A pressure detection and early warning method and system based on pipeline network data

Abstract

The present invention belongs to the technical field of pipeline network detection. The present invention provides a pressure detection and early warning method and system based on pipeline network data, including: by comparing and analyzing the positions of historical abnormal pipeline network detection nodes and the set pipeline network detection nodes, finding out the missed detection nodes of the pipeline network, and predicting the pressure abnormal time points of the missed detection nodes of the pipeline network to determine whether the missed detection nodes of the pipeline network will have pressure abnormalities during the current pipeline network detection period, improving the accuracy of pipeline network pressure detection, simulating the pressure distribution of the pipeline network under different working conditions, finding the normal pressure ranges of each pipeline network detection node under different working conditions, solving the problem of inaccurate or misjudged pipeline network pressure detection caused by different normal pressure ranges of pipeline network detection nodes due to different working conditions, and identifying emergency warning pipeline network detection nodes and necessary warning pipeline network detection nodes, improving the timeliness of pipeline network pressure detection and early warning.

Description

Technical Field

[0001] The present invention belongs to the technical field of pipeline network detection, and specifically relates to a pressure detection and early warning method and system based on pipeline network data. Background Art

[0002] There are many problems in the existing technology regarding the setting of pipeline network detection nodes and pressure detection. For example, in the setting of pipeline network detection nodes, there is a lack of systematic location comparison and analysis of historical abnormal nodes and existing nodes, making it difficult to accurately discover missing detection nodes, resulting in ineffective monitoring of some key positions, leaving blind spots in pipeline network detection and preventing a comprehensive understanding of the operation status of the pipeline network. Moreover, the existing technology does not predict the time points of pressure anomalies for missing detection nodes, and cannot judge in advance whether these nodes will have pressure anomalies during the current detection period, which is not conducive to the reasonable planning of detection work and resource allocation. In terms of pipeline network pressure detection, since the pressure distribution of the pipeline network is different under different working conditions, and the existing technology fails to simulate the pressure distribution of the pipeline network for each working condition to determine the normal pressure range of each detection node under different working conditions, it is easy to cause inaccurate or misjudged pressure detection due to changes in working conditions, and it is impossible to detect pipeline network pressure anomalies in a timely and accurate manner. At the same time, the existing technology can only identify pipeline network detection nodes that have already had pressure anomalies, lacking predictive analysis of pressure changes, and cannot identify in advance nodes that may have pressure anomalies, resulting in untimely early warning of pipeline network pressure detection, unable to take effective measures in the initial stage of pressure anomalies, which may lead to the expansion of problems and affect the safe and stable operation of the pipeline network.

[0003] Therefore, the present invention provides a pressure detection and early warning method and system based on pipeline network data. Summary of the Invention

[0004] To make up for the deficiencies of the existing technology and solve at least one of the technical problems proposed in the background art.

[0005] The technical solution adopted by the present invention to solve its technical problems is as follows:

[0006] A pressure detection and early warning method based on pipeline network data, comprising:

[0007] Comparing the positions of historical abnormal pipeline network detection nodes and the already set pipeline network detection nodes to identify missing pipeline network detection nodes;

[0008] Predicting whether the missing pipeline network detection nodes will have pressure anomalies during the current pipeline network detection period through the pressure anomaly interval time series of the missing pipeline network detection nodes in the historical operation cycle of the pipeline network. If so, marking the missing pipeline network detection nodes as necessary pipeline network detection nodes;

[0009] Adding new pipeline network detection nodes according to the positions of the necessary pipeline network detection nodes to optimize the setting of pipeline network detection nodes;

[0010] After the optimal setting of the pipeline network detection nodes is completed, the pressure distribution of the pipeline network under different working conditions is simulated through the pipeline network pressure model to determine the normal pressure range of each pipeline network detection node under different working conditions;

[0011] Compare the pressure of each detected pipeline network detection node with the normal pressure range of each pipeline network detection node under the corresponding working condition to identify emergency warning pipeline network detection nodes and necessary warning pipeline network detection nodes.

[0012] Preferably, the identification method of the pipeline network omission detection node is as follows:

[0013] Mark the historical abnormal pipeline network detection nodes that are different from the positions of all the set pipeline network detection nodes as pipeline network omission detection nodes.

[0014] Preferably, the process of predicting whether the pipeline network omission detection node will have abnormal pressure during the current pipeline network detection period is as follows:

[0015] Conduct a stationarity analysis on the pressure anomaly interval time series to determine the predicted value of the anomaly interval time, and combine it with the last pressure anomaly time point of the pipeline network omission detection node to determine the predicted pressure anomaly time point;

[0016] If the predicted pressure anomaly time point is within the current pipeline network detection period, it means that the pipeline network omission detection node will have abnormal pressure during the current pipeline network detection period.

[0017] Preferably, the process of conducting a stationarity analysis on the pressure anomaly interval time series is as follows:

[0018] Calculate the standard deviation and mean of the pressure anomaly interval time series, and perform a ratio calculation on the standard deviation and mean to obtain the stationarity value of the pressure anomaly interval time series;

[0019] If the stationarity value is greater than or equal to the stationarity threshold, it means it is non-stationary; then use the LSTN network to process the pressure anomaly interval time series to determine the predicted value of the anomaly interval time

[0020] If the stationarity value is less than the stationarity threshold, it means it is stationary, then use the historical average method to perform mean processing on the pressure anomaly interval time series to determine the predicted value of the anomaly interval time.

[0021] Preferably, the determination method of the normal pressure range of each pipeline network detection node under different working conditions is as follows:

[0022] Through the simulation of the pressure distribution of the pipeline network under different working conditions, obtain the maximum pressure and minimum pressure of each pipeline network detection node during the normal operation of the pipeline network under different working conditions, and construct the normal pressure range of each pipeline network detection node under different working conditions.

[0023] Preferably, the identification method for the emergency warning pipeline network detection nodes and the necessary warning pipeline network detection nodes is as follows:

[0024] If the pressure of the pipeline network detection node is not within the normal pressure range, then mark the pipeline network detection node as an emergency warning pipeline network detection node; otherwise, mark the pipeline network detection node as a non-emergency warning pipeline network detection node.

[0025] Conduct a pressure critical analysis on the non-emergency warning pipeline network detection nodes to determine whether pressure prediction is required. If so, predict whether the pressure of the non-emergency warning pipeline network detection node will reach the edge of the normal pressure range endpoint within the current pipeline network detection cycle. If it will, then mark the non-emergency warning pipeline network detection node as a necessary warning pipeline network detection node.

[0026] Preferably, the pressure critical analysis process is as follows:

[0027] If the pressure of the non-emergency warning pipeline network detection node is not within the pressure critical range, then generate a proximity analysis signal.

[0028] Based on the proximity analysis signal, if the pressure value of the non-emergency warning pipeline network detection node exceeds the maximum value within the pressure critical range, then perform an absolute difference operation on the pressure value of the non-emergency warning pipeline network detection node and the maximum value within the normal pressure range to obtain the pressure critical value of the non-emergency warning pipeline network detection node.

[0029] If the pressure value of the non-emergency warning pipeline network detection node is lower than the minimum value within the pressure critical range, then perform an absolute difference operation on the pressure value of the non-emergency warning pipeline network detection node and the minimum value within the normal pressure range to obtain the pressure critical value of the non-emergency warning pipeline network detection node.

[0030] If the pressure critical value is less than or equal to the pressure critical threshold, then generate a prediction signal.

[0031] Preferably, the process of predicting whether the pressure of the non-emergency warning pipeline network detection node will reach the edge of the normal pressure range endpoint within the current pipeline network detection cycle includes:

[0032] Obtain the departure time point through the time point when the pressure value of the non-emergency warning pipeline network detection node departs from the pressure critical range.

[0033] Fit the pressure values of the non-emergency warning pipeline network detection node within the period from the departure time point to the prediction signal generation time point to obtain a pressure fitting line and calculate the coefficient of determination.

[0034] Judge whether the pressure value of the non-emergency warning pipeline network detection node changes linearly or non-linearly through the coefficient of determination and determine the pressure edge arrival time point.

[0035] If the pressure edge arrival time point is within the current pipeline network detection period, mark the non-emergency warning pipeline network detection node as a necessary warning pipeline network detection node.

[0036] Preferably, the determination method of the pressure edge arrival time point is as follows:

[0037] If it shows a linear change, find the time point corresponding to the end value of the normal pressure range closest to the pressure value of the non-emergency warning pipeline network detection node on the pressure fitting line, that is, the pressure edge arrival time point;

[0038] If it shows a non-linear change, calculate the pressure change rate of the non-emergency warning pipeline network detection node at adjacent times, and extract the fastest pressure change rate;

[0039] Perform a proportional calculation on the pressure critical value and the fastest pressure change rate to obtain the remaining pressure critical duration, and combine it with the predicted signal generation time point to determine the pressure edge arrival time point.

[0040] A pressure detection and warning system based on pipeline network data, characterized in that it includes:

[0041] Detection omission analysis module: Compare the historical abnormal pipeline network detection nodes with the set pipeline network detection nodes in terms of location to identify the pipeline network nodes with omitted detection;

[0042] Necessary detection analysis module: Through the pressure anomaly interval time series of the pipeline network nodes with omitted detection in the historical operation period of the pipeline network, predict whether the pipeline network nodes with omitted detection will have pressure anomalies in the current pipeline network detection period. If so, mark the pipeline network nodes with omitted detection as necessary pipeline network detection nodes;

[0043] Detection node optimization module: Add new pipeline network detection nodes according to the positions of the necessary pipeline network detection nodes to optimize the setting of the pipeline network detection nodes;

[0044] Working condition pressure simulation module: After the optimization setting of the pipeline network detection nodes is completed, simulate the pressure distribution of the pipeline network under different working conditions through the pipeline network pressure model to determine the normal pressure range of each pipeline network detection node under different working conditions;

[0045] Pressure detection and warning module: Compare the pressure of each detected pipeline network detection node with the normal pressure range of each pipeline network detection node under the corresponding working condition to identify the emergency warning pipeline network detection nodes and the necessary warning pipeline network detection nodes.

[0046] The beneficial effects of the present invention are as follows:

[0047] 1. By comparing the positions of historical abnormal pipeline network detection nodes with the set pipeline network detection nodes, the missing pipeline network detection nodes are found. Through predicting the pressure abnormal time points of the missing pipeline network detection nodes, it is determined whether the missing pipeline network detection nodes will have pressure abnormalities during the current pipeline network detection cycle, thereby completing the optimization of the setting of pipeline network detection nodes, improving the rationality of the setting of pipeline network detection nodes, optimizing the cost of pipeline network pressure detection resources, and improving the accuracy of pipeline network pressure detection.

[0048] 2. Simulate the pressure distribution of the pipeline network under different working conditions to find the normal pressure range of each pipeline network detection node under different working conditions, solving the problem of inaccurate or misjudged pipeline network pressure detection caused by different normal pressure ranges of pipeline network detection nodes due to different working conditions. And through the comparison of pressures, identify emergency warning pipeline network detection nodes and necessary warning pipeline network detection nodes, not only identifying the pipeline network detection nodes that already have pressure abnormalities, but also identifying the pipeline network detection nodes that may have pressure abnormalities through predictive analysis of pressure changes and making necessity markings, improving the timeliness of pipeline network pressure detection and warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The present invention will be further described below with reference to the accompanying drawings.

[0050] Figure 1 is a flowchart of the steps of a pressure detection and warning method based on pipeline network data according to an embodiment of the present invention;

[0051] Figure 2 is a program block diagram of a pressure detection and warning system based on pipeline network data according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0052] In order to make the technical means, creative features, achieved purposes and effects of the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments.

[0053] Embodiment 1

[0054] Please refer to Figure 1 shown, a pressure detection and warning method based on pipeline network data according to an embodiment of the present invention includes the following steps:

[0055] Step 1: Compare the positions of historical abnormal pipeline network detection nodes with the set pipeline network detection nodes to identify the missing pipeline network detection nodes;

[0056] In Step 1, the set pipeline network detection nodes are set in advance by those skilled in the art. For example, the setting of pipeline network detection nodes includes, but is not limited to, the following methods:

[0057] Method 1: In a tree-shaped and unidirectional water supply pipe network structure, such as a mountainous tree-shaped water supply system, due to the large elevation difference in mountainous areas, the pressure at the end is prone to being low; the water flow distribution pressure change is reflected at the bifurcation, and the settings of the pipe network detection nodes include but are not limited to: the outlet of the water plant (point 1), the bifurcation of the main pipe (point 2), the entrance of the village with the highest elevation (point 1), etc.;

[0058] Method 2: In a ring-shaped and multi-directional water supply pipe network structure, such as an urban ring-shaped water supply system, due to the need to detect the pressure fluctuations at the intersection of the ring main lines, track and dispatch the response before and after the pumping station, and calibrate the water distribution deviation at the demarcation line, the settings of the pipe network detection nodes include but are not limited to: the intersection of the ring main lines (point 3), before and after the secondary pressurization pumping station (point 2), at the water supply demarcation line (point 2), etc.;

[0059] Method 3: In a drainage pipe network structure, such as a commercial area drainage pipe network system, due to the high oil content in the catering sewage, it is necessary to monitor the water quality; the flow rate at the inlet of the storage tank reflects the drainage efficiency, and the settings of the pipe network detection nodes include but are not limited to: the drainage outlets in the catering concentrated area (point 2), the intersection of the rainwater pipes (point 2), the inlet of the storage tank (point 3), etc.;

[0060] In step one, the historical abnormal pipe network detection nodes refer to the pipe network detection nodes where pressure anomalies occur during the historical operation cycle of the pipe network;

[0061] In step one, the process of identifying the missing pipe network detection nodes includes:

[0062] Compare the positions of the already set pipe network detection nodes with the historical abnormal pipe network detection nodes,

[0063] Mark the historical abnormal pipe network detection nodes that are different from the positions of all the already set pipe network detection nodes as the missing pipe network detection nodes;

[0064] Conversely, if the historical abnormal pipe network detection node has the same position as any one of the already set pipe network detection nodes, then mark the historical abnormal pipe network detection node as a non-missing pipe network detection node;

[0065] Step two: Predict whether the missing pipe network detection nodes will have pressure anomalies in the current pipe network detection cycle through the pressure anomaly interval time series of the missing pipe network detection nodes during the historical operation cycle of the pipe network. If so, mark the missing pipe network detection nodes as the necessary pipe network detection nodes;

[0066] In step two, the pressure anomaly interval time series of the missing pipe network detection nodes during the historical operation cycle of the pipe network is constructed by the interval time when the missing pipe network detection nodes have pressure anomalies during the historical operation cycle of the pipe network;

[0067] Exemplarily, assume that the time points when the pipeline network leakage detection node has abnormal pressure during the historical operation period of the pipeline network are (ysj 1 、ysj 2 、ysj 3 ......ysj n ), where ysjn represents the time point when the pipeline network leakage detection node has abnormal pressure for the nth time during the historical operation period of the pipeline network. Then, the constructed time series of abnormal pressure intervals is: (ysj 2 -ysj 1 、ysj 3 -ysj 2 、......ysj n -ysj n-1 )

[0068] In step two, the process of predicting whether the pipeline network leakage detection node will have abnormal pressure during the current pipeline network detection period includes:

[0069] Calculate the standard deviation and mean of the time series of abnormal pressure intervals of the pipeline network leakage detection node during the historical operation period of the pipeline network, and perform a ratio calculation on the standard deviation and the mean to obtain the stationary value of the time series of abnormal pressure intervals of the pipeline network leakage detection node during the historical operation period of the pipeline network;

[0070] It can be understood that by calculating the ratio of the standard deviation and the mean, and calculating the stationary value of the time series of abnormal pressure intervals of the pipeline network leakage detection node during the historical operation period of the pipeline network by obtaining the coefficient of variation, it is to judge the stationarity of the occurrence of abnormal pressure of the pipeline network leakage detection node during the historical operation period of the pipeline network. Its role is conducive to more accurately predicting whether the pipeline network leakage detection node will have abnormal pressure during the current pipeline network detection period according to the stationarity of the occurrence of abnormal pressure of the pipeline network leakage detection node during the historical operation period of the pipeline network;

[0071] In some preferred embodiments, compare the stationary value of the time series of abnormal pressure intervals of the pipeline network leakage detection node during the historical operation period of the pipeline network with the stationary threshold;

[0072] If the stationary value is greater than or equal to the stationary threshold, it means that the time series of abnormal pressure intervals of the pipeline network leakage detection node during the historical operation period of the pipeline network is non-stationary;

[0073] If the stationary value is less than the stationary threshold, it means that the time series of abnormal pressure intervals of the pipeline network leakage detection node during the historical operation period of the pipeline network is stationary;

[0074] If the pressure anomaly interval time series of the pipeline network leakage detection node is stationary during the historical operation period of the pipeline network, the historical average method is used to process the pressure anomaly interval time series, and the mean value of the pressure anomaly interval time series is used as the predicted value of the anomaly interval time;

[0075] If the pressure anomaly interval time series of the pipeline network leakage detection node is non-stationary during the historical operation period of the pipeline network, the LSTN network is used to process the pressure anomaly interval time series to determine the predicted value of the anomaly interval time;

[0076] Exemplarily, the process of using the LSTN network to process the pressure anomaly interval time series includes:

[0077] Assume that the pressure anomaly interval time series is: [2, 3, 4, 5, 6, 7, 8, 9, 10, 11] (unit: days);

[0078] Convert the pressure anomaly interval time series into the LSTM input format. For example, use the first 3 pressure anomaly intervals to predict the 4th pressure anomaly interval to form training samples:

[0079] Input sequence: [[2,3,4],[3,4,5],...,[8,9,10]];

[0080] Output label: [5,6,7,8,9,10,11];

[0081] Normalization: Scale the data to the range [0,1] (such as using max-min normalization);

[0082] Design a neural network containing an LSTM layer:

[0083] Input layer: 3 nodes (corresponding to the sliding window length of 3);

[0084] LSTM layer: 64 units;

[0085] Fully connected layer: 1 node (output the predicted pressure anomaly interval time);

[0086] Model training;

[0087] Divide the dataset: The first 70% of the samples are used as the training set, and the last 30% are used as the validation set;

[0088] Training parameters: Set the number of training epochs (epochs = 50) and the batch size (batch_size = 32);

[0089] Predict the next interval;

[0090] Input the last 3 intervals [9,10,11] into the trained LSTM model;

[0091] Output predicted value: The pressure anomaly interval time after the 11th interval predicted by the model output is obtained to get the predicted value of the anomaly interval time;

[0092] Obtain the last pressure anomaly time point of the pipeline network omission detection node before the current pipeline network detection cycle, and sum it with the predicted value of the anomaly interval time to obtain the subsequent pressure anomaly prediction time point of the pipeline network omission detection node;

[0093] If the pressure anomaly prediction time point is within the current pipeline network detection cycle, it means that the pipeline network omission detection node will have a pressure anomaly within the current detection cycle, and mark the pipeline network omission detection node as a necessary pipeline network detection node;

[0094] If the pressure anomaly prediction time point is not within the current pipeline network detection cycle, it means that the pipeline network omission detection node will not have a pressure anomaly within the current detection cycle, and mark the pipeline network omission detection node as a non-necessary pipeline network detection node;

[0095] It can be understood that by predicting whether the pipeline network omission detection node will have a pressure anomaly within the current detection cycle and identifying the necessary pipeline network detection nodes according to the prediction results, it is beneficial to optimize the addition of the set pipeline network detection nodes based on the necessary pipeline network detection nodes, improve the reasonable accuracy of the pipeline network detection node settings, and through the necessity analysis of the pipeline network omission detection nodes, it is beneficial to improve the reasonable accuracy of the pipeline network detection node settings while optimizing the resource cost of the pipeline network detection node settings;

[0096] Step 3: Set and add new pipeline network detection nodes according to the positions of the necessary pipeline network detection nodes to optimize the pipeline network detection node settings;

[0097] Specifically, set new pipeline network detection nodes at the positions corresponding to the necessary pipeline network detection nodes;

[0098] The technical solution and beneficial points of the embodiments of this application are as follows: Compare the positions of historical abnormal pipeline network detection nodes with the already set pipeline network detection nodes to identify missing pipeline network detection nodes. Based on the pressure anomaly interval time series of the missing pipeline network detection nodes during the historical operation period of the pipeline network, predict whether the missing pipeline network detection nodes will have pressure anomalies during the current pipeline network detection period. If so, mark the missing pipeline network detection nodes as necessary pipeline network detection nodes, add new pipeline network detection nodes according to the settings of the necessary pipeline network detection nodes, and optimize the settings of the pipeline network detection nodes. This application analyzes and compares the positions of historical abnormal pipeline network detection nodes with the already set pipeline network detection nodes to find the missing pipeline network detection nodes, and predicts whether the missing pipeline network detection nodes will have pressure anomalies during the current pipeline network detection period through the pressure anomaly time points of the missing pipeline network detection nodes, thereby completing the optimization of the settings of the pipeline network detection nodes, improving the rationality of the settings of the pipeline network detection nodes, optimizing the cost of pipeline network pressure detection resources, and improving the accuracy of pipeline network pressure detection.

[0099] Embodiment 2

[0100] Please refer to Figure 1 As shown, based on the technical solution of the optimized setting of the pipeline network detection nodes described in Embodiment 1 above, the embodiments of the present invention perform simulations of the operation of the pipeline network under different working conditions after the optimized setting of the pipeline network detection nodes, and compare them with the actual operating conditions of the pipeline network during the current pipeline network detection period, so as to more accurately realize the early warning of pipeline network pressure detection. Therefore, a pressure detection and early warning method based on pipeline network data described in the embodiments of the present invention includes the following steps:

[0101] Step Four: After the optimized setting of the pipeline network detection nodes is completed, simulate the pressure distribution of the pipeline network under different working conditions through the pipeline network pressure model, and determine the normal pressure range of each pipeline network detection node under different working conditions;

[0102] In Step Four, the construction method of the pipeline network pressure model is as follows:

[0103] A1. Collect pipeline network information: Obtain the topological structure of the pipeline network, clarify the connection relationships of each node and pipe section. For example, for a looped pipeline network, it includes multiple water source nodes, several intermediate nodes, and a large number of end nodes, and record information such as the pipe diameter and pipe material of each pipe section. For example, the pipe diameter of a certain pipe section is 400 mm, and the pipe material is ductile iron pipe;

[0104] A2. Analyze the historical water consumption data to determine the water usage patterns in different time periods. For example, from 7 to 9 am and from 17 to 19 pm on weekdays are peak water usage periods, the water consumption on weekends is relatively scattered but overall higher than that during non-peak periods on weekdays, and the water consumption in summer is greater than that in winter due to high temperatures, etc.;

[0105] A3. Collect the pressure data of each detection node in the collection pipeline network at different times. These data include the pressure values in different seasons, different time periods, as well as normal and abnormal conditions.

[0106] A4. Select a mathematical model: Based on the principle of fluid mechanics, use the Darcy - Weisbach formula combined with the node flow balance equation to establish a pipeline network pressure model.

[0107] In step four, the process of simulating the pressure distribution of the pipeline network under different working conditions and determining the normal pressure range of each pipeline network detection node under different working conditions includes:

[0108] Set multiple working conditions. For example, including but not limited to the peak working condition on a summer weekday, the low - valley working condition on a winter weekend, etc. Input the pre - set working condition parameters into the pipeline network pressure model and calculate the pressure distribution of each detection node in the pipeline network under this working condition.

[0109] According to the simulation results, obtain the maximum pressure and minimum pressure of each pipeline network detection node under the normal operation of different working conditions, and construct the normal pressure range of each pipeline network detection node under different working conditions based on the maximum pressure and minimum pressure.

[0110] It should be noted that each pipeline network detection node corresponds to a normal pressure range under each working condition.

[0111] Step five: Compare the pressure of each pipeline network detection node detected with the normal pressure range of each pipeline network detection node under the corresponding working condition, and identify the emergency warning pipeline network detection nodes and the necessary warning pipeline network detection nodes.

[0112] In step five, it should be noted that the meaning of the corresponding working condition is that the working condition of the pipeline network detection node in the current pipeline network detection period is consistent with the working condition of the pipeline network detection node corresponding to the compared normal pressure range.

[0113] In step five, the process of identifying the emergency warning pipeline network detection nodes and the necessary warning pipeline network detection nodes includes:

[0114] Compare the pressure of the pipeline network detection node with the normal pressure range of the pipeline network detection node under the corresponding working condition.

[0115] If the pressure of the pipeline network detection node is within the normal pressure range, mark the pipeline network detection node as a non - emergency warning pipeline network detection node.

[0116] If the pressure of the pipeline network detection node is not within the normal pressure range, mark the pipeline network detection node as an emergency warning pipeline network detection node.

[0117] It should be noted that the emergency warning pipeline detection node indicates that the pipeline detection node has deviated from the normal pressure range, and it is necessary to issue an emergency warning for this pipeline detection node to prevent pipeline accidents from occurring;

[0118] Based on the non-emergency warning pipeline detection node, compare the pressure of the non-emergency warning pipeline detection node with the pressure critical range within the normal pressure range;

[0119] If the pressure of the non-emergency warning pipeline detection node is within the pressure critical range, no treatment is performed;

[0120] If the pressure of the non-emergency warning pipeline detection node is not within the pressure critical range, a proximity analysis signal is generated;

[0121] It should be noted that the pressure critical range is set by those skilled in the art. The maximum value within the pressure critical range is less than the maximum value within the normal pressure range, and the minimum value within the pressure critical range is greater than the minimum value within the normal pressure range. Exemplarily, for a certain end pipeline detection node, under the peak working conditions of multiple summer working days in the past, the normal pressure range is between 0.22 - 0.28 MPa, then the set pressure critical range may be 0.24 - 0.26 MPa;

[0122] Based on the proximity analysis signal, if the pressure value of the non-emergency warning pipeline detection node exceeds the maximum value within the pressure critical range, take the absolute value of the difference between the pressure value of the non-emergency warning pipeline detection node and the maximum value within the normal pressure range to obtain the pressure critical value of the non-emergency warning pipeline detection node;

[0123] Exemplarily, assume that the pressure value of the non-emergency warning pipeline detection node is 0.27 MPa, which exceeds the maximum value of 0.26 MPa within the pressure critical range. Then take the absolute value of the difference between 0.27 MPa and 0.28 MPa, and the obtained pressure critical value is 0.01 MPa;

[0124] If the pressure value of the non-emergency warning pipeline detection node is lower than the minimum value within the pressure critical range, take the absolute value of the difference between the pressure value of the non-emergency warning pipeline detection node and the minimum value within the normal pressure range to obtain the pressure critical value of the non-emergency warning pipeline detection node;

[0125] Exemplarily, assume that the pressure value of the non-emergency warning pipeline detection node is 0.235 MPa, which is lower than the minimum value of 0.24 MPa within the pressure critical range. Then take the absolute value of the difference between 0.235 MPa and 0.22 MPa, and the obtained pressure critical value is 0.015 MPa;

[0126] Compare the pressure critical value with the pressure critical threshold;

[0127] If the pressure critical value is less than or equal to the pressure critical threshold, a prediction signal is generated;

[0128] If the pressure critical value is greater than the pressure critical threshold, no processing is performed;

[0129] It should be noted that the meaning of generating the prediction signal is as follows: The prediction signal is generated because the pressure critical value is less than or equal to the pressure critical threshold. Since the pressure critical value is less than or equal to the pressure critical threshold, it means that the pressure value of the non - emergency warning pipeline network detection node has approached the edge of the normal pressure range endpoint of the pipeline network detection node. In order to prevent the pressure value of the non - emergency warning pipeline network detection node from exceeding the normal pressure range, the pressure prediction of the non - emergency warning pipeline network detection node is carried out in advance, which is conducive to the early warning of the pipeline network pressure detection and improves the timeliness of the warning;

[0130] Based on the prediction signal, obtain the time point when the pressure value of the non - emergency warning pipeline network detection node deviates from the pressure critical range, and get the deviation time point;

[0131] It should be noted that the pressure value of the non - emergency warning pipeline network detection node deviating from the pressure critical range means that the pressure value of the non - emergency warning pipeline network detection node exceeds the maximum value within the pressure critical range or the pressure value of the non - emergency warning pipeline network detection node is lower than the minimum value within the pressure critical range;

[0132] Obtain the pressure values of the non - emergency warning pipeline network detection node at different times during the period from the deviation time point to the prediction signal generation time point, and integrate them into a pressure data set;

[0133] Use the least - squares method to fit the pressure values of the non - emergency warning pipeline network detection node at different times in the pressure data set, obtain the pressure fitting line, and calculate the determination coefficient R 2 ;

[0134] Exemplarily, the calculation method of the determination coefficient R 2 can be:

[0135] Using the least - squares method to fit the pressure values of the non - emergency warning pipeline network detection node at different times in the pressure data set, a pressure fitting line and a pressure fitting line model (y = kx + b) can be obtained, where y represents pressure, x represents time, and b represents the intercept; Taking different times as independent variables and inputting them, the pressure prediction values at different times are obtained; Calculate three sums of squares according to the pressure values of the non - emergency warning pipeline network detection node at different times and the obtained pressure prediction values at different times, including the total sum of squared deviations (SST), the regression sum of squares (SSR), and the residual sum of squares (SSE), and calculate the determination coefficient through the three sums of squares. The specific calculation formula is as follows:

[0136]

[0137] Compare the coefficient of determination R of the pressure fitting line 2 with the coefficient of determination threshold;

[0138] If the coefficient of determination R of the pressure fitting line 2 is greater than or equal to the coefficient of determination threshold, it indicates that the fitting effect of the pressure fitting line is good, and the pressure values of the non-emergency warning pipeline network detection nodes change linearly;

[0139] If the coefficient of determination R of the pressure fitting line 2 is less than the coefficient of determination threshold, it indicates that the fitting effect of the pressure fitting line is good, and the pressure values of the non-emergency warning pipeline network detection nodes change non-linearly;

[0140] Based on the linear change of the pressure values of the non-emergency warning pipeline network detection nodes, find the time point corresponding to the end point value of the normal pressure range closest to the pressure value of the non-emergency warning pipeline network detection node on the pressure fitting line, that is, the pressure edge arrival time point;

[0141] If the pressure edge arrival time point is within the current pipeline network detection period, it indicates that the pressure value of the non-emergency warning pipeline network detection node will reach the edge of the normal pressure range end point within the current pipeline network detection period, and early warning is required. Then mark the non-emergency warning pipeline network detection node as a necessary warning pipeline network detection node;

[0142] If the pressure edge arrival time point is within the current pipeline network detection period, it indicates that the pressure value of the non-emergency warning pipeline network detection node will not reach the edge of the normal pressure range end point within the current pipeline network detection period, and no operation is performed;

[0143] Based on the non-linear change of the pressure values of the non-emergency warning pipeline network detection nodes, calculate the pressure change rate of the non-emergency warning pipeline network detection nodes at adjacent moments in the pressure data set, and extract the fastest pressure change rate, where the fastest pressure change rate includes the fastest pressure increase rate and the fastest pressure decrease rate; among them, the pressure change rate is calculated by the ratio of the pressure difference between the non-emergency warning pipeline network detection nodes at adjacent moments to the time interval between adjacent moments;

[0144] Perform a proportional calculation on the pressure critical value and the fastest pressure change rate to obtain the remaining pressure critical duration, and combine it with the prediction signal generation time point, and sum to obtain the pressure edge arrival time point, that is: pressure edge arrival time point = prediction signal generation time point + remaining pressure critical duration;

[0145] It should be noted that when the predicted signal is generated, if the pressure value of the non-emergency warning pipeline detection node is lower than the minimum value within the pressure critical range, the pressure critical value is calculated proportionally with the fastest pressure reduction rate; when the predicted signal is generated, if the pressure value of the non-emergency warning pipeline detection node exceeds the maximum value within the pressure critical range, the pressure critical value is calculated proportionally with the fastest pressure increase rate;

[0146] If the pressure edge arrival time point is within the current pipeline network detection cycle, it means that the pressure value of the non-emergency warning pipeline detection node will reach the edge of the normal pressure range endpoint within the current pipeline network detection cycle, and early warning is required. Then, the non-emergency warning pipeline detection node is marked as a necessary warning pipeline detection node;

[0147] If the pressure edge arrival time point is within the current pipeline network detection cycle, it means that the pressure value of the non-emergency warning pipeline detection node will not reach the edge of the normal pressure range endpoint within the current pipeline network detection cycle, and no operation is performed;

[0148] The technical solution and beneficial points of the embodiments of the present application are as follows: After the optimal setting of the pipeline network detection nodes is completed, the pressure distribution of the pipeline network under different working conditions is simulated through the pipeline network pressure model, and the normal pressure range of each pipeline network detection node under different working conditions is determined. Within the current pipeline network detection cycle, the pressure of each pipeline network detection node is detected and compared with the normal pressure range of each pipeline network detection node under the corresponding working conditions to identify the emergency warning pipeline detection nodes and the necessary warning pipeline detection nodes. The present invention simulates the pressure distribution of the pipeline network under different working conditions to find the normal pressure range of each pipeline network detection node under different working conditions, solves the problem of inaccurate pipeline network pressure detection or misjudgment caused by different normal pressure ranges of pipeline network detection nodes due to different working conditions, and through the comparison of pressures, identifies the emergency warning pipeline detection nodes and the necessary warning pipeline detection nodes, not only identifying the pipeline network detection nodes with abnormal pressure, but also through the predictive analysis of pressure changes, identifying the pipeline network detection nodes that may have abnormal pressure and making necessary markings, improving the timeliness of pipeline network pressure detection and warning.

[0149] Embodiment 3

[0150] Please refer to Figure 2 As shown, based on the pressure detection and warning method based on pipeline network data described in the above Embodiment 1 and Embodiment 2, the present invention further provides a pressure detection and warning system based on pipeline network data, including the following modules:

[0151] Detection omission analysis module: Compare the historical abnormal pipeline network detection nodes with the set pipeline network detection nodes in terms of position to identify the pipeline network detection nodes with omitted detection;

[0152] The process of identifying the pipeline network detection nodes with omitted detection includes:

[0153] Compare the positions of the set pipeline network detection nodes with those of the historical abnormal pipeline network detection nodes.

[0154] Mark the historical abnormal pipeline network detection nodes that are different from all the positions of the set pipeline network detection nodes as pipeline network omission detection nodes.

[0155] Conversely, if the position of a historical abnormal pipeline network detection node is the same as that of any one of the set pipeline network detection nodes, mark the historical abnormal pipeline network detection node as a non-pipeline network omission detection node.

[0156] Necessary detection and analysis module: Predict whether a pipeline network omission detection node will have a pressure anomaly in the current pipeline network detection period through the pressure anomaly interval time series of the pipeline network omission detection node in the historical operation period of the pipeline network. If so, mark the pipeline network omission detection node as a pipeline network necessary detection node.

[0157] The pressure anomaly interval time series of the pipeline network omission detection node in the historical operation period of the pipeline network is constructed by the interval time when the pipeline network omission detection node has a pressure anomaly in the historical operation period of the pipeline network.

[0158] The process of predicting whether a pipeline network omission detection node will have a pressure anomaly in the current pipeline network detection period includes:

[0159] Calculate the standard deviation and mean of the pressure anomaly interval time series of the pipeline network omission detection node in the historical operation period of the pipeline network, and calculate the ratio of the standard deviation and the mean to obtain the stationary value of the pressure anomaly interval time series of the pipeline network omission detection node in the historical operation period of the pipeline network.

[0160] If the stationary value is greater than or equal to the stationary threshold, it means that the pressure anomaly interval time series of the pipeline network omission detection node in the historical operation period of the pipeline network is non-stationary.

[0161] If the stationary value is less than the stationary threshold, it means that the pressure anomaly interval time series of the pipeline network omission detection node in the historical operation period of the pipeline network is stationary.

[0162] If the pressure anomaly interval time series of the pipeline network omission detection node in the historical operation period of the pipeline network is stationary, use the historical average method to perform mean processing on the pressure anomaly interval time series, and take the mean of the pressure anomaly interval time series as the anomaly interval time prediction value.

[0163] If the pressure anomaly interval time series of the pipeline network omission detection node in the historical operation period of the pipeline network is non-stationary, use the LSTN network to process the pressure anomaly interval time series to determine the anomaly interval time prediction value.

[0164] Obtain the last pressure anomaly time point of the pipeline network leakage detection node before the current pipeline network detection period, and sum it with the predicted value of the anomaly interval time to obtain the subsequent predicted pressure anomaly time point of the pipeline network leakage detection node;

[0165] If the predicted pressure anomaly time point is within the current pipeline network detection period, it means that the pipeline network leakage detection node will have a pressure anomaly within the current detection period, and mark the pipeline network leakage detection node as a necessary pipeline network detection node;

[0166] If the predicted pressure anomaly time point is not within the current pipeline network detection period, it means that the pipeline network leakage detection node will not have a pressure anomaly within the current detection period, and mark the pipeline network leakage detection node as a non-necessary pipeline network detection node;

[0167] Detection node optimization module: Add new pipeline network detection nodes according to the positions of the necessary pipeline network detection nodes to optimize the settings of the pipeline network detection nodes;

[0168] Set new pipeline network detection nodes at the positions corresponding to the necessary pipeline network detection nodes;

[0169] Operating condition pressure simulation module: After the optimization of the pipeline network detection node settings is completed, simulate the pressure distribution of the pipeline network under different operating conditions through the pipeline network pressure model to determine the normal pressure range of each pipeline network detection node under different operating conditions;

[0170] According to the simulation results, obtain the maximum pressure and minimum pressure of each pipeline network detection node under the normal operation of the pipeline network under different operating conditions, and construct the normal pressure range of each pipeline network detection node under different operating conditions based on the maximum pressure and minimum pressure;

[0171] Pressure detection and early warning module: Compare the pressure of each detected pipeline network detection node with the normal pressure range of each pipeline network detection node under the corresponding operating conditions to identify emergency early warning pipeline network detection nodes and necessary early warning pipeline network detection nodes.

[0172] The process of identifying emergency early warning pipeline network detection nodes and necessary early warning pipeline network detection nodes includes:

[0173] If the pressure of the pipeline network detection node is within the normal pressure range, mark the pipeline network detection node as a non-emergency early warning pipeline network detection node;

[0174] If the pressure of the pipeline network detection node is not within the normal pressure range, mark the pipeline network detection node as an emergency early warning pipeline network detection node;

[0175] Based on non-emergency early warning pipeline network detection nodes;

[0176] If the pressure of the non-emergency early warning pipeline network detection node is within the pressure critical range, no processing is performed;

[0177] If the pressure of the non - emergency warning pipeline network detection node is not within the pressure critical range, a proximity analysis signal is generated;

[0178] Based on the proximity analysis signal, if the pressure value of the non - emergency warning pipeline network detection node exceeds the maximum value within the pressure critical range, the absolute value of the difference between the pressure value of the non - emergency warning pipeline network detection node and the maximum value within the normal pressure range is obtained to get the pressure critical value of the non - emergency warning pipeline network detection node;

[0179] If the pressure value of the non - emergency warning pipeline network detection node is lower than the minimum value within the pressure critical range, the absolute value of the difference between the pressure value of the non - emergency warning pipeline network detection node and the minimum value within the normal pressure range is obtained to get the pressure critical value of the non - emergency warning pipeline network detection node;

[0180] If the pressure critical value is less than or equal to the pressure critical threshold, a prediction signal is generated;

[0181] Based on the prediction signal, the time point when the pressure value of the non - emergency warning pipeline network detection node deviates from the pressure critical range is obtained to get the deviation time point;

[0182] The pressure values of the non - emergency warning pipeline network detection node at different moments within the period from the deviation time point to the prediction signal generation time point are obtained and integrated into a pressure data set;

[0183] Using the least - squares method to fit the pressure values of the non - emergency warning pipeline network detection node at different moments in the pressure data set to obtain a pressure fitting straight line, and calculate the determination coefficient R of the pressure fitting straight line 2 ;

[0184] If the determination coefficient R of the pressure fitting straight line 2 is greater than or equal to the determination coefficient threshold, it indicates that the fitting effect of the pressure fitting straight line is good and the pressure value of the non - emergency warning pipeline network detection node changes linearly;

[0185] If the determination coefficient R of the pressure fitting straight line 2 is less than the determination coefficient threshold, it indicates that the fitting effect of the pressure fitting straight line is good and the pressure value of the non - emergency warning pipeline network detection node changes non - linearly;

[0186] Based on the linear change of the pressure value of the non - emergency warning pipeline network detection node, the time point corresponding to the normal pressure range endpoint value closest to the pressure value of the non - emergency warning pipeline network detection node on the pressure fitting straight line is found, that is, the pressure edge arrival time point;

[0187] If the pressure edge arrival time point is within the current pipe network detection period, it means that the pressure value of the non-emergency warning pipe network detection node will reach the edge of the normal pressure range within the current pipe network detection period, and early warning is required. Then, the non-emergency warning pipe network detection node is marked as a necessary warning pipe network detection node;

[0188] Based on the fact that the pressure value of the non-emergency warning pipe network detection node changes non-linearly, calculate the pressure change rate of the non-emergency warning pipe network detection node at adjacent moments in the pressure data set, and extract the fastest pressure change rate, where the fastest pressure change rate includes the fastest pressure increase rate and the fastest pressure decrease rate; among them, the pressure change rate is obtained by proportionally calculating the pressure difference between the non-emergency warning pipe network detection nodes at adjacent moments and the time interval between adjacent moments;

[0189] Perform a proportional calculation on the pressure critical value and the fastest pressure change rate to obtain the remaining pressure critical duration, and combine it with the prediction signal generation time point, and sum to obtain the pressure edge arrival time point, that is: pressure edge arrival time point = prediction signal generation time point + remaining pressure critical duration;

[0190] If the pressure edge arrival time point is within the current pipe network detection period, it means that the pressure value of the non-emergency warning pipe network detection node will reach the edge of the normal pressure range within the current pipe network detection period, and early warning is required. Then, the non-emergency warning pipe network detection node is marked as a necessary warning pipe network detection node.

[0191] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art of this industry should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A pressure detection and early warning method based on pipe network data, characterized in that: include: Compare the locations of historical abnormal pipe network detection nodes with the set pipe network detection nodes to identify missed detection nodes in the pipe network; Through the time series of pressure anomaly intervals of the missed detection nodes in the historical operation cycle of the pipeline network, it is predicted whether the missed detection nodes in the pipeline network will have pressure anomalies in the current pipeline network detection cycle. If so, the missed detection nodes in the pipeline network are marked as necessary detection nodes in the pipeline network. Add new pipe network detection nodes according to the locations of necessary detection nodes in the pipe network and optimize the settings of pipe network detection nodes; After the optimization setting of the pipeline network detection nodes is completed, the pressure distribution of the pipeline network under different working conditions is simulated through the pipeline network pressure model to determine the normal pressure range of each pipeline network detection node under different working conditions; Compare and predict the pressure of each pipeline network detection node detected with the normal pressure range of each pipeline network detection node under the corresponding working conditions, and identify the emergency warning pipeline network detection nodes and necessary warning pipeline network detection nodes; The identification method of the emergency warning pipe network detection node and the necessary warning pipe network detection node is: If the pressure of the pipeline network detection node is not within the normal pressure range, the pipeline network detection node is marked as an emergency warning pipeline network detection node, otherwise, the pipeline network detection node is marked as a non-emergency warning pipeline network detection node; Perform critical pressure analysis on non-emergency warning pipe network detection nodes to determine whether pressure prediction is needed. If necessary, obtain the departure time point through the time point when the pressure value of the non-emergency warning pipe network detection node leaves the critical pressure range; Fit the pressure values ​​of the non-emergency warning pipe network detection nodes in the period between the separation time point and the prediction signal generation time point to obtain the pressure fitting straight line and calculate the determination coefficient; The determination coefficient is used to determine whether the pressure value of the non-emergency warning pipe network detection node changes linearly or nonlinearly and to determine the time point when the pressure edge arrives; If the pressure edge arrival time point is within the current pipe network detection cycle, the non-emergency warning pipe network detection node will be marked as a necessary warning pipe network detection node.

2. A pressure detection and early warning method based on pipe network data according to claim 1, characterized in that: The method for identifying the missing detection nodes of the pipe network is as follows: The historical abnormal pipeline network detection nodes that are different from the positions of all the set pipeline network detection nodes are marked as pipeline network missed detection nodes.

3. The pressure detection and early warning method based on pipe network data according to claim 1 is characterized in that: The process of predicting whether a pipe network missed detection node will have pressure anomaly within the current pipe network detection cycle is as follows: Perform stationarity analysis on the pressure anomaly interval time series to determine the predicted value of the anomaly interval time, and determine the pressure anomaly prediction time point in combination with the last pressure anomaly time point of the missed detection node in the pipeline network; If the pressure anomaly prediction time point is within the current pipe network detection cycle, it means that the missed detection node of the pipe network will cause pressure anomaly within the current pipe network detection cycle.

4. A pressure detection and early warning method based on pipe network data according to claim 3, characterized in that: The process of performing stationarity analysis on the pressure anomaly interval time series is as follows: The standard deviation and mean of the pressure anomaly interval time series are calculated, and the standard deviation and mean are proportionally calculated to obtain the stable value of the pressure anomaly interval time series; If the stability value is greater than or equal to the stability threshold, it indicates instability; then the LSTM network is used to process the pressure abnormality interval time series to determine the abnormality interval time prediction value; If the stable value is less than the stable threshold, it means stability, and the historical average method is used to process the pressure anomaly interval time series to determine the predicted value of the anomaly interval time.

5. The pressure detection and early warning method based on pipe network data according to claim 1 is characterized in that: The normal pressure range of each pipe network detection node under different working conditions is determined as follows: By simulating the pressure distribution of the pipeline network under different working conditions, the maximum and minimum pressures of each pipeline network detection node under normal operation of the pipeline network under different working conditions are obtained, and the normal pressure range of each pipeline network detection node under different working conditions is constructed.

6. The pressure detection and early warning method based on pipe network data according to claim 1 is characterized by: The pressure critical analysis process is: If the pressure of the non-emergency warning pipe network detection node is not within the critical pressure range, a proximity analysis signal is generated; Based on the proximity analysis signal, if the pressure value of the non-emergency warning pipeline network detection node exceeds the maximum value in the critical pressure range, the pressure value of the non-emergency warning pipeline network detection node is processed with the maximum value in the normal pressure range to obtain the critical pressure value of the non-emergency warning pipeline network detection node; If the pressure value of the non-emergency warning pipe network detection node is lower than the minimum value in the critical pressure range, the pressure value of the non-emergency warning pipe network detection node is processed with the minimum value in the normal pressure range to obtain the critical pressure value of the non-emergency warning pipe network detection node; If the pressure critical value is less than or equal to the pressure critical threshold, a prediction signal is generated.

7. The pressure detection and early warning method based on pipe network data according to claim 1 is characterized by: The method for determining the pressure edge arrival time point is: If it changes linearly, find the time point corresponding to the endpoint value of the normal pressure range closest to the pressure value of the non-emergency warning pipeline network detection node on the pressure fitting line, that is, the pressure edge arrival time point; If it changes nonlinearly, the pressure change rate of the non-emergency warning pipeline network detection nodes at adjacent moments is calculated, and the fastest pressure change rate is extracted; The critical pressure value and the fastest rate of pressure change are proportionally calculated to obtain the critical remaining pressure duration, and the pressure edge arrival time is determined in combination with the predicted signal generation time point.

8. A pressure detection and early warning system based on pipe network data, characterized in that: The system is used to execute the detection and early warning method described in any one of claims 1 to 7, including: Missed detection analysis module: compare the positions of historical abnormal pipe network detection nodes with the set pipe network detection nodes to identify missed detection nodes of the pipe network; Necessary detection analysis module: through the pressure anomaly interval time series of the pipeline network missed detection node in the historical operation cycle of the pipeline network, it is predicted whether the pipeline network missed detection node will have pressure anomaly in the current pipeline network detection cycle. If so, the pipeline network missed detection node is marked as a necessary detection node of the pipeline network; Detection node optimization module: add new pipeline network detection nodes according to the necessary detection node positions of the pipeline network, and optimize the settings of pipeline network detection nodes; Working condition pressure simulation module: After the optimization setting of the pipeline network detection nodes is completed, the pressure distribution of the pipeline network under different working conditions is simulated through the pipeline network pressure model to determine the normal pressure range of each pipeline network detection node under different working conditions; Pressure detection and early warning module: compare and predict the pressure of each pipeline network detection node detected with the normal pressure range of each pipeline network detection node under the corresponding working conditions, and identify emergency early warning pipeline network detection nodes and necessary early warning pipeline network detection nodes.

Images