Gas pipe network abnormal data intelligent identification and early warning method and system

Abstract

The invention relates to the technical field of gas pipe network monitoring and early warning, in particular to an intelligent recognition and early warning method and system for abnormal data of a gas pipe network, and the system specifically comprises a pipe network abnormality early warning center, a multi-dimensional data module, a multi-dimensional processing module, a troubleshooting false alarm module, an abnormality marking module, an operation and maintenance division module, a dynamic early warning module and an early warning module. Through multi-dimensional data acquisition and processing and preliminary screening and spatial correlation analysis and verification of a pipe network identification model, missing judgment of single data can be avoided, equipment fault misinformation can be accurately eliminated, meanwhile, by means of a triangulation positioning method and historical data deviation correction, a system can upgrade abnormal points from preliminary positioning to accurate coordinates, and the accuracy of the abnormal points is improved. Through GIS map visual marking, blind troubleshooting of operation and maintenance personnel is avoided, the operation and maintenance personnel are guided to deal with high-level risk events preferentially based on early warning level division, resource waste in a low-risk scene is avoided, operation and maintenance pertinence is improved, and management cost is reduced.

Description

Technical Field

[0001] This invention relates to the field of gas pipeline network monitoring and early warning technology, and in particular to a method and system for intelligent identification and early warning of abnormal data in gas pipeline networks. Background Technology

[0002] As a core infrastructure for urban energy supply, gas pipeline networks play a crucial role in delivering gas to residents and industrial users. Their operational safety and stability are directly related to public safety, people's livelihood, and the continuity of industrial production. With the acceleration of urbanization, the scale of gas pipeline networks continues to expand (covering multiple levels of pipelines, including main pipelines, branch pipelines, and user access pipelines). Furthermore, some pipelines suffer from aging and corrosion due to their age. These issues, coupled with changes in the surrounding environment (such as soil subsidence and construction vibration) and fluctuations in operating parameters (pressure and flow), significantly increase the risk of abnormal events such as leaks and blockages.

[0003] However, traditional gas pipeline anomaly monitoring and early warning models have many limitations and are no longer suitable for the safety management needs of large-scale and complex pipeline networks. The specific problems are as follows:

[0004] Firstly, data collection is incomplete: it often only monitors operating parameters such as pressure and flow rate, ignoring data related to the surrounding environment (soil settlement, vibration) and pipeline location, which can easily lead to missed detection of abnormal causes due to missing information;

[0005] Secondly, the anomaly identification is inaccurate: it relies on a single parameter threshold for judgment, without considering the correlation between parameters and environmental interference, resulting in prominent problems of false alarms (normal fluctuations are judged as abnormal) and false alarms (hidden blockages are not detected).

[0006] Third, the problem of unclear positioning and disordered operation and maintenance management is due to the reliance on manual inspection and positioning, which can only roughly determine the area. Finding fault points in complex pipelines is time-consuming, delays the handling of anomalies, increases safety risks, and preventive maintenance is not supported by historical data. Furthermore, there is no graded warning system, high-risk anomalies are not prioritized, and low-risk anomalies waste resources, resulting in low operation and maintenance efficiency.

[0007] To address the aforementioned technical shortcomings, a solution is proposed. Summary of the Invention

[0008] The purpose of this invention is to provide a method and system for intelligent identification and early warning of abnormal data in gas pipeline networks. It takes "data acquisition - preprocessing - preliminary identification - verification and investigation - precise positioning - operation and maintenance early warning" as the core link. Each module is closely connected through signal response to form a complete closed loop from anomaly detection to handling and optimization. At the same time, through continuous monitoring of signals and dynamic early warning adjustments, it can adapt to changes in pipeline operation in real time, realize long-term and dynamic safety management and control of gas pipeline networks, ensure stable operation of pipeline networks, and solve the technical defects mentioned above.

[0009] The objective of this invention can be achieved through the following technical solution: a gas pipeline network abnormal data intelligent identification and early warning system, comprising a pipeline network abnormal early warning center, a multi-dimensional data module, a multi-dimensional processing module, a false alarm detection module, an abnormal marking module, an operation and maintenance division module, a dynamic early warning module, and an early warning module;

[0010] The multidimensional data module is used to collect multidimensional raw data streams from the gas pipeline network and send them to the pipeline network anomaly early warning center for storage; the multidimensional processing module is used to preprocess the collected multidimensional raw data streams to obtain standardized preprocessed data, and is accompanied by an anomaly identification feedback process.

[0011] The false alarm detection module is used to build a spatial correlation analysis model and perform false alarm feedback analysis to obtain minor leakage signals or manual detection signals.

[0012] The anomaly marking module is used to analyze the anomaly location and visual feedback of anomaly monitoring points, and finally obtain a visual GIS management map;

[0013] The operation and maintenance division module is used to perform targeted operation and maintenance analysis on the historical abnormal data of the collected leak point P, and to process the obtained pipeline fault index to obtain regular signals or operation and maintenance management signals.

[0014] The dynamic early warning module is used to perform dynamic early warning classification and analysis on the acquired leak points, process the dynamic early warning scores, and output low-level, medium-level, or high-level early warning signals.

[0015] Preferably, the analysis process of the multi-dimensional processing module is as follows: inputting pre-processed data into a pre-set pipeline identification model to obtain the output anomaly identification result; judging whether anomaly exists in the anomaly identification result; if it exists, generating an anomaly signal; if it does not exist, generating a continuous monitoring signal.

[0016] Preferably, the analysis process of the false alarm detection module is as follows:

[0017] TT1: Based on the pipeline network topology, each monitoring point is regarded as a node and the pipeline is regarded as an edge to establish a spatial correlation analysis model;

[0018] TT2: Calculate the spatial distance, pipeline length and media transmission time of adjacent monitoring points based on GIS information, and set parameter association thresholds for adjacent monitoring points;

[0019] TT3: Obtain the abnormal monitoring point corresponding to the abnormal signal, and extract the preprocessed data of the adjacent monitoring points of the identified abnormal monitoring point.

[0020] Preferably, it also includes TT4: determining whether there is a parameter change trend matching the abnormal monitoring point among adjacent monitoring points. If there is a matching trend among adjacent monitoring points, it is determined to be a substantial abnormal event in the pipeline network. If there is no matching trend among adjacent monitoring points, the status data of the sensor of the abnormal monitoring point is further obtained, and the parameters in the status data are compared and analyzed with the preset threshold to obtain false alarm signals or suspicious monitoring points. The abnormal influence radius of the suspicious monitoring point is obtained, and the abnormal influence radius = preset pipeline coefficient × pipeline pressure change.

[0021] TT5: Obtain the monitoring distance between the suspected monitoring point and the adjacent monitoring point, and process the abnormal influence radius and the monitoring distance to obtain a slight leakage signal or a manual investigation signal.

[0022] Preferably, the analysis process of the anomaly marking module is as follows:

[0023] Step 1: Mark the abnormal monitoring point as A, and simultaneously retrieve the adjacent monitoring points B and C that generated the abnormal signal successively, and the three monitoring points are not collinear;

[0024] Step 2: Set the GIS coordinates (x, y), A(x1, y1), B(x2, y2), and C(x3, y3) of the leak point P, and obtain the propagation speed v;

[0025] Obtain the anomaly detection times t1, t2, and t3 from the leak point P to A, B, and C.

[0026] Step 3: The time difference Δt1 = t1 - t0 when monitoring point A detects the anomaly corresponds to the distance d1 = v × Δt1 from A to P, where t0 represents the time when the leak occurred;

[0027] The time difference Δt2 = t2 - t0 when monitoring point B senses the anomaly corresponds to the distance d2 = v × Δt2 from B to P.

[0028] The time difference Δt3 = t3 - t0 when monitoring point C senses the anomaly corresponds to the distance d3 = v × Δt3 from C to P. Eliminating t0 yields the distance difference d1 - d2 = v × (t1 - t2) between A and B and the distance difference d1 - d3 = v × (t1 - t3) between A and C, respectively denoted as the distance difference ΔAB = v × ΔtAB between A and B, where ΔtAB represents the absolute value of t1 - t2, and the distance difference ΔAC = v × ΔtAC between A and C, where ΔtAC represents the absolute value of t1 - t3.

[0029] Preferably, the process also includes a fourth step: obtaining the distance equation between A and P: (x-x1) 2 +(y-y1) 2 =d1 2 Similarly, the distance equations between B and P and between C and P are obtained.

[0030] Substituting d2 = d1 - ΔAB and d3 = d1 - ΔAC into the distance equations between B and P and between C and P, and combining them with the distance equation between A and P, the leakage point P(x initial, y initial) is obtained by solving the above system of two linear equations and filtering based on the propagation direction A to B and then to C.

[0031] Step 5: Statistically analyze historical leakage data for the same area, calculate the average deviation value (deviation longitude, deviation latitude), and superimpose the current initial coordinates with the average deviation value of the area to obtain the corrected final leakage point P coordinates (x final, y final).

[0032] Step 6: Based on the spatial correlation analysis model, mark the leakage point P(x final, y final) on the spatial correlation analysis model with red circles to obtain a visual GIS management map.

[0033] Preferably, the analysis process of the operation and maintenance division module is as follows: Based on the visual GIS pipe map, a pipe of a preset length with the leak point P as the center point is obtained, and the pipe of the corresponding preset length is set as the target pipe. The historical abnormal data of the target pipe is obtained, including the number of faults and the fault repair time. The number of faults and the fault repair time are normalized and multiplied by the corresponding preset weight coefficient to obtain the value as the pipe fault index. The pipe fault index is then processed to obtain the normal signal or operation and maintenance management signal.

[0034] Preferably, the analysis process of the dynamic early warning module is as follows: Based on the visual GIS pipe map, the main pipe nodes, branch pipe nodes, and user access nodes are divided into high importance, medium importance, and low importance, and assigned values ​​to obtain the importance level corresponding to the current leakage point P; at the same time, the diffusion impact data and leakage risk level of leakage point P are obtained; based on the analysis of the diffusion impact data and leakage risk level of leakage point P, a leakage impact score and a leakage risk assignment are obtained; the value calculated by weighting the importance level assignment × m1 + leakage impact score × m2 + leakage risk assignment × m3 is set as the dynamic early warning score; and the dynamic early warning score is compared and analyzed with the preset maximum dynamic early warning score Jmax and the preset maximum dynamic early warning score Jmin to obtain a low-level early warning signal, a medium-level early warning signal, or a high-level early warning signal.

[0035] The beneficial effects of this invention are as follows:

[0036] (1) This invention uses multi-dimensional data collection to cover all dimensions of pipeline operation, environment and location information, combined with data preprocessing to optimize input quality, and then performs preliminary screening and spatial correlation analysis verification through pipeline identification model. This can avoid missing judgments of single data and accurately eliminate false alarms of equipment failures, effectively identify substantial anomalies such as leaks and blockages, and greatly reduce safety accidents such as gas explosions and supply interruptions caused by missing or misjudged anomalies, thus protecting residents' lives, industrial production and public safety.

[0037] (2) The present invention also uses the triangulation method combined with historical data deviation correction. The system can upgrade the abnormal point from preliminary positioning to precise coordinates, and visually present the fault location and pipeline topology relationship through GIS map visualization, avoiding blind troubleshooting by maintenance personnel. By analyzing the historical fault data of pipelines, the system can distinguish between regular pipelines and frequently faulty pipelines, and realize the concentration of maintenance resources in high-risk areas.

[0038] (3) This invention classifies early warning levels based on the importance of the pipeline network, the scope of impact, and the degree of risk, guiding maintenance personnel to prioritize the handling of high-risk events and avoid wasting resources in low-risk scenarios. This not only improves the targeting of maintenance but also reduces management costs such as manpower and material resources. Attached Figure Description

[0039] The invention will now be further described with reference to the accompanying drawings;

[0040] Figure 1 This is a flowchart of the system of the present invention;

[0041] Figure 2 This is a reference diagram of the method of the present invention;

[0042] Figure 3 This is a partial analysis diagram of the present invention. Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments;

[0045] Example 1: Please refer to Figures 1 to 3 As shown, this invention is an intelligent identification and early warning system for abnormal data in gas pipeline networks, including a pipeline network abnormality early warning center, a multi-dimensional data module, a multi-dimensional processing module, a false alarm detection module, an anomaly marking module, an operation and maintenance division module, a dynamic early warning module, and an early warning module. The pipeline network abnormality early warning center has bidirectional communication connections with both the multi-dimensional data module and the multi-dimensional processing module. The pipeline network abnormality early warning center has a unidirectional communication connection with the false alarm detection module. Both the multi-dimensional processing module and the false alarm detection module have a unidirectional communication connection with the early warning module. The pipeline network abnormality early warning center has a bidirectional communication connection with the anomaly marking module. The anomaly marking module has a unidirectional communication connection with the operation and maintenance division module. The operation and maintenance division module has a unidirectional communication connection with the dynamic early warning module. The dynamic early warning module has a unidirectional communication connection with the early warning module.

[0046] The multidimensional data module is used to collect multidimensional raw data streams from the gas pipeline network and send them to the pipeline anomaly early warning center for storage;

[0047] That is, setting a monitoring period and collecting multi-dimensional raw data streams of the gas pipeline network within the monitoring period, wherein the multi-dimensional raw data streams include:

[0048] Pipeline network operating parameters: pressure, flow rate, temperature, medium composition concentration, etc.;

[0049] Surrounding environmental data: ambient temperature, humidity, soil subsidence, surface vibration frequency, etc.;

[0050] Location-related data: GIS coordinates of each monitoring point and their relationship with the pipeline network topology (such as pipe diameter, connection method, etc.);

[0051] This means simultaneously collecting three core types of data: pipeline operation, surrounding environment, and location correlation. This avoids the limitations of a single data dimension, provides complete data support for subsequent anomaly identification, and reduces the risk of missed detection due to missing data.

[0052] The multidimensional processing module is used to preprocess the collected multidimensional raw data stream to obtain standardized preprocessed data. Preprocessing includes data cleaning, standardization, etc.

[0053] The pre-processed data is input into a pre-set pipeline identification model to obtain the output anomaly identification results, including whether the anomaly exists, the anomaly type (leakage and blockage), and the GIS coordinates of the anomaly monitoring point.

[0054] The system determines whether an anomaly exists in the anomaly identification results. If an anomaly exists, an anomaly signal is generated; otherwise, a continuous monitoring signal is generated.

[0055] Upon responding to abnormal signals or continuous monitoring signals, the early warning module immediately performs the preset early warning operation corresponding to the abnormal signal or continuous monitoring signal, so as to achieve continuous monitoring and abnormal identification and early warning of the gas pipeline network.

[0056] When an abnormal signal is generated, a spatial correlation analysis model is constructed by combining the pipeline network topology and GIS information. The abnormality identification results are verified to eliminate false alarms caused by equipment failure and accurately locate the substantive abnormality. That is, the false alarm investigation module is used to construct the spatial correlation analysis model and to carry out false alarm investigation feedback analysis. The specific false alarm investigation feedback analysis process is as follows:

[0057] TT1: Based on the pipeline network topology, each monitoring point is regarded as a node and the pipeline is regarded as an edge to establish a spatial correlation analysis model;

[0058] TT2: Calculate the spatial distance, pipeline length and media transmission time of adjacent monitoring points based on GIS information, and set parameter association thresholds for adjacent monitoring points (e.g., the pressure difference between adjacent monitoring points should be ≤0.2MPa);

[0059] TT3: Obtain the abnormal monitoring point corresponding to the abnormal signal. For the identified abnormal monitoring point, extract the preprocessed data of the adjacent monitoring points (select upstream or downstream monitoring points with medium connectivity and parameter consistency based on the abnormal type, that is, monitoring points with medium transmission correlation or spatial topology correlation and can synchronously sense the same pipeline network abnormality. For example, if the abnormal type is blockage, select the upstream monitoring point with medium connectivity and parameter consistency as the adjacent monitoring point. If the abnormal type is leakage, select the downstream monitoring point with medium connectivity and parameter consistency as the adjacent monitoring point).

[0060] TT4: Determine whether there is a parameter change trend at adjacent monitoring points that matches the abnormal monitoring point (e.g., when the pressure at the abnormal monitoring point drops sharply, do the pressure at adjacent monitoring points drop synchronously)?

[0061] If there is a matching trend between adjacent monitoring points, it is determined to be a substantial abnormal event in the pipeline network (leakage, blockage, etc.);

[0062] If there is no matching trend between adjacent monitoring points, the status data of the abnormal monitoring point sensor (such as power supply voltage, signal strength, etc.) is further obtained, and the parameters in the status data are compared and analyzed with the preset threshold. If there is a parameter deviation in the status data corresponding to the preset threshold, a false alarm signal is generated. That is, the early warning module is used to respond to the false alarm signal and immediately display the preset early warning text corresponding to the false alarm signal to remind the operation and management personnel that the abnormality is a false alarm caused by equipment failure and can be eliminated.

[0063] If the parameter offset in the status data does not correspond to the preset threshold, the abnormal monitoring point is determined to be a suspicious monitoring point.

[0064] The abnormal influence radius of the suspected monitoring point is obtained. The abnormal influence radius = preset pipeline coefficient × pipeline pressure change.

[0065] TT5: Obtain the monitoring distance between the suspected monitoring point and the adjacent monitoring point, and process the abnormal influence radius and the monitoring distance. If the abnormal influence radius is less than the monitoring distance, a slight leakage signal is generated. If the abnormal influence radius is greater than or equal to the monitoring distance, a manual investigation signal is generated.

[0066] The early warning module is used to respond to minor leak signals or manual inspection signals, immediately display the preset early warning text corresponding to the minor leak signal or manual inspection signal, remind the operation and management personnel of the specific verification results of the abnormality, and carry out targeted management of the gas pipeline network based on the verification results;

[0067] That is, it automatically analyzes and processes the data to quickly determine whether there are any anomalies, the type of anomaly (leakage / blockage), and the preliminary location, replacing manual investigation and greatly improving the efficiency of anomaly identification;

[0068] Based on the identification results, "abnormal signals" or "continuous monitoring signals" are generated to clarify the direction of subsequent processing and avoid ineffective investment of operation and maintenance resources.

[0069] Example 2: Accurate identification of substantial anomalies in the pipeline network (such as leaks and blockages) enables maintenance personnel to locate the fault point in a timely and accurate manner and take immediate measures to deal with it. For example, if a gas pipeline leak is not detected in time, it may cause serious safety accidents such as gas explosions. However, by accurately locating the leak point through this method, such safety risks can be greatly reduced, and the safety of residents, industrial production and public safety can be protected.

[0070] Based on the parameter variation amplitude and medium transmission time of the abnormal monitoring point and its adjacent monitoring points in the spatial correlation analysis model, the specific location of the substantial anomaly is calculated using the triangulation method, and then linked to the GIS map to generate anomaly location markers. The anomaly marker module is used to analyze the anomaly location and visual feedback of the abnormal monitoring point. The specific analysis process is as follows:

[0071] Step 1: Mark the abnormal monitoring point as A, and simultaneously retrieve the adjacent monitoring points B and C that sensed and generated the abnormal signal in sequence (i.e., A sensed it first, B second, and C last), and the three monitoring points are not collinear, that is, the propagation direction is A to B and then to C.

[0072] Step 2: Set the GIS coordinates (x, y), A(x1, y1), B(x2, y2), and C(x3, y3) of the leak point P, and obtain the propagation speed v;

[0073] Obtain the anomaly detection times t1, t2, and t3 from the leak point P to A, B, and C.

[0074] Step 3: The time difference Δt1 = t1 - t0 when monitoring point A detects the anomaly corresponds to the distance d1 = v × Δt1 from A to P, where t0 represents the time when the leak occurred;

[0075] The time difference Δt2 = t2 - t0 when monitoring point B senses the anomaly corresponds to the distance d2 = v × Δt2 from B to P.

[0076] The time difference Δt3 = t3 - t0 when monitoring point C senses the anomaly corresponds to the distance d3 = v × Δt3 from C to P;

[0077] Eliminating t0 yields the distance difference between A and B, d1-d2 = v × (t1-t2), and the distance difference between A and C, d1-d3 = v × (t1-t3), which are denoted as the distance difference between A and B, ΔAB = v × ΔtAB, where ΔtAB represents the absolute value of t1-t2, and the distance difference between A and C, ΔAC = v × ΔtAC, where ΔtAC represents the absolute value of t1-t3.

[0078] Step 4: Obtain the distance equation between A and P: (x-x1) 2 +(y-y1) 2 =d1 2 Similarly, the distance equations between B and P and between C and P are obtained.

[0079] Substitute d2 = d1 - ΔAB and d3 = d1 - ΔAC into the distance equations between B and P and between C and P, and combine them with the distance equation between A and P.

[0080] The above system of two linear equations is solved using a linear equation solver (such as Gaussian elimination). At the same time, the leakage point P (initial x, initial y) is obtained by filtering based on the propagation direction A to B and then to C.

[0081] Step 5: Statistically analyze the historical leakage data of the same area (such as a main pipeline), calculate the average deviation value (deviation longitude, deviation latitude), and superimpose the current initial coordinates with the average deviation value of the area to obtain the corrected final leakage point P coordinates (x final, y final).

[0082] Step 6: Based on the spatial correlation analysis model, the leak point P(x end, y end) is marked with a red circle on the spatial correlation analysis model to obtain a visual GIS pipeline map. The pipeline anomaly early warning center is used to respond to and display the visual GIS pipeline map, which helps to provide intuitive spatial guidance for operation and maintenance personnel.

[0083] That is, by associating the parameter trends of adjacent monitoring points and checking the sensor status, it can accurately distinguish between false alarms of equipment failure and real pipeline anomalies, reduce the interference of "false warnings" on operation and maintenance work, and reduce the waste of manpower;

[0084] The system calculates the impact radius of suspicious monitoring points and compares the spacing to further classify scenarios into "minor leaks" and "requiring manual investigation," thereby enabling graded handling of anomalies and improving the targeted nature of operation and maintenance.

[0085] The operations and maintenance (O&M) partitioning module is used to perform targeted O&M analysis on the historical anomaly data of the collected leak point P. The specific targeted O&M analysis process is as follows:

[0086] Based on the visual GIS pipe map, a pipe of a preset length centered at the leak point P is obtained, and this preset length pipe is set as the target pipe. Historical anomaly data for the target pipe is acquired, including the number of faults and the fault repair time. The number of faults and the fault repair time are normalized and multiplied by the corresponding preset weight coefficients to calculate the value, which is set as the pipe fault index. The pipe fault index is then processed for discrimination. If the pipe fault index is less than the preset pipe fault index threshold, a normal signal is generated. If the pipe fault index is greater than or equal to the preset pipe fault index threshold, the target management is set to frequent fault management, and an operation and maintenance management signal is generated simultaneously.

[0087] The early warning module is used to respond to regular signals or operation and maintenance management signals and immediately display the preset early warning text corresponding to the regular signals or operation and maintenance management signals, so as to make targeted operation and maintenance management decisions for the target pipeline;

[0088] That is, by calculating the pipeline failure index, "routine pipelines" and "frequently failing pipelines" can be distinguished, avoiding excessive maintenance of low-risk pipelines and concentrating resources on high-risk areas to improve maintenance efficiency.

[0089] Develop targeted operation and maintenance strategies based on historical data to reduce recurring failures in frequently faulty pipelines, reduce the overall operational risk of the pipeline network, and extend the service life of the pipelines.

[0090] Example 3: The dynamic early warning module is used to perform dynamic early warning classification and analysis on the acquired leakage points. The specific dynamic early warning classification and analysis process is as follows:

[0091] Based on the visual GIS pipeline map and according to the main pipeline nodes, branch pipeline nodes, and user access nodes, the gas pipeline network is classified into three importance levels. The main pipeline nodes are classified as high importance, the branch pipeline nodes as medium importance, and the user access nodes as low importance. The high importance, medium importance, and low importance are assigned a score of 10, 6, and 2 respectively (the value range is 0-10).

[0092] The importance level of the leak point P is assigned, and the diffusion impact data of the leak point P (such as the affected pipeline length, number of users, and area) and the degree of leak risk are obtained.

[0093] The diffusion impact data is normalized. The sum of the normalized pipeline length, number of users, and area multiplied by the corresponding preset weight coefficient is set as the leakage impact level. The leakage impact score (0-10) is obtained based on the leakage impact level being located in the corresponding preset leakage impact level range. It should be noted that the preset leakage impact level range is set in ascending order, and the leakage impact score is also set in ascending order.

[0094] The leakage risk level is assigned a value: 10 points for severe risk, 6 points for moderate risk, and 2 points for mild risk (the value range is 0-10), thus obtaining the leakage risk assignment for the current leakage risk level.

[0095] The preset weight coefficients m1, m2, and m3 for the importance level assignment, leakage impact score, and leakage risk assignment are obtained. m1, m2, and m3 are all greater than zero. The value calculated by assigning importance level × m1 + leakage impact score × m2 + leakage risk assignment × m3 is set as the dynamic warning score. The dynamic warning score is compared with the preset maximum dynamic warning score Jmax and the preset maximum dynamic warning score Jmin. If the dynamic warning score < Jmin, a low-level warning signal is generated. If Jmin ≤ dynamic warning score ≤ Jmax, a medium-level warning signal is generated. If the dynamic warning score > Jmax, a high-level warning signal is generated.

[0096] The early warning module is used to respond to low-level, medium-level, or high-level early warning signals, immediately perform the preset early warning operations corresponding to the low-level, medium-level, or high-level early warning signals, and make reasonable operation and maintenance management based on the degree of early warning feedback.

[0097] It should be noted that the warning intensity is as follows: low-level warning signal < medium-level warning signal < high-level warning signal;

[0098] That is, by combining triangulation and historical data deviation correction, the coordinates of abnormal points are upgraded from "preliminary location" to "precise location", avoiding maintenance personnel blindly searching for fault points and shortening the handling time;

[0099] By marking anomalies on the GIS map, the location of the fault and its relationship with the pipeline topology can be presented intuitively, reducing the spatial cognitive difficulty for maintenance personnel and helping them to quickly reach the site.

[0100] Example 4: This invention also proposes a method for intelligent identification and early warning of abnormal data in gas pipeline networks, comprising the following steps:

[0101] Step 1: Acquisition and preprocessing of multi-dimensional raw data streams to obtain standardized preprocessed data;

[0102] Step 2: Anomaly identification and feedback process for gas pipeline network preprocessing data;

[0103] Step 3: Combining the pipeline network topology and GIS information, construct a spatial correlation analysis model and conduct a false alarm feedback analysis process.

[0104] Step 4: Based on the parameter change range and medium transmission time of the abnormal monitoring point and adjacent monitoring points in the spatial correlation analysis model, the leakage point P (x final, y final) is calculated using the triangulation method, and a visual GIS pipeline map is constructed at the same time.

[0105] Step 5: Targeted operation and maintenance analysis based on visual GIS management maps and historical anomaly data;

[0106] Step Six: Determine the early warning level (low / medium / high) by comprehensively considering the importance level assignment, leakage impact score, and leakage risk assignment, and issue corresponding early warning signals to guide operation and maintenance.

[0107] In summary, by collecting multi-dimensional data covering all aspects of pipeline operation, environment, and location, and combining this with data preprocessing to optimize input quality, followed by preliminary screening and spatial correlation analysis using a pipeline identification model, the system can avoid missing single data points and accurately eliminate false alarms from equipment malfunctions. This effectively identifies substantive anomalies such as leaks and blockages, significantly reducing safety accidents such as gas explosions and supply interruptions caused by missed or misjudged anomalies, thus protecting residential life, industrial production, and public safety. Furthermore, by using triangulation combined with historical data deviation correction, the system can upgrade anomaly points from preliminary location to precise coordinates, and visualize the fault location and pipeline topology relationship through GIS maps, avoiding blind troubleshooting by maintenance personnel. By analyzing historical pipeline fault data, the system distinguishes between routine and frequently faulty pipelines, allowing maintenance resources to be concentrated in high-risk areas. Additionally, by classifying early warning levels based on pipeline importance, impact range, and risk level, the system guides maintenance personnel to prioritize high-risk events, avoiding resource waste in low-risk scenarios. This improves the targeting of maintenance efforts while reducing management costs such as manpower and materials.

[0108] The threshold is set for comparative analysis of results to determine whether they are good or bad. The value of the threshold is determined by a combination of large-scale model analysis of sample data and human experience. It can also be adjusted appropriately based on seasonal or common-sense influencing factors.

[0109] The size of the coefficient is a specific value obtained by quantifying each parameter to facilitate subsequent comparison. The size of the coefficient depends on the amount of sample data and the corresponding operating coefficient initially set by those skilled in the art for each set of sample data; as long as it does not affect the proportional relationship between the parameter and the quantified value.

[0110] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A gas pipeline network abnormal data intelligent identification and early warning system, characterized in that, The system comprises a pipeline network anomaly early warning center, a multi-dimensional data module, a multi-dimensional processing module, a false alarm elimination module, an anomaly marking module, an operation and maintenance division module, a dynamic early warning module, and an early warning module. The multi-dimensional data module is used for collecting multi-dimensional original data streams of the gas pipeline network and sending the data streams to the pipeline network anomaly early warning center for storage; the multi-dimensional processing module is used for pre-processing the collected multi-dimensional original data streams to obtain standardized pre-processed data, and accompanying the anomaly recognition feedback process; The false alarm elimination module is used for constructing a spatial correlation analysis model and accompanying false alarm elimination feedback analysis to obtain a slight leakage signal or a manual elimination signal; The anomaly marking module is used for analyzing the anomaly position and visual feedback of the anomaly monitoring point, and finally obtaining a visual GIS pipe map; The operation and maintenance division module is used for performing targeted operation and maintenance analysis on the historical anomaly data of the collected leakage point P, performing discriminant processing on the obtained pipeline fault index, and obtaining a conventional signal or an operation and maintenance management signal; The dynamic early warning module is used for performing dynamic early warning division analysis on the obtained leakage point, performing discriminant processing on the dynamic early warning score, and outputting a low-level early warning signal, a medium-level early warning signal, or a high-level early warning signal.

2. The intelligent identification and early warning system for abnormal data of a gas pipeline network according to claim 1, characterized in that, The analysis process of the multi-dimensional processing module is as follows: the pre-processed data is input into a pre-set pipeline network recognition model to obtain an output anomaly recognition result, and the presence or absence of anomaly in the anomaly recognition result is determined; if the anomaly exists, an anomaly signal is generated; if the anomaly does not exist, a continuous monitoring signal is generated.

3. The intelligent identification and early warning system for abnormal data of gas pipeline network according to claim 1, characterized in that, The analysis process of the false alarm elimination module is as follows: TT1: Based on the pipeline network topology structure, each monitoring point is regarded as a node, and the pipeline network pipeline is regarded as an edge, and a spatial correlation analysis model is established; TT2: Based on GIS information, the spatial distance, pipeline length, and medium transmission time of adjacent monitoring points are calculated, and the parameter correlation threshold of adjacent monitoring points is set; TT3: The anomaly signal corresponding to the anomaly monitoring point is obtained, and the pre-processed data of the adjacent monitoring points of the identified anomaly monitoring point is extracted.

4. The intelligent identification and early warning system for abnormal data of a gas pipeline network according to claim 3, characterized in that, TT4: It is judged whether the adjacent monitoring points have a parameter change trend matching the anomaly monitoring point; if the adjacent monitoring points have a matching trend, it is determined that there is a pipeline substantive anomaly event; if the adjacent monitoring points do not have a matching trend, the state data of the anomaly monitoring point sensor is further obtained, and the parameters in the state data are compared and analyzed with the pre-set preset threshold to obtain a false alarm signal or a suspicious monitoring point; the anomaly influence radius of the suspicious monitoring point is obtained, and the anomaly influence radius = preset pipeline coefficient × pipeline pressure change amount; TT5: The monitoring distance between the suspicious monitoring point and the adjacent monitoring point is obtained, and the anomaly influence radius and the monitoring distance are determined to obtain a slight leakage signal or a manual elimination signal.

5. The intelligent identification and early warning system for abnormal data of gas pipeline network according to claim 1, characterized in that, The analysis process of the anomaly marking module is as follows: First step: mark the anomaly monitoring point as A, and simultaneously call the adjacent monitoring points B and C that generate anomaly signals in sequence, and the three monitoring points are not collinear; Second step: set the GIS coordinates (x, y) of the leakage point P, A (x1, y1), B (x2, y2), and C (x3, y3), and obtain the propagation speed v; The abnormal perception time t1, t2 and t3 of the leakage point P to A, B and C is obtained, Third step: the time difference of abnormality sensed by monitoring point A Δ t1 = t1 - t0, corresponding to the distance d1 = v x At1 from A to P, t0 represents the time of leakage occurrence; The time difference Δt2=t2-t0 that the monitoring point B perceives the abnormality corresponds to the distance d2=v×Δt2 from B to P; The time difference Δt3=t3-t0 that the monitoring point C perceives the abnormality corresponds to the distance d3=v×Δt3 from C to P; eliminating t0 obtains the distance difference d1-d2=v×(t1-t2) between A and B and the distance difference d1-d3=v×(t1-t3) between A and C, which are respectively denoted as the distance difference ΔAB=v×ΔtAB between A and B, ΔtAB representing the absolute value of t1-t2, and the distance difference ΔAC=v×ΔtAC between A and C, ΔtAC representing the absolute value of t1-t3.

6. The intelligent identification and early warning system for abnormal data of gas pipeline network according to claim 1, characterized in that, Also include the fourth step: get to A and P distance equation: (x-x1) 2 +(y-y1) 2 =d1 2 , similarly get B and P distance equation and C and P distance equation; Substituting d2=d1-ΔAB and d3=d1-ΔAC into the distance equations of B and P and C and P and combining with the distance equation of A and P, the leakage point P(xinitial, yinitial) is obtained by solving the above binary linear equation set and screening based on the propagation direction A to B and then to C; Step 5: Statistics are performed on historical leakage data of the same region to calculate the average deviation value (deviation longitude, deviation latitude), and the current initial coordinates are superimposed with the average deviation value of the region to obtain the corrected final leakage point P coordinates (xterminal, yterminal); Step 6: Based on the spatial correlation analysis model, the leakage point P(xterminal, yterminal) is marked on the spatial correlation analysis model by a red circle to obtain a visual GIS pipe diagram.

7. The intelligent identification and early warning system for abnormal data of gas pipeline network according to claim 1, characterized in that, The analysis process of the operation and maintenance division module is as follows: based on the visual GIS pipe diagram, a pipeline of a preset length with the leakage point P as the center point is obtained, and the pipeline corresponding to the preset length is set as a target pipeline; historical abnormal data of the target pipeline is obtained, the historical abnormal data including the number of faults and the fault repair duration; the value obtained by multiplying the number of faults and the fault repair duration after normalization processing with the corresponding preset weight coefficient is set as the pipeline fault index; and the pipeline fault index is subjected to discrimination processing to obtain a conventional signal or an operation and maintenance management signal.

8. The intelligent identification and early warning system for abnormal data of gas pipeline network according to claim 1, characterized in that, The analysis process of the dynamic early warning module is as follows: based on the visual GIS pipe diagram, the main pipe nodes, branch pipe nodes and user access nodes are sequentially divided into high importance, medium importance and low importance, and the importance level corresponding to the current leakage point P is valued; At the same time, the diffusion influence data and the leakage risk degree of the leakage point P are obtained, the diffusion influence data and the leakage risk degree of the leakage point P are analyzed to obtain a leakage influence score and a leakage risk value, the value obtained by multiplying the importance level value×m1, the leakage influence score×m2 and the leakage risk value×m3 is set as the dynamic early warning score, and the dynamic early warning score is compared and analyzed with the preset maximum dynamic early warning score Jmax and the preset minimum dynamic early warning score Jmin to obtain a low-level warning signal, a medium-level warning signal or a high-level warning signal.

9. An intelligent identification and early warning method for abnormal data of a gas pipeline network, applied to an intelligent identification and early warning system for abnormal data of a gas pipeline network according to any one of claims 1-8, comprising the following steps: Step one: Collect and preprocess multi-dimensional raw data to obtain standardized preprocessed data; Step two: Abnormal identification and feedback process of gas pipeline network preprocessed data; Step three: Combined with the topology of the pipeline network and GIS information, a spatial correlation analysis model is constructed, accompanied by a false alarm feedback analysis process; Step four: Based on the parameter variation amplitude and medium transmission time of the abnormal monitoring point and the adjacent monitoring point in the spatial correlation analysis model, the leakage point P(xend, yend) is calculated by using the triangular positioning method, and a visual GIS pipe map is constructed; Step five: Targeted operation and maintenance analysis process based on the visual GIS pipe map and historical abnormal data; Step six: The warning level is determined comprehensively from the importance level assignment, leakage impact score, and leakage risk assignment, and the corresponding warning signal is sent to guide operation and maintenance disposal.

Images