Petroleum pipeline network abnormity identification and early warning system

Abstract

The invention relates to the technical field of pipeline anomaly detection, in particular to a petroleum pipeline network anomaly recognition and early warning system which comprises a pressure fluctuation monitoring module, a flow anomaly analysis module, a state association optimization module, an early warning strategy adjustment module and an anomaly early warning generation module. According to the method, deep correlation analysis is carried out on the relation between pressure fluctuation and flow abnormity, the change trend of pipeline pressure and flow can be accurately captured, the abnormal state in pipe network operation can be recognized in time, the abnormity early warning structure is dynamically adjusted, the early warning strategy can be flexibly optimized along with the change of the working condition, and therefore the detection precision and the response speed are improved; the state can be mastered in real time in a complex pipe network environment, missing detection and false alarm phenomena caused by insufficient data interpretation or early warning strategy solidification are effectively reduced, the adaptability of the system to emergency situations is enhanced, and the problems of response lag and insufficient sensitivity caused by single data source under variable working conditions are avoided.

Description

Technical Field

[0001] This invention relates to the field of pipeline anomaly detection technology, and in particular to an oil pipeline network anomaly identification and early warning system. Background Technology

[0002] The field of pipeline anomaly detection technology involves the condition monitoring and anomaly identification of pipeline systems transporting media. Its core aspects include the continuous acquisition, feature extraction, and pattern analysis of operating parameters in long-distance or complex pipeline networks for oil, natural gas, and chemical media. This enables the identification and judgment of operating conditions such as leaks, blockages, corrosion, and pressure anomalies. This field encompasses a complete technical system from sensor signal acquisition, data preprocessing, feature information extraction to the implementation of anomaly identification strategies, and is commonly applied in energy transmission, industrial production, and public safety scenarios. Among these, the traditional oil pipeline anomaly identification and early warning system refers to a system that identifies and alerts to anomalies such as pressure fluctuations, flow anomalies, and changes in media composition that occur during the operation of oil transmission pipelines. This is achieved by deploying pressure sensors, flow meters, temperature sensors, and other measuring devices along the pipeline, combined with fixed threshold comparisons, statistical feature calculations, and pattern matching based on historical operating parameters.

[0003] In practical applications, existing technologies rely too heavily on data collected by sensors deployed along the pipeline and use fixed thresholds and historical data patterns for anomaly detection. This approach is difficult to adjust flexibly when pipeline operating conditions are complex and changeable, especially under subtle pressure fluctuations and flow rate variations, which can easily lead to missed detections and false alarms. Traditional methods' pattern matching and threshold setting ignore the dynamic changes in the overall pipeline network's operating status, failing to respond promptly to abnormal fluctuations in the pipeline environment. This results in the inability to detect potential problems in their early stages, increasing the risk of emergencies such as oil leaks and blockages. Especially when dealing with large-scale, complex pipeline systems, traditional methods exhibit a lag in responding to uncertainties and cannot efficiently process real-time changing data, impacting overall monitoring effectiveness. Summary of the Invention

[0004] To address the technical problems existing in the prior art, this invention provides a system for identifying and warning anomalies in oil pipeline networks. The technical solution is as follows: On the one hand, an anomaly identification and early warning system for oil pipeline networks is provided, the system comprising: The pressure fluctuation monitoring module is based on pressure sensor data in the pipeline network. By capturing the amplitude and frequency of pressure changes and combining them with the characteristics of flow fluctuations, it analyzes the occurrence pattern of pressure anomalies and obtains a pressure anomaly trend data table. Based on the pressure anomaly trend data table, the flow anomaly analysis module extracts the relationship between flow change characteristics and pressure fluctuations, analyzes the distribution characteristics of flow anomalies, identifies key nodes of flow anomalies by combining pressure fluctuation trends, and constructs a flow anomaly distribution model. Based on the aforementioned abnormal flow distribution model, the state association optimization module extracts pipeline pressure fluctuation deviation values ​​and abnormal flow adjustment data, compares pipeline pressure changes with flow distribution characteristics, and obtains the pipeline network abnormal state identification results through pipeline pressure distribution deviation and abnormal flow adjustment. Based on the pipeline abnormality identification results, the early warning strategy adjustment module analyzes the impact of abnormality adjustment on pipeline operation, extracts pressure distribution change and flow fluctuation adjustment data, corrects the abnormality early warning structure, and obtains an abnormality early warning adjustment data list.

[0005] As a further embodiment of the present invention, the pressure anomaly trend data table includes pressure change amplitude, pressure fluctuation frequency and anomaly occurrence time; the flow anomaly distribution model includes flow change characteristics, pressure fluctuation relationship and flow anomaly key nodes; the pipeline network anomaly status identification result includes pressure fluctuation deviation value, flow anomaly adjustment data and pressure and flow distribution matching relationship; and the anomaly early warning adjustment data list includes pressure distribution change data, flow fluctuation adjustment data and early warning structure adjustment data.

[0006] As a further aspect of the present invention, the pressure fluctuation monitoring module includes: The pressure change capture submodule captures the pressure change amplitude and frequency over different time periods based on pressure sensor data in the pipeline network, and statistically analyzes the pressure fluctuations on a periodic basis to obtain pipeline pressure change trend data. The flow fluctuation analysis submodule, based on the pipeline pressure change trend data, analyzes the distribution pattern of flow fluctuations by examining the relationship between flow fluctuation characteristics and pressure changes, extracts flow fluctuation feature data, analyzes the temporal distribution of flow anomalies, and obtains a flow fluctuation trend dataset. The anomaly pattern identification submodule analyzes the relationship between the time of anomaly occurrence and flow fluctuation based on the flow fluctuation trend dataset, and identifies the anomaly occurrence pattern by combining it with the pressure change trend, and obtains a pressure anomaly trend data table.

[0007] As a further aspect of the present invention, the traffic anomaly analysis module includes: The flow distribution visualization submodule extracts the flow change characteristics of each node based on the pressure anomaly trend data table, monitors the relationship between flow distribution and pressure fluctuations, and performs node analysis on the flow distribution to obtain a flow anomaly distribution dataset. The pressure fluctuation matching submodule, based on the traffic anomaly distribution dataset and combined with traffic fluctuation characteristics, analyzes the matching degree between traffic anomalies and pressure fluctuations, identifies the relationship between traffic distribution and pressure fluctuations, and obtains the traffic and pressure matching results. The distribution model construction submodule, based on the flow and pressure matching results, combined with the flow fluctuation characteristics and pressure change trends, analyzes the mutual influence between pressure and flow to obtain a flow anomaly distribution model.

[0008] As a further aspect of the present invention, the state association optimization module includes: The pressure fluctuation analysis submodule extracts pressure fluctuation deviation values ​​based on the flow anomaly distribution model, analyzes the range and frequency of pressure fluctuations, identifies the occurrence pattern of pressure fluctuation anomalies, and obtains pressure fluctuation deviation values. The flow distribution matching submodule, based on the pressure fluctuation deviation value, combined with the flow fluctuation characteristics and pressure distribution data, compares the matching between the flow distribution and pressure fluctuation, identifies the imbalance of the flow distribution, and obtains the flow distribution matching data. The abnormal state optimization submodule adjusts the abnormal state identification based on the flow distribution matching data, compares the matching degree between pressure fluctuation and flow distribution, optimizes the abnormal state identification, and obtains the pipeline abnormal state identification result.

[0009] As a further aspect of the present invention, the pressure fluctuation deviation value is expressed by the formula: ; in, This represents the pressure fluctuation deviation value. Representing the The pressure fluctuation deviation value in the sampling sequence at any given time. This represents the arithmetic mean of the pressure fluctuation deviations at all times in the sampling sequence. Representing the The weight of the pressure fluctuation at any given moment. Represents the total number of moments in the sampling sequence. This represents the pressure fluctuation reference value obtained based on the abnormal flow distribution model.

[0010] As a further aspect of the present invention, the early warning strategy adjustment module includes: The pressure distribution analysis submodule extracts pressure distribution change data based on the pipeline abnormality identification results, analyzes the range of pressure distribution changes, monitors pressure distribution fluctuations, identifies pressure distribution change trends, and obtains pressure distribution change results. Based on the pressure distribution change results, the flow fluctuation adjustment submodule collects flow fluctuation adjustment data, analyzes the changes in pressure distribution before and after flow fluctuation, identifies the adjustment trend of flow fluctuation, and obtains the flow fluctuation adjustment dataset. The early warning structure correction submodule adjusts the abnormal early warning structure based on the traffic fluctuation adjustment dataset, corrects the matching relationship between pressure distribution and traffic fluctuation, optimizes the abnormal early warning structure, and obtains an abnormal early warning adjustment data list.

[0011] As a further aspect of the present invention, the pressure distribution change data refers to the data on the pressure change over time at each location in the pipeline network; The pressure distribution change trend refers to the direction or pattern of pressure distribution change over a period of time; The flow fluctuation adjustment data refers to the data related to changes in pressure distribution during the flow fluctuation process; The aforementioned anomaly warning structure refers to a framework or model structure that uses an early warning mechanism to detect abnormal situations.

[0012] As a further aspect of the present invention, the system also includes an anomaly warning generation module: Based on the abnormal warning adjustment data list, the abnormal warning generation module adjusts the order of abnormal warnings according to pressure fluctuation characteristics and flow distribution, extracts the time of abnormal occurrence and pressure distribution change data, and constructs an intelligent abnormal warning scheme. The intelligent anomaly early warning scheme includes the time of anomaly occurrence, pressure distribution change data, and early warning sequence.

[0013] As a further aspect of the present invention, the anomaly warning generation module includes: The early warning sequence adjustment submodule adjusts the sequence of abnormal early warnings based on the abnormal early warning adjustment data list, evaluates the matching relationship between pressure fluctuations and flow distribution, analyzes the impact of the abnormal early warning sequence on pipeline operation, optimizes the abnormal early warning sequence, and obtains the abnormal early warning sequence adjustment result. Based on the abnormal warning sequence adjustment results, the pressure distribution analysis submodule analyzes the relationship between the pressure distribution of each node and the abnormal warning time, analyzes the changing trend of pressure distribution, and obtains pressure distribution change data. Based on the pressure distribution change data, the anomaly warning generation submodule adjusts the anomaly warning sequence and pressure distribution ratio, optimizes the anomaly warning process and pressure distribution configuration, and obtains an intelligent anomaly warning solution.

[0014] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: By conducting in-depth analysis of the relationship between pressure fluctuations and flow anomalies, the changing trends of pressure and flow in pipelines can be accurately captured, and abnormal states of the pipeline network can be identified in a timely manner. This optimizes the limitations of traditional systems that rely on threshold and statistical feature calculations. By dynamically adjusting the anomaly warning structure, it can respond more intelligently to different operating conditions of the pipeline network, effectively improving the accuracy of anomaly detection and the timeliness of warnings. In practical applications of complex pipeline networks, it can achieve real-time monitoring of the pipeline network status, avoiding missed detections or false alarms caused by improper operation or unreasonable system settings. It enhances the system's adaptability to emergencies and avoids the limitations of traditional methods that rely solely on data from a single sensor, effectively preventing the system from experiencing lag and insufficient sensitivity when facing complex operating conditions. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of an oil pipeline network anomaly identification and early warning system provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the system framework of the present invention; Figure 3 This is a flowchart of the pressure fluctuation monitoring module in this invention; Figure 4 This is a flowchart of the traffic anomaly analysis module in this invention; Figure 5 This is a flowchart of the state association optimization module in this invention; Figure 6 This is a flowchart of the early warning strategy adjustment module in this invention; Figure 7 This is a flowchart of the anomaly warning generation module in this invention. Detailed Implementation

[0017] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0018] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0019] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0020] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0021] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0022] This invention provides an oil pipeline network anomaly identification and early warning system, such as... Figure 1-2 The diagram shown illustrates an anomaly identification and early warning system for oil pipeline networks. This system includes: The pressure fluctuation monitoring module is based on pressure sensor data in the pipeline network. By capturing the amplitude and frequency of pressure changes and combining them with the characteristics of flow fluctuations, it analyzes the occurrence pattern of pressure anomalies and obtains a pressure anomaly trend data table. The flow anomaly analysis module extracts the relationship between flow change characteristics and pressure fluctuations based on the pressure anomaly trend data table, analyzes the distribution characteristics of flow anomalies, identifies key nodes of flow anomalies by combining pressure fluctuation trends, and constructs a flow anomaly distribution model. The state association optimization module is based on the flow anomaly distribution model. It extracts pipeline pressure fluctuation deviation value and flow anomaly adjustment data, compares pipeline pressure change with flow distribution characteristics, and obtains the pipeline network anomaly state identification result through pipeline pressure distribution deviation and flow anomaly adjustment. The early warning strategy adjustment module analyzes the impact of abnormal state adjustment on pipeline operation based on the pipeline abnormal state identification results, extracts pressure distribution change and flow fluctuation adjustment data, corrects the abnormal early warning structure, and obtains an abnormal early warning adjustment data list. The anomaly warning generation module adjusts the anomaly warning data list based on the anomaly warning adjustment data list, adjusts the anomaly warning order according to pressure fluctuation characteristics and flow distribution, extracts the anomaly occurrence time and pressure distribution change data, and constructs an intelligent anomaly warning scheme.

[0023] The pressure anomaly trend data table includes pressure change amplitude, pressure fluctuation frequency, and anomaly occurrence time; the flow anomaly distribution model includes flow change characteristics, pressure fluctuation relationship, and key nodes of flow anomaly; the pipeline anomaly status identification results include pressure fluctuation deviation value, flow anomaly adjustment data, and pressure and flow distribution matching relationship; the anomaly early warning adjustment data list includes pressure distribution change data, flow fluctuation adjustment data, and early warning structure adjustment data; and the intelligent anomaly early warning scheme includes anomaly occurrence time, pressure distribution change data, and early warning sequence.

[0024] Specifically, such as Figure 2 , 3 As shown, the pressure fluctuation monitoring module includes: The pressure change capture submodule captures the pressure change amplitude and frequency over different time periods based on pressure sensor data in the pipeline network, and statistically analyzes the pressure fluctuations on a periodic basis to obtain pipeline pressure change trend data. Continuous acquisition of pipeline pressure sensor data is performed. For example, node P1 acquires data once per second. The raw data undergoes quality checks and timestamp alignment. For instance, the pressure at P1 is 0.45 MPa at one moment and 0.46 MPa the next second. Subsequently, the data is divided into preset 5-minute time intervals, and the maximum and minimum pressure differences within each time interval are calculated to obtain the pressure change amplitude. For instance, the maximum is 0.52 MPa and the minimum is 0.48 MPa, with an amplitude of 0.04 MPa. At the same time, a pressure change threshold of 0.005 MPa is set. This threshold is based on the historical normal fluctuation range and is used to identify effective pressure fluctuations. The number of times the pressure exceeds the threshold by 0.50 MPa above or below the average value of this segment is counted as the frequency. For instance, if it exceeds the threshold by 8 times, the pressure change amplitude and frequency data are then summarized by hourly cycle, and their average, maximum, and minimum values ​​are calculated. For instance, if the average amplitude of P1 is 0.035 MPa and the average frequency is 6 times / 5 minutes, the statistical data is the pipeline pressure change trend data.

[0025] The flow fluctuation analysis submodule is based on pipeline pressure change trend data. By analyzing the relationship between flow fluctuation characteristics and pressure changes, it analyzes the distribution pattern of flow fluctuation, extracts flow fluctuation feature data, analyzes the temporal distribution of flow anomalies, and obtains a flow fluctuation trend dataset. Based on pipeline pressure change trend data, such as the hourly average pressure change amplitude and frequency data of node P1 over the past 24 hours, a relationship model between flow fluctuation characteristics and pressure changes is constructed. This model, based on fluid mechanics principles and historical data regression analysis, describes the impact of pressure change amplitude and frequency on flow fluctuations. For example, when the pressure change amplitude is 0.035 MPa and the frequency is 6 times / 5 minutes, the predicted flow fluctuation amplitude is 5.35 m. 3 / h, then, extract flow fluctuation characteristic data, for example, identify situations where the flow fluctuation exceeds ±15% of the normal flow, this threshold is based on historical pipe bursts or large valve operations, for example, flow from 100m 3 / h dropped to 60m 3 / h, fluctuation range 40m 3 / h, lasting 3 minutes, marked as abnormal, and time series analysis is performed on the extracted traffic anomaly feature data. For example, it is identified that the traffic anomalies are mainly concentrated between 2:00 am and 4:00 am, and finally the traffic fluctuation trend dataset is obtained.

[0026] The anomaly pattern identification submodule analyzes the relationship between the time of anomaly occurrence and traffic fluctuation based on the traffic fluctuation trend dataset, and identifies the anomaly occurrence pattern by combining it with the pressure change trend, and obtains a pressure anomaly trend data table; Based on a traffic fluctuation trend dataset, for example, if the frequency of abnormal traffic increases at node P1 between 2:00 AM and 4:00 AM, the time of the anomaly is correlated with the characteristics of traffic fluctuation. For example, between 2:30 AM and 2:45 AM, the traffic of node P1 suddenly increases from the normal range to 30m. 3 / h, and simultaneously combined with pressure change trend data during that period, for example, the pressure drops sharply from 0.55MPa to 0.35MPa, and the frequency increases from 4 times every 5 minutes to 15 times. By comparing and matching the abnormal flow characteristics with the pressure change trend during the same period, the pattern of abnormal occurrence can be identified. For example, when the flow fluctuation is "significantly prominent" (exceeding 3 times the normal range or an absolute value change of 20m), the abnormality can be identified. 3 If a pipe bursts and leaks (the threshold is based on historical leakage data) and the pressure drops sharply (the rate exceeds 0.05 MPa / min and the total pressure exceeds 0.1 MPa), it is identified as a "pipe burst leak," and a pressure anomaly trend data table is finally obtained.

[0027] Specifically, such as Figure 2 , 4 As shown, the traffic anomaly analysis module includes: The traffic distribution visualization submodule extracts the traffic change characteristics of each node based on the pressure anomaly trend data table, monitors the relationship between traffic distribution and pressure fluctuations, and performs node analysis on the traffic distribution to obtain a traffic anomaly distribution dataset. Based on the pressure anomaly trend data table, for example, containing the "pipe burst leakage" anomaly record of node P1, the flow change characteristics of each node are extracted. For example, the flow rate of node P1 during the abnormal period decreased from the normal 100m³ / h. 3 / h dropped to 60m 3 / h, the standard deviation of flow increases. The instantaneous rate of change is obtained by comparing real-time and historical flow data. Then, the relationship between flow distribution and pressure fluctuation is monitored. The flow characteristics of nodes are compared with the pressure fluctuations of the same period. For example, when the pressure drops by 0.20MPa, the flow of P1 drops synchronously. There is a strong correlation between the two. The correlation rule is that when the pressure drops by more than 10%, the flow drops by more than 20%. This rule is based on historical pipeline fault data. Then, node analysis is performed on the flow distribution to check the flow relationship between each node and its adjacent nodes. For example, when the flow of P1 drops, the flow of P2 and P3 also drops. Flow imbalance is identified by graph theory algorithm and flow balance equation. For example, the inflow-outflow imbalance of P1 exceeds 10% and is marked as abnormal. Finally, the abnormal flow distribution dataset is obtained.

[0028] The pressure fluctuation matching submodule analyzes the matching degree between flow anomalies and pressure fluctuations based on the flow anomaly distribution dataset and the characteristics of flow fluctuations, identifies the relationship between flow distribution and pressure fluctuations, and obtains the flow and pressure matching results. Based on a traffic anomaly distribution dataset, for example, the traffic of node P1 drops by 40m. 3 / h, P2 drops 30m 3 / h, combining flow fluctuation characteristics, analyze the matching degree between flow anomalies and pressure fluctuations. Quantify the synchronicity of flow anomalies and concurrent pressure fluctuation events in terms of time, amplitude, and duration. For example, a time overlap rate exceeding 80% is considered "high," and an amplitude correlation coefficient exceeding 0.8 is also considered "high." This threshold is based on historical pipeline data analysis. Calculate matching scores for each matching quantitative indicator. For example, set the matching score weights as follows: time matching 0.4, amplitude matching 0.3, and duration matching 0.3. The weights are based on expert experience and historical data. For example, if the matching score of node P1 reaches 0.9, a strong correlation between flow distribution and pressure fluctuation is identified. This threshold of 0.7 is used to distinguish between strong and weak correlations, ultimately obtaining the flow and pressure matching results.

[0029] The distribution model construction submodule, based on the flow and pressure matching results, combined with the flow fluctuation characteristics and pressure change trends, analyzes the mutual influence between pressure and flow to obtain a flow anomaly distribution model; Based on the matching results of flow and pressure, such as the high matching degree and strong correlation between nodes P1, P2, and P3, and combined with flow fluctuation characteristics and pressure change trend data, a graph neural network model based on the pipeline topology and historical operating data is constructed. This model captures the complex nonlinear relationship between pressure and flow. For example, in a pipe burst leakage scenario, the model learns that a large flow leakage causes a sharp drop in regional pressure, which is positively correlated with the leakage flow rate. For example, if P1 leaks 40m... 3When the pressure drops by 0.2 MPa per hour, the model infers the abnormal flow rate that caused the pressure fluctuations. For example, the periodic small oscillations of P1 are related to the instability of the control valve. Finally, a flow rate anomaly distribution model is generated, which can predict or explain the flow rate anomaly distribution pattern and propagation path in real time.

[0030] Specifically, such as Figure 2 , 5 As shown, the state association optimization module includes: The pressure fluctuation analysis submodule is based on the flow anomaly distribution model, extracts pressure fluctuation deviation values, analyzes the range and frequency of pressure fluctuations, identifies the occurrence pattern of pressure fluctuation anomalies, and obtains pressure fluctuation deviation values. The pressure fluctuation deviation value is calculated using the following formula: ; in, This represents the pressure fluctuation deviation value. Representing the The pressure fluctuation deviation value in the sampling sequence at any given time. This represents the arithmetic mean of the pressure fluctuation deviations at all times in the sampling sequence. Representing the The weight of the pressure fluctuation at any given moment. Represents the total number of moments in the sampling sequence. This represents a reference value for pressure fluctuations obtained based on a traffic flow anomaly distribution model. Based on the flow anomaly distribution model, a model capable of predicting or explaining flow anomaly distribution patterns in the pipeline network is obtained. Pressure fluctuation deviation values ​​are extracted, representing the difference between the actual monitored pressure fluctuations and the normal pressure fluctuations predicted based on the flow anomaly distribution model. By inputting real-time collected pressure data from each node into the flow anomaly distribution model, the model outputs a reference value for the expected pressure fluctuation under the current flow distribution. Then, the actual pressure fluctuations were monitored. and Compare and calculate the deviation; for example, if the model predicts that the pressure fluctuation at node P1 should be under the current flow distribution... Within, while the actual monitored pressure fluctuations are Then it exists The deviation was calculated by statistically analyzing the pressure fluctuation deviation values ​​at each moment in the sampling sequence, and the total number of moments in the sampling sequence was [not specified]. Set as (Each sample corresponds to 5 minutes of data, 1 sample per second). For example, the pressure fluctuation data of node P1 within a specific time period is shown in Table 1. Regarding the pressure fluctuation deviation value... The calculation uses the formula. ,in This represents the pressure fluctuation deviation value, a comprehensive indicator that reflects the degree to which actual pressure fluctuations deviate from their average and reference values. This represents the pressure fluctuation reference value obtained based on the flow anomaly distribution model. For example, according to the flow anomaly distribution model, under the current flow conditions, the pressure fluctuation reference value of node P1 is... Set as This value is the center value of the normal pressure fluctuation range inferred by the model based on the current pipeline flow distribution and historical data, ensuring... The rationality and reproducibility, Representing the The pressure fluctuation deviation value at time 1 in the sampling sequence is defined as the value of the pressure fluctuation deviation at time 2. The absolute difference between the actual pressure value at a given moment and the average pressure value over the 5-minute time interval containing that moment. For example, if the actual pressure at a certain moment is... The average pressure over this 5-minute period is ,but , This represents the arithmetic mean of the pressure fluctuation deviations at all times in the sampling sequence, i.e. For example, for the data in Table 1, , Representing the The weight of the pressure fluctuation at any given moment. The settings are based on the degree of deviation between the pressure fluctuation at that moment and the historical normal fluctuation distribution. The greater the deviation, the greater the weight. For example, it can be set to... ,in It is the average of historical normal pressure fluctuations, while threshold is a benchmark value used to quantify the degree of deviation. For example, Set as Threshold is set to Experiments have shown that when the weight increases linearly with the degree of deviation, it can more accurately reflect the importance of abnormal fluctuations. For the data in Table 1, it is assumed that all... All are set to 1 (i.e., weight differences are not considered). This represents the total number of moments in the sampling sequence. For example, within the aforementioned 5 minutes, 10 sampling points are used as an example. The advantage of this formula lies in the introduction of... This item can effectively reflect the degree to which a single pressure fluctuation deviates from the average level, combined with... The weights highlight the contribution of abnormal fluctuations to the overall deviation, while the denominator... This term normalizes the deviation value by taking into account normal fluctuation levels, making the deviation values ​​under different operating conditions comparable. For example, substituting the above example data into the formula for calculation: First, calculate the molecular part: ; Then calculate the denominator: ; Final calculation : ; This result indicates that the current pressure fluctuation deviates to the normal reference value and average value to a certain extent. The numerical result It is the pressure fluctuation deviation value, which directly quantifies the instability of pressure fluctuations; a higher value indicates a higher risk. This value indicates a significant anomaly in pressure fluctuations. The value is compared with a preset anomaly threshold. For example, if the anomaly threshold is set to 1... ,but Exceeding the threshold indicates abnormal pressure fluctuations. Subsequently, the range and frequency of these fluctuations are analyzed. For example, based on the calculated pressure fluctuation deviation, further analysis is conducted to determine whether the amplitude of the actual pressure fluctuation exceeds a preset normal range (e.g., 0.01-0.05 MPa) and whether the number of fluctuations per unit time exceeds the normal frequency (e.g., 1-5 fluctuations per minute). For instance, if a pressure fluctuation amplitude of [value missing] is detected within a 5-minute timeframe... Furthermore, the fluctuation frequency reached as high as 8 times per minute, which exceeded the normal range. Subsequently, the occurrence pattern of the pressure fluctuation anomaly was identified. By performing time series analysis on the continuously monitored pressure fluctuation deviation values, ranges, and frequencies, the periodicity, suddenness, or persistence of the pressure fluctuation anomaly was identified. For example, [the text abruptly ends here, likely due to an incomplete sentence or missing information]. The value remains at If the pressure fluctuation amplitude and frequency remain high, it is identified as a persistent pressure anomaly, which indicates that there are long-term unstable factors in the pipeline network. The pressure fluctuation deviation value is obtained, and the deviation value and its corresponding analysis results are recorded as key inputs for subsequent flow distribution matching. Table 1: Example Table of Pressure Fluctuation Data at Node P1

[0031] Table 1 lists the actual pressure, 5-minute average pressure, and calculated pressure fluctuation deviation of node P1 at a specific sampling time. The data is used to calculate the pressure fluctuation deviation value. .

[0032] The flow distribution matching submodule, based on the pressure fluctuation deviation value, combines the flow fluctuation characteristics and pressure distribution data to compare the matching between the flow distribution and pressure fluctuation, identify the imbalance of the flow distribution, and obtain the flow distribution matching data. Based on the pressure fluctuation deviation value, for example, Q=0.3616 for node P1, combined with flow fluctuation characteristics and real-time pressure distribution data, the matching between flow distribution and pressure fluctuation is compared. By calculating the similarity between the pressure fluctuation feature vector and the flow fluctuation feature vector, for example, by calculating the cosine similarity, if it is higher than 0.85, this threshold is used to distinguish between high matching and low matching, and the matching degree is considered "high", indicating that there is a high degree of consistency between the two. For example, the similarity between the pressure fluctuation and the flow drop at node P1 reaches 0.9. Subsequently, the imbalance of flow distribution is identified. When the matching degree is high, the pipeline flow distribution is further analyzed to see if there is a local "prominence" (absolute deviation of regional flow exceeds 20% and affects more than 3 nodes, this threshold identifies local imbalance caused by anomalies) or "significantly low" (flow deviation exceeds 30% and lasts for more than 10 minutes, this threshold identifies leakage or blockage). For example, the flow of downstream P2 and P3 of P1 decreases synchronously, indicating that there is a flow imbalance. Finally, the flow distribution matching data is obtained.

[0033] The abnormal state optimization submodule adjusts the abnormal state identification based on flow distribution matching data, compares the matching degree between pressure fluctuation and flow distribution, optimizes the abnormal state identification, and obtains the pipeline abnormal state identification result. Based on traffic distribution matching data, for example, node P1 has Q=0.3616 and a traffic decrease of 40m. 3 Based on the matching degree of 0.9 and the identification results of regional flow imbalance, the initially identified abnormal state is adjusted. For example, if it is initially identified as "local blockage", but the data shows that the pressure drop and flow drop have a high matching degree and involve multiple nodes, it is adjusted to "large-scale leakage". This adjustment is based on more comprehensive data fusion verification, comparing the matching degree of pressure fluctuation and flow distribution, and evaluating the correlation strength of different regions or abnormal events. For example, a regional matching degree higher than 0.8 is "high" and lower than 0.5 is "low". A high matching degree indicates a high homology of anomalies. Subsequently, based on the matching degree comparison results, the classification and severity of abnormal states are further optimized. For example, areas with high matching degree concentration are identified as "regional large-scale leakage", and confidence levels are assigned according to the matching degree score. For example, a confidence level of 0.95 is "high", and the final pipeline abnormal state identification result is obtained.

[0034] Specifically, such as Figure 2 , 6 As shown, the early warning strategy adjustment module includes: The pressure distribution analysis submodule extracts pressure distribution change data based on the pipeline abnormality identification results, analyzes the range of pressure distribution changes, monitors pressure distribution fluctuations, identifies pressure distribution change trends, and obtains pressure distribution change results. Pressure distribution change data refers to the data on how the pressure at each location in the pipeline network changes over time. The trend of pressure distribution change refers to the direction or pattern of pressure distribution change over a period of time; Based on the results of pipeline network anomaly identification, for example, if a "large-scale leak" is identified as occurring near node P1, pressure distribution change data is extracted, i.e., the original pressure sensor data of all monitoring points in the pipeline network before and after the anomaly occurred, such as all pressure readings per second from 2:00 AM to 3:00 AM. Subsequently, the range of pressure distribution change is analyzed, and the minimum, maximum, and average pressure values ​​of each monitoring point, as well as the overall average pressure change amplitude of the pipeline network, are calculated. For example, the pressure at P1 drops from 0.55 MPa to 0.35 MPa, a change of 0.20 MPa. Then, the fluctuation of pressure distribution is monitored, and pressure-time curves are plotted in real time, and the sliding window standard deviation is calculated. For example, when the standard deviation of the pressure at a certain node exceeds 0.01 MPa within 1 minute, it is judged as "violent" pressure fluctuation. This threshold is based on normal pipeline network operation experience. Subsequently, the trend of pressure distribution change is identified. For example, through trend line fitting, it is found that the pressure at P1 shows a rapid linear downward trend with a rate of decrease of 0.005 MPa / second, and the final pressure distribution change result is obtained.

[0035] The flow fluctuation adjustment submodule collects flow fluctuation adjustment data based on the pressure distribution change results, analyzes the changes in pressure distribution before and after flow fluctuation, identifies the adjustment trend of flow fluctuation, and obtains the flow fluctuation adjustment dataset. Flow fluctuation adjustment data refers to data related to changes in pressure distribution during flow fluctuations; Based on pressure distribution changes, such as the sudden pressure drop trend in node P1 caused by a "large-scale leak," flow fluctuation adjustment data is collected. This is achieved by real-time monitoring of the operating status and adjustment parameters of flow regulation equipment such as pumps and valves, as well as the flow changes at each node before and after the adjustment operation. For example, if the water supply pump power is increased from 500kW to 600kW, the flow rate at P1 is recorded as decreasing from 60m³ / h. 3 / h rises to 80m 3 / h, then, analyze the changes in pressure distribution before and after the flow fluctuation, compare the flow adjustment data with the corresponding pressure distribution change data, for example, increasing the pump power by 100kW causes the P1 pressure to rise by 0.05MPa, identify the adjustment trend of flow fluctuation, and by analyzing multiple adjustment operations and their pressure effects, identify the impact pattern of different flow adjustment strategies on pipeline pressure, for example, increasing the upstream flow by x% can cause the pressure in the affected area to rise by y%, and finally obtain the flow fluctuation adjustment dataset.

[0036] The early warning structure correction submodule adjusts the abnormal early warning structure based on the traffic fluctuation adjustment dataset, corrects the matching relationship between pressure distribution and traffic fluctuation, optimizes the abnormal early warning structure, and obtains an abnormal early warning adjustment data list. Anomaly warning structure refers to a framework or model structure that uses an early warning mechanism to detect abnormal situations; Based on the flow fluctuation adjustment dataset, such as data on the impact of pump power and valve opening adjustments on pressure and flow, the abnormal early warning structure is adjusted. The parameters, thresholds, or logical rules of the current early warning model are corrected according to the actual adjustment effect. For example, if the adjustment successfully alleviates a sudden pressure drop, the pressure drop threshold for high-level alarms is corrected from 0.1 MPa to 0.08 MPa. Subsequently, the matching relationship between pressure distribution and flow fluctuation is corrected. Based on the actual monitored linkage changes, the matching weight in the early warning structure is recalibrated. For example, the matching coefficient between the flow fluctuation pattern and a specific pressure distribution change is increased from 0.8 to 0.9 to ensure early warning accuracy. Then, the abnormal early warning structure is optimized, and the early warning triggering logic, level classification, and priority are adjusted. For example, the high-priority "pipe burst leakage" trigger condition is more sensitive and has a shorter response time. Finally, an abnormal early warning adjustment data list is obtained.

[0037] Specifically, such as Figure 2 , 7 As shown, the anomaly warning generation module includes: The early warning sequence adjustment submodule adjusts the order of abnormal early warnings based on the abnormal early warning adjustment data list, evaluates the matching relationship between pressure fluctuations and flow distribution, analyzes the impact of the abnormal early warning sequence on pipeline operation, optimizes the abnormal early warning sequence, and obtains the abnormal early warning sequence adjustment results. Based on the abnormal warning adjustment data list, for example, including the list with the highest priority "large-scale leakage" warning after optimization, the order of abnormal warnings is adjusted, with "pipe burst leakage" being prioritized to ensure that the most impactful events are handled first. Subsequently, the matching relationship between pressure fluctuations and flow distribution is evaluated to verify whether the matching degree between the pressure drop and flow drop corresponding to high-priority warning events (such as "pipe burst leakage") is always at a "high" level (e.g., matching degree score greater than 0.9), avoiding false alarms. The impact of the abnormal warning order on pipeline operation is analyzed. By simulating and analyzing historical data, the impact of the adjusted warning order on the stable operation of the pipeline network, water supply efficiency, and user experience is evaluated. For example, the priority of warnings for small-scale pressure fluctuations is reduced to reduce alarm fatigue. Finally, the abnormal warning order is optimized to obtain the result of the abnormal warning order adjustment.

[0038] The pressure distribution analysis submodule analyzes the relationship between the pressure distribution of each node and the abnormal warning time based on the abnormal warning sequence adjustment results, analyzes the changing trend of pressure distribution, and obtains pressure distribution change data. Based on the results of adjusting the order of abnormal warnings, such as including an optimized and sorted list of abnormal warnings, the relationship between the pressure distribution of each node and the abnormal warning time is analyzed. Historical pressure distribution patterns are associated with abnormal warning timestamps. For example, if the P1 pressure starts to drop sharply 5 minutes before the "large-scale leakage" warning is triggered, and the warning is triggered when the pressure drops by 0.1 MPa, the pressure change and the warning trigger time delay are quantified. Subsequently, the trend of pressure distribution change is analyzed, and the pressure distribution of each node is analyzed in depth. For example, if the P1 pressure drops sharply from 0.55 MPa to 0.35 MPa at a rate of 0.005 MPa / second for 40 seconds, linear regression is used to identify that the pressure shows a rapid linear downward trend. Finally, pressure distribution change data is obtained, and the pressure value, change magnitude, rate, and trend type before and after the abnormality of each node are recorded in detail.

[0039] The anomaly warning generation submodule adjusts the anomaly warning sequence and pressure distribution ratio based on pressure distribution change data, optimizes the anomaly warning process and pressure distribution configuration, and obtains an intelligent anomaly warning solution. Based on pressure distribution change data, such as the rate and trend of pressure drop before and after the "large-scale leakage" warning is triggered at node P1, the order of abnormal warnings and the proportion of pressure distribution are adjusted. If the pressure drop rate in a certain area is "very fast" (exceeding 0.01 MPa / second, this threshold captures emergencies), the priority of the "large-scale leakage" warning in that area is temporarily raised to the highest level. At the same time, according to the magnitude, speed and scope of pressure change, the proportion of attention given to different areas or types of anomalies in the warning information is dynamically adjusted. For example, if the pressure drop in area P1 reaches 0.20 MPa, a higher display priority is assigned. Subsequently, the abnormal warning process and pressure distribution configuration are optimized to shorten the response time of high-priority warnings, and related flow regulation equipment is dynamically configured. For example, when the pressure is predicted to drop to a dangerous threshold, it is recommended to immediately start the backup pump or close the relevant valves, thus obtaining an intelligent abnormal warning solution.

[0040] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A system for identifying and warning of anomalies in an oil pipeline network, characterized in that, The system includes: The pressure fluctuation monitoring module is based on pressure sensor data in the pipeline network. By capturing the amplitude and frequency of pressure changes and combining them with the characteristics of flow fluctuations, it analyzes the occurrence pattern of pressure anomalies and obtains a pressure anomaly trend data table. Based on the pressure anomaly trend data table, the flow anomaly analysis module extracts the relationship between flow change characteristics and pressure fluctuations, analyzes the distribution characteristics of flow anomalies, identifies key nodes of flow anomalies by combining pressure fluctuation trends, and constructs a flow anomaly distribution model. Based on the aforementioned abnormal flow distribution model, the state association optimization module extracts pipeline pressure fluctuation deviation values ​​and abnormal flow adjustment data, compares pipeline pressure changes with flow distribution characteristics, and obtains the pipeline network abnormal state identification results through pipeline pressure distribution deviation and abnormal flow adjustment. Based on the pipeline abnormality identification results, the early warning strategy adjustment module analyzes the impact of abnormality adjustment on pipeline operation, extracts pressure distribution change and flow fluctuation adjustment data, corrects the abnormality early warning structure, and obtains an abnormality early warning adjustment data list.

2. The oil pipeline network anomaly identification and early warning system according to claim 1, characterized in that: The pressure anomaly trend data table includes pressure change amplitude, pressure fluctuation frequency, and anomaly occurrence time. The flow anomaly distribution model includes flow change characteristics, pressure fluctuation relationship, and key nodes of flow anomaly. The pipeline anomaly status identification results include pressure fluctuation deviation value, flow anomaly adjustment data, and pressure and flow distribution matching relationship. The anomaly early warning adjustment data list includes pressure distribution change data, flow fluctuation adjustment data, and early warning structure adjustment data.

3. The oil pipeline network anomaly identification and early warning system according to claim 1, characterized in that: The pressure fluctuation monitoring module includes: The pressure change capture submodule captures the pressure change amplitude and frequency over different time periods based on pressure sensor data in the pipeline network, and statistically analyzes the pressure fluctuations on a periodic basis to obtain pipeline pressure change trend data. The flow fluctuation analysis submodule, based on the pipeline pressure change trend data, analyzes the distribution pattern of flow fluctuations by examining the relationship between flow fluctuation characteristics and pressure changes, extracts flow fluctuation feature data, analyzes the temporal distribution of flow anomalies, and obtains a flow fluctuation trend dataset. The anomaly pattern identification submodule analyzes the relationship between the time of anomaly occurrence and flow fluctuation based on the flow fluctuation trend dataset, and identifies the anomaly occurrence pattern by combining it with the pressure change trend, and obtains a pressure anomaly trend data table.

4. The oil pipeline network anomaly identification and early warning system according to claim 3, characterized in that: The traffic anomaly analysis module includes: The flow distribution visualization submodule extracts the flow change characteristics of each node based on the pressure anomaly trend data table, monitors the relationship between flow distribution and pressure fluctuations, and performs node analysis on the flow distribution to obtain a flow anomaly distribution dataset. The pressure fluctuation matching submodule, based on the traffic anomaly distribution dataset and combined with traffic fluctuation characteristics, analyzes the matching degree between traffic anomalies and pressure fluctuations, identifies the relationship between traffic distribution and pressure fluctuations, and obtains the traffic and pressure matching results. The distribution model construction submodule, based on the flow and pressure matching results, combined with the flow fluctuation characteristics and pressure change trends, analyzes the mutual influence between pressure and flow to obtain a flow anomaly distribution model.

5. The oil pipeline network anomaly identification and early warning system according to claim 4, characterized in that: The state association optimization module includes: The pressure fluctuation analysis submodule extracts pressure fluctuation deviation values ​​based on the flow anomaly distribution model, analyzes the range and frequency of pressure fluctuations, identifies the occurrence pattern of pressure fluctuation anomalies, and obtains pressure fluctuation deviation values. The flow distribution matching submodule, based on the pressure fluctuation deviation value, combined with the flow fluctuation characteristics and pressure distribution data, compares the matching between the flow distribution and pressure fluctuation, identifies the imbalance of the flow distribution, and obtains the flow distribution matching data. The abnormal state optimization submodule adjusts the abnormal state identification based on the flow distribution matching data, compares the matching degree between pressure fluctuation and flow distribution, optimizes the abnormal state identification, and obtains the pipeline abnormal state identification result.

6. The oil pipeline network anomaly identification and early warning system according to claim 5, characterized in that: The pressure fluctuation deviation value is calculated using the following formula: ； in, This represents the pressure fluctuation deviation value. Representing the The pressure fluctuation deviation value in the sampling sequence at any given time. This represents the arithmetic mean of the pressure fluctuation deviations at all times in the sampling sequence. Representing the The weight of the pressure fluctuation at any given moment. Represents the total number of moments in the sampling sequence. This represents the pressure fluctuation reference value obtained based on the abnormal flow distribution model.

7. The oil pipeline network anomaly identification and early warning system according to claim 5, characterized in that: The early warning strategy adjustment module includes: The pressure distribution analysis submodule extracts pressure distribution change data based on the pipeline abnormality identification results, analyzes the range of pressure distribution changes, monitors pressure distribution fluctuations, identifies pressure distribution change trends, and obtains pressure distribution change results. Based on the pressure distribution change results, the flow fluctuation adjustment submodule collects flow fluctuation adjustment data, analyzes the changes in pressure distribution before and after flow fluctuation, identifies the adjustment trend of flow fluctuation, and obtains the flow fluctuation adjustment dataset. The early warning structure correction submodule adjusts the abnormal early warning structure based on the traffic fluctuation adjustment dataset, corrects the matching relationship between pressure distribution and traffic fluctuation, optimizes the abnormal early warning structure, and obtains an abnormal early warning adjustment data list.

8. The oil pipeline network anomaly identification and early warning system according to claim 7, characterized in that: The pressure distribution change data refers to the data on the pressure change at each location in the pipeline network over time. The pressure distribution change trend refers to the direction or pattern of pressure distribution change over a period of time; The flow fluctuation adjustment data refers to the data related to changes in pressure distribution during the flow fluctuation process; The aforementioned anomaly warning structure refers to a framework or model structure that uses an early warning mechanism to detect abnormal situations.

9. The oil pipeline network anomaly identification and early warning system according to claim 1, characterized in that: The system also includes an anomaly warning generation module: Based on the abnormal warning adjustment data list, the abnormal warning generation module adjusts the order of abnormal warnings according to pressure fluctuation characteristics and flow distribution, extracts the time of abnormal occurrence and pressure distribution change data, and constructs an intelligent abnormal warning scheme. The intelligent anomaly early warning scheme includes the time of anomaly occurrence, pressure distribution change data, and early warning sequence.

10. The oil pipeline network anomaly identification and early warning system according to claim 9, characterized in that: The anomaly warning generation module includes: The early warning sequence adjustment submodule adjusts the sequence of abnormal early warnings based on the abnormal early warning adjustment data list, evaluates the matching relationship between pressure fluctuations and flow distribution, analyzes the impact of the abnormal early warning sequence on pipeline operation, optimizes the abnormal early warning sequence, and obtains the abnormal early warning sequence adjustment result. Based on the abnormal warning sequence adjustment results, the pressure distribution analysis submodule analyzes the relationship between the pressure distribution of each node and the abnormal warning time, analyzes the changing trend of pressure distribution, and obtains pressure distribution change data. Based on the pressure distribution change data, the anomaly warning generation submodule adjusts the anomaly warning sequence and pressure distribution ratio, optimizes the anomaly warning process and pressure distribution configuration, and obtains an intelligent anomaly warning solution.

Images