe network abnormal pressure detection data; wherein step S3 comprises the following steps: Step S31: frequency inversion of the pipe body structure dynamic resonance spectrum to generate resonance frequency-structure coupling data; Step S32: calculating the stress of the resonance frequency-structure coupling data to obtain the pipe wall equivalent stress value; Step S33: stress distribution mapping of the pipe topology network based on the pipe wall equivalent stress value to generate pipe node stress map data; Step S34: internal pressure inversion calculation of the pipe node stress map data in combination with the pipe network flow data to generate the pipe internal pressure value; Step S35: time series comparison and analysis of the pipe internal pressure value data and the pipe network flow data to generate pressure time series deviation data; joint threshold discrimination processing of the pressure time series deviation data and the pipe wall equivalent stress value to generate the pipe network abnormal pressure detection data; Step S4: pipe fitting screening of the pipe topology network based on the pipe network abnormal pressure detection data to obtain pipe network fitting data; analyzing the two-way fluid-structure coupling of the pipe network fitting data to obtain water hammer coupled pulse pressure peak value; and accurately positioning the abnormal pipe network pressure area in the pipe topology network through the pipe network abnormal pressure detection data and the water hammer coupled pulse pressure peak value to realize automatic monitoring of the pipe network pressure.

2. The method of automatic monitoring of data processing of pressure in pipe networks according to claim 1, characterized by, Step S1 comprises the following steps: Step S11: obtaining a pipe network original engineering drawing and extracting pipe network structure information in the pipe network original engineering drawing to obtain pipe network structure data; Step S12: node connectivity analysis of the pipe network structure data to generate pipe topology network data; Step S13: flow path clustering of the pipe topology network data to generate candidate main path data; Step S14: path weight judgment is performed on the candidate trunk path data, and trunk pipeline and branch pipeline division is performed on the candidate trunk path data according to the judgment result, to obtain trunk and branch pipeline division data; Step S15: flow logic of the trunk and branch pipeline division data is extracted, so as to generate pipeline network flow data.

3. The method of automatically monitoring data processing of pressure in a pipe network according to claim 2, characterized in that, Step S14 includes the following steps: Step S141: path weight judgment is performed on the candidate trunk path data, to obtain a judgment result, wherein the path weight includes the geometric length of the path, the flow demand on the path, and the pressure demand of the path; Step S142: when the following conditions occur at the same time, the candidate trunk path data is judged as trunk pipeline division data: the geometric length of the path is between 500 meters and 5000 meters, the flow demand on the path is between 50 cubic meters / hour and 1000 cubic meters / hour, and the pressure demand of the path is between 0.5 MPa and 5 MPa; Step S143: when any of the following conditions occurs, the candidate trunk path data is judged as branch pipeline division data: the geometric length of the path is less than 500 meters, the flow demand on the path is less than 50 cubic meters / hour, and the pressure demand of the path is less than 0.5 MPa or higher than 5 MPa; Step S144: the trunk pipeline division data and the branch pipeline division data are integrated into the trunk and branch pipeline division data.

4. The method of claim 1, wherein, The pipeline wall microwave resonator deployment based on the pipeline network flow data in step S2 includes: Screening key nodes of the pipeline network flow data; Performing trunk and branch segmentation mapping on the key nodes to generate segmentation deployment unit data; Analyzing waveguide adaptability of the segmentation deployment unit data, and performing multi-point frequency distribution optimization on the segmentation deployment unit data according to the waveguide adaptability to generate resonator configuration scheme data; Performing position calibration on the resonator configuration scheme data to generate microwave resonator deployment parameter data; Performing pipeline wall microwave resonator deployment according to the microwave resonator deployment parameter data, so as to obtain the deployed microwave resonator.

5. The method of claim 1, wherein, The pipe body structure dynamic resonance spectrum construction in step S2 includes: Synchronously calibrating the longitudinal wave propagation time delay data and the pipe circumferential deformation harmonic to generate time-frequency joint reference data; Extracting resonance response characteristics of the time-frequency joint reference data; Performing pipe segment structure mapping on the pipe through the resonance response characteristics to generate structure resonance distribution data; Performing dynamic energy clustering analysis on the structure resonance distribution data to generate dynamic resonance clustering data; Constructing the pipe body structure dynamic resonance spectrum based on the structure resonance distribution data, the dynamic resonance clustering data, and the resonance response characteristics.

6. The method of claim 1, wherein, Step S33 includes the following steps: Step S331: performing grid reconstruction on the pipe wall equivalent stress data to generate pipe wall gridded stress data; Step S332: performing spatial projection on the pipe wall gridded stress data to generate stress space mapping data; Step S333: performing node fusion on the stress space mapping data and the pipe topology network to generate initial node stress data; performing direction tensor decomposition on the initial node stress data to generate node principal stress vector data; Step S334: performing full map fusion and connectivity weighting on the node principal stress vector data to generate pipe node stress map data.

7. The method of claim 1, wherein, The joint threshold discrimination processing of the pressure time sequence deviation data and the pipe wall equivalent stress value in step S35 includes: When any of the following conditions occurs, it is determined that there is an abnormal sudden pressure fluctuation, and abnormal sudden pressure fluctuation data is obtained: the pressure change rate within any 5 minutes exceeds 0.5 MPa / min; the single-point pressure value increases or decreases by more than 1.2 MPa within 10 minutes; the pressure fluctuation frequency is greater than 0.8 Hz, and the duration is more than 15 minutes; When the following conditions occur simultaneously, it is determined that there is an abnormal periodic pressure fluctuation, and periodic pressure fluctuation abnormal data is obtained: the pressure peak deviation in adjacent monitoring periods exceeds ± 20%; the wave peak / trough time interval change rate in three consecutive periods exceeds ± 25%; after comparison with the historical baseline waveform, the dynamic time warping deviation exceeds 10%; When the following conditions are met simultaneously, it is determined that there is an abnormal pipe wall equivalent stress overrun, and equivalent stress overrun abnormal data is obtained: the pipe wall equivalent stress value exceeds 80% of the design material yield limit, i.e., more than 160 MPa, wherein the design material yield limit is 200 MPa; the stress concentration coefficient is higher than 2.5; the equivalent stress sudden increase rate exceeds 15 MPa / 10 min within any 30-minute window, and the abnormal section length exceeds 20 meters; Integrate the abnormal sudden pressure fluctuation data, the periodic pressure fluctuation abnormal data, and the equivalent stress overrun abnormal data, align the spatial position and the time stamp, and generate pipe network abnormal pressure detection data.

8. The method of claim 1, wherein, Step S4 includes the following steps: Step S41: analyze the spatial coupling of the pipe network abnormal pressure detection data to generate abnormal node screening data; and perform topological path extraction on the abnormal node screening data; Step S42: component recognition is performed on the pipe topological network according to the extracted topological path to generate pipe network component data; and bidirectional fluid-structure coupling modeling is performed on the pipe network component data to generate water hammer effect structure response data; Step S43: time domain features of the water hammer effect structure response data are extracted to obtain water hammer coupled pulse pressure peaks; joint positioning processing is performed on the water hammer coupled pulse pressure peaks and the abnormal node screening data to generate pipe network abnormal pressure region data; Step S44: adaptive monitoring deployment processing is performed based on the pipe network abnormal pressure region data to realize automatic monitoring of the pipe network pressure.

Images