A method for monitoring fluid pipeline leakage based on shortest path planning

Abstract

The invention discloses a method for monitoring leakage of a fluid pipe network based on the shortest path planning, comprising the following steps: establishing a propagation model of a negative pressure wave in a single straight pipe; mapping key points of the pipe, obtaining the longitude and latitude coordinates of several points, making a pipe network distribution data table, establishing an adjacency matrix according to the connection relationship between different points, and making a pipe network diagram in MATLAB; taking points at fixed intervals in all pipe sections of the pipe network, and subdividing the entire pipe network into several discrete points; solving the shortest path from each discrete point to each deployed sensor, and then calculating in turn the arrival time of the negative pressure wave signal detected by each deployed pressure sensor when a leak occurs at each discrete point. The arrival time is formed into a data column, and a time delay standard library is established therefrom. When a leak actually occurs at a certain position, the arrival time of the negative pressure wave measured by each sensor is compared with the time delay standard library, and the leak point coordinates are screened.

Description

Technical Field

[0001] The present invention relates to the technical field of pipeline detection, and in particular to a fluid pipeline network leakage monitoring method based on shortest path planning. Background Art

[0002] With the rapid development of pipeline construction, pipeline transportation plays an increasingly important role in our lives and industrial production. Due to factors such as pipeline wear and human damage, pipeline leakage incidents occur from time to time, which causes huge economic losses, waste of resources, and even causes casualties, resulting in adverse effects. Therefore, it is of great significance to monitor fluid transportation pipelines in real time, give timely alarms for pipeline leakage, and locate the leakage point.

[0003] At present, the most commonly used method for pipeline leakage detection in my country is the negative pressure wave detection method. Most of them deploy pressure sensors at both ends of a pipeline. When the pipeline leaks, there will be an obvious pressure drop at these two ends. At this time, the leakage point can be located based on the time delay of the pressure drop.

[0004] Transportation pipelines are often complex in structure, and traditional negative pressure wave leak detection methods are not applicable to pipelines. Taking the urban water supply network as an example, in order to meet the water needs of multiple users, the pipeline needs to contain multiple branches. In areas with dense populations, the number of pipeline branches will also increase accordingly. In this case, if the traditional negative pressure wave detection method is still used for pipeline leak detection, pressure sensors need to be deployed at both ends of each branch, which is costly and not applicable. If the deployed sensor and the leak point are not in the same pipeline, there will be many pipeline combinations from the leak point to the sensor, forming multiple propagation paths for the leakage negative pressure wave; when calculating the propagation time of the negative pressure wave generated by the leak, it is necessary to find the shortest path from the sensor to the leak point.

[0005] In view of this, it is necessary to design a multi-path pipe network leakage detection algorithm based on shortest path planning to solve the above problems. Summary of the invention

[0006] The present invention provides a fluid pipeline leakage monitoring method based on the shortest path planning, which can realize the leakage location of complex urban pipelines and has a wider range of applicability, as described below:

[0007] A method for monitoring leakage of a fluid pipe network based on shortest path planning, the method comprising the following steps:

[0008] Establish a negative pressure wave propagation model in a single straight pipeline; map key points of the pipeline, obtain the longitude and latitude coordinates of several points, make a pipeline network distribution data table, establish an adjacency matrix based on the connection relationship between different points, and make a pipeline network diagram in MATLAB;

[0009] Points are taken at fixed intervals in all pipe sections of the pipe network to subdivide the entire pipe network into a number of discrete points;

[0010] The shortest path from each discrete point to each deployed sensor is solved, and then the arrival time of the negative pressure wave signal detected by each deployed pressure sensor when a leak occurs at each discrete point is calculated in turn.

[0011] The arrival times are formed into a data column, and a time delay standard library is established based on this. When a leak actually occurs at a certain location, the arrival time of the negative pressure wave measured by each sensor is compared with the time delay standard library, and the coordinates of the leak point are screened.

[0012] The pipeline network distribution data table is prepared, and an adjacency matrix is ​​established according to the connection relationship between different points. The pipeline network diagram is made in MATLAB as follows:

[0013] Suppose there are n distinct points, numbered 1, 2, 3, ... n, then the adjacency matrix C has n rows and n columns. Let the element in the i-th row and j-th column of the adjacency matrix be c ij , then:

[0014]

[0015] The adjacency matrix C is established by searching the matrix N row by row and column by column. When searching in each row, starting from the second pair of coordinate points, the adjacency matrix elements corresponding to each coordinate point and its previous point are set to 1.

[0016] Define the coordinate matrix Coordinate. Every time a longitude and latitude coordinate that does not appear is searched, add it to the matrix Coordinate. The number of its rows in the matrix is ​​the number of the point.

[0017] In the process of establishing the adjacency matrix C, the repeated points are merged. Let the point searched in a certain cycle be P1, and the point searched before point P1 be P0. The possible existence of points P0 and P1 is analyzed:

[0018] (1) Points P0 and P1 are the coordinate points that appear for the first time. Let the point numbers of points P0 and P1 be a and b, then c ab =1;

[0019] (2) Point P0 is not the first coordinate point to appear. The point with the same coordinate value is numbered k. However, point P1 is the first coordinate point to appear. The point number of point P1 is b. Then c kb =1;

[0020] (3) Point P1 is not the first coordinate point to appear. The point with the same coordinate value is numbered t. However, point P0 is the first coordinate point to appear. The point number of point P0 is a. Then c at =1;

[0021] (4) Point P0 is not the first coordinate point to appear, and the point with the same coordinate value is numbered k. Point P1 is not the first coordinate point to appear, and the point with the same coordinate value is numbered t. Then c kt =1.

[0022] The process of forming a data column from the arrival time and establishing a delay standard library is as follows:

[0023] The line connecting two adjacent points in the shortest path is regarded as a straight pipe section. The propagation time is obtained by taking the pipe length and the wave speed as the quotient. Multiple propagation times are accumulated and the arrival time of the negative pressure wave signal detected by each pressure sensor is calculated. All arrival times form a data column, and a time delay standard library is established based on this.

[0024] Furthermore, two adjacent path points in the shortest path are A and B, and the situations of A and B are analyzed:

[0025] (1) A and B are not multi-pass locations, and the inner diameter of the pipe where A or B is located is directly used as D for calculation;

[0026] (2) If A or B is a multi-pass position, but B or A corresponds to only one pipe inner diameter, the pipe inner diameter corresponding to B or A shall be used as D for calculation;

[0027] (3) A and B are both multi-pass locations, and they have the same set of pipe diameters. This value is used as D in the calculation;

[0028] (4) If both A and B are multi-pass locations and have two or more sets of identical pipe diameters, search for the pipelines where points A and B are located in the original pipeline data table, obtain the pipe diameter information, and use it as D in the calculation.

[0029] The solution of the shortest path is:

[0030] (1) Initialize the number of nodes, node access flag variables, and traverse the cache queue;

[0031] (2) The starting node is set to a discrete point denoted as u1, and the ending node is set to a sensor position point numbered as v1 in the segmentation area. The starting node number is added to the traversal cache queue, and the access flag variable corresponding to the point is set to 1;

[0032] (3) Define the first point of the cache queue as m, traverse all points in the network in order of number, record the traversed point number as s, check the adjacency matrix, add all nodes connected to the first point of the queue to the end of the queue in sequence, and set the corresponding access flag variable to 1 for each visit; store the first and last elements of the queue after each visit in the matrix O;

[0033] (4) If the sensor position v1 does not appear in the cache queue queue, the first element of the cache queue queue is set to empty and the remaining elements are moved forward by one position;

[0034] (5) Repeat step (3) until the sensor position v1 appears in the cache queue queue. Through the path point information stored in the matrix O, trace the source level by level to obtain the complete shortest path from the discrete point u1 to the sensor position v1.

[0035] (6) By traversing the starting vertex and the end point, changing the position of the starting node and the end point, and repeating the above steps, all required shortest paths are obtained.

[0036] The method further includes: determining the repeated coordinate points, specifically:

[0037] When searching the pipeline coordinate data one by one, determine whether the longitude and latitude of the location are consistent with the existing coordinates in the coordinate matrix. For consistent longitude data, record the row number of the existing coordinates in the Coordinate matrix and store it in the vector ko; for consistent latitude data, record the row number of the existing coordinates in the Coordinate matrix and store it in the vector ki; if the intersection of vectors ko and ki is not empty, it indicates that the coordinate point currently being searched is a duplicate point and needs to be merged.

[0038] The beneficial effects of the technical solution provided by the present invention are:

[0039] 1. The present invention uses the shortest path planning and builds a corresponding experimental platform, which can monitor the urban water supply pipeline in real time and locate the leakage of complex pipe networks, with high feasibility;

[0040] 2. The present invention uses a small number of pressure sensors and has a low application cost, and can be used to optimize pipeline maintenance work and avoid the occurrence of serious safety and environmental accidents such as casualties caused by pipeline leakage. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a flow chart of the fluid pipeline leakage monitoring algorithm based on the shortest path planning;

[0042] Figure 2 This is a diagram of the urban pipe network and sensor distribution;

[0043] Figure 3 To delineate the regional version map;

[0044] Figure 4 Schematic diagram of a multi-pass pipe section. DETAILED DESCRIPTION

[0045] In order to make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention are described in further detail below.

[0046] Example 1

[0047] A method for monitoring fluid pipeline leakage based on shortest path planning, see Figure 1 , the method comprises the following steps:

[0048] Step 101: Establish a negative pressure wave propagation model in a single straight pipe;

[0049] Among them, by analyzing the movement of fluid in the pipeline, a pressure wave propagation model is established and the wave velocity formula is obtained.

[0050] Step 102: Survey and map the key points of the pipeline, obtain the longitude and latitude coordinates of several points, make a pipeline network distribution data table, establish an adjacency matrix based on the connection relationship between different points, and make a pipeline network diagram in Matlab;

[0051] The pipe network diagram can provide an intuitive understanding of the distribution of pipes.

[0052] Step 103: taking points at fixed intervals in all pipe sections of the pipe network, and subdividing the entire pipe network into a number of discrete points;

[0053] Step 104: Calculate the shortest path from each discrete point to each deployed sensor, and then calculate the arrival time of the negative pressure wave signal detected by each deployed pressure sensor when a leak occurs at each discrete point.

[0054] Step 105: The arrival time of step 104 is formed into a data column, and a time delay standard library is established therefrom. When a leak actually occurs at a certain location, the arrival time of the negative pressure wave measured by each sensor is compared with the time delay standard library, and the coordinates of the leak point are screened.

[0055] In summary, the embodiment of the present invention provides a fluid pipeline leakage monitoring algorithm based on shortest path planning through the above steps 101-105, which can realize leakage positioning of complex urban pipelines and has wider applicability.

[0056] Example 2

[0057] The scheme in Example 1 is further introduced below in combination with specific examples and calculation formulas, as described below for details:

[0058] Step 201: Establish a negative pressure wave propagation model in a single straight pipe;

[0059] By analyzing the movement of fluid in the pipeline, a pressure wave propagation model is established to obtain the wave velocity formula, which specifically includes the following steps:

[0060] By analyzing the fluid propagation process in the pipeline, the pressure wave propagation velocity formula in the thin-walled tube is obtained:

[0061]

[0062] Where K is the elastic coefficient of the liquid, Pa; ρ is the liquid density, kg / m 3 ; δ is the pipe wall thickness, m; D is the pipe inner diameter, m; E is the elastic modulus of the pipe, which is about 2.0×10 11 Pa; C1 is the constraint coefficient of the pipe, which depends on the constraint conditions of the pipe.

[0063] In this embodiment, the main parameters are set as follows: elastic coefficient of water K: 2.1×10 9 Pa; water density ρ: 1.0×10 3 kg / m 3 ; Pipe wall thickness δ: 0.05m; Elastic modulus of steel pipe E: 2.0×10 11 Pa; constraint coefficient of pipe C1: 0.81.

[0064] Among them, the inner diameter D of the pipeline varies with different pipe sections.

[0065] Step 202: Survey and map the key points of the pipeline, obtain the longitude and latitude coordinates of several points, make a pipeline network distribution data table, establish an adjacency matrix based on the connection relationship between different points, and select appropriate locations of the pipeline to deploy several pressure sensors, and make a pipeline network diagram in Matlab, which specifically includes the following steps:

[0066] Step 2.1: Process the known pipe network distribution data and convert the data table into a matrix (the blank part in the table is recorded as NaN), named N. Each row of data in N represents the longitude and latitude coordinates of the endpoint or bend point of a pipeline (not necessarily a straight pipe).

[0067] Step 2.2: Establish an adjacency matrix C based on the connection relationship between different points in the pipe network.

[0068] Suppose there are n distinct points, numbered 1, 2, 3, ... n, then the adjacency matrix C has n rows and n columns. Let the element in the i-th row and j-th column of the adjacency matrix be cij , then:

[0069]

[0070] The adjacency matrix C is established by searching N row by row and column by column. When searching in each row, starting from the second pair of coordinate points, the adjacency matrix elements corresponding to each coordinate point and its previous point are set to 1 (set to 0 if this condition is not met). Define the coordinate matrix Coordinate. At the same time, every time a longitude and latitude coordinate that does not appear is searched, it is added to the matrix Coordinate. The number of its rows in the matrix is ​​the number of the point.

[0071] In order to reduce the amount of calculation, the repeated points are merged in the process of establishing the adjacency matrix C. Suppose the point searched in a certain cycle is P1 (where P1 is not the first point in the row), and the point searched before point P1 is P0. The possible existence of points P0 and P1 is analyzed:

[0072] (1) Points P0 and P1 are the coordinate points that appear for the first time. Let the point numbers of points P0 and P1 be a and b, then c ab =1;

[0073] (2) Point P0 is not the first coordinate point to appear. The point with the same coordinate value is numbered k. However, point P1 is the first coordinate point to appear. The point number of point P1 is b. Then c kb =1;

[0074] (3) Point P1 is not the first coordinate point to appear. The point with the same coordinate value is numbered t. However, point P0 is the first coordinate point to appear. The point number of point P0 is a. Then c at =1;

[0075] (4) Point P0 is not the first coordinate point to appear, and the point with the same coordinate value is numbered k. Point P1 is not the first coordinate point to appear, and the point with the same coordinate value is numbered t. Then c kt =1.

[0076] The above rules are illustrated with a set of examples. Table 1 shows a portion of sample data of Table N. The three rows of coordinates represent three pipelines. Table 2 shows a version of these points arranged in numerical order.

[0077] Take point 2 (117.3205°E, 39.1299°N) and point 3 (117.3205°E, 39.12999°N) as examples. Since they are located on the same pipeline and are adjacent points, we have c 23 =1;

[0078] For points 5 and 6, since they are not in the same pipeline, there is c 56 =0;

[0079] In the third row of data in Table 1, observe point 11 and point 2 (117.3205°E, 39.1299°N). Since the two are connected, we have c 11,2 =1; Similarly, point 2 is connected to point 12, c 2，12 =1.

[0080] Table 1 Example of pipe network latitude and longitude table (partial)

[0081]

[0082] Table 2 Point number list

[0083]

[0084]

[0085] The following is an introduction to the software implementation process of determining duplicate coordinate points. When searching for pipeline coordinate data one by one, determine whether the longitude and latitude of the location are consistent with the existing coordinates in the coordinate matrix. For consistent longitude data, record the row number of the existing coordinates in the coordinate matrix and store it in the vector ko; for consistent latitude data, record the row number of the existing coordinates in the coordinate matrix and store it in the vector ki; if the intersection of ko and ki is not empty, it indicates that the coordinate point currently being searched is a duplicate point and needs to be merged.

[0086] Step 2.3, based on the coordinate matrix Coordinate and the adjacency matrix C established in step 2.2, realize the pipe network drawing in matlab;

[0087] Since the latitude and longitude of the pressure sensor are known, it can be marked directly on the pipe network map.

[0088] In this embodiment, the actual pipe network diagram and sensor distribution are as follows: Figure 2 In the version shown, 10 pressure sensors are deployed (marked with dots in the figure, and the numbers represent the sensor numbers).

[0089] Figure 2 The 10 pressure sensors are mainly distributed in the middle section of the urban pipe network. The sensors are sparsely arranged, and the straight-line distance between adjacent units ranges from 300 meters to 5,500 meters.

[0090] Step 203: Select points at fixed intervals in all pipe sections of the pipe network to subdivide the entire pipe network into a number of discrete points. The specific steps are as follows:

[0091] The pipeline formed by any two adjacent points obtained in step 202 is regarded as a straight pipe segment, and the longitude and latitude of all two adjacent points can be linearly interpolated to obtain a number of new pipe segment coordinates, thereby subdividing the entire pipe network into a number of discrete points.

[0092] In this embodiment, in order to reduce the amount of calculation, only a specific area of ​​the pipe network is intercepted and regarded as a zone where leakage may occur. Figure 3 The pipeline is subdivided, and the interval between subdivision points is about 5 meters after interpolation.

[0093] Step 204: Calculate the arrival time of the negative pressure wave signal detected by each deployed pressure sensor when a leak occurs at each discrete point position, form a data column, and thereby establish a delay standard library. The specific steps are:

[0094] Step 4.1.1: After the pipe network is subdivided into several discrete points, a larger adjacency matrix is ​​re-established based on the connection relationship between each point.

[0095] Step 4.1.2: With the help of the new adjacency matrix, calculate the shortest path from each discrete point to each deployed sensor.

[0096] The specific implementation of solving the shortest path in the embodiment of the present invention is:

[0097] (1) Initialize the number of nodes (the number of discrete points), the node access flag variable (if the node has not been visited, the variable is 0, otherwise it is set to 1), and traverse the cache queue (referred to as queue, which is an empty vector at the beginning);

[0098] (2) The starting node is set to a discrete point (the point number is denoted as u1), the ending node is set to a sensor position in the segmented area (the point number is denoted as v1), the starting node number is added to the traversal cache queue, and the access flag variable corresponding to the point is set to 1;

[0099] (3) Define the first point in the queue as m, traverse all points in the network in order of number, record the traversed point number as s, and check the adjacency matrix to find all points that meet c ms = 1 (i.e. the nodes connected to the first point of the queue) are added to the end of the queue in sequence. Each time a point is visited, its corresponding access flag variable is set to 1. Then the first and last elements of the queue after each visit are stored in the matrix O;

[0100] (4) After step (3), if the sensor position v1 does not appear in the queue, the first element of the queue is set to empty and the remaining elements are moved forward by one position;

[0101] (5) Repeat step (3) until the sensor position v1 appears in the queue. Then, through the path point information stored in the matrix O, the complete shortest path from the discrete point u1 to the sensor position v1 is obtained by tracing back step by step.

[0102] Here we take an example to illustrate the step-by-step tracing. If the discrete point u1 is 2, the sensor position v1 is 8, and the matrix O is:

[0103]

[0104] The matrix O has 5 rows and 2 columns, and the elements in each row form pairs (2, 4), (2, 5), (4, 7), (5, 9), (7, 8), which are the first and last elements of the queue after each access. When the sensor position v1 (corresponding to 8) is accessed, the search stops and the tracing begins. First, the first element of the row where 8 is located (i.e. 7) is found. At this time, jump to the third row of the matrix O, find the first element is 4, jump to the first row, access the discrete point u1 (i.e. 2), and the tracing ends. The overall tracing process is 8-7-4-2, and the corresponding shortest path from the discrete point u1 to the sensor position v1 is 2-4-7-8.

[0105] (6) By traversing the starting vertex and the end point, changing the position of the starting node and the end point, and repeating the above steps, all required shortest paths are obtained.

[0106] Step 4.2: The line connecting two adjacent points in the shortest path obtained in step 4.1.2 is regarded as a straight pipe segment. The corresponding propagation time can be obtained by directly using the pipe length and the wave speed calculated by the model formula (1) in step 201 as the quotient (the elastic coefficient, constraint coefficient and other parameters of all pipe segments are set to be the same during calculation, and only the inner diameter and length of the pipe are different). The arrival time of the negative pressure wave signal detected by each pressure sensor can be calculated by accumulating multiple propagation times. All arrival times form a data column, and a delay standard library is established from them.

[0107] The parameter of the pipe inner diameter D must be taken into account in actual calculations. There are many tee and cross positions in the pipe network, that is, one point may correspond to multiple pipe inner diameter values. A matrix Mul is established to store the numbers of these points and the corresponding inner diameters in the record matrix Mul.

[0108] Assume that two adjacent path points in the shortest path to be solved are A and B, and the schematic diagram of the case with multiple pipe fittings is as follows Figure 4 As shown, the possible situations of A and B are analyzed below:

[0109] (1) Neither A nor B is a multi-pass position, and the inner diameter of the pipe where A (or B) is located is directly used as D for calculation;

[0110] (2) If A (or B) is a multi-pass position, but B (or A) corresponds to only one pipe inner diameter, the pipe inner diameter corresponding to B (or A) shall be used as D for calculation;

[0111] (3) If both A and B are multi-pass locations and have the same pipe diameter, this value is directly used as D in the calculation;

[0112] (4) If both A and B are multi-pass locations and have two or more sets of identical pipe diameters, search for the pipelines where points A and B are located in the original pipeline data table, obtain the pipe diameter information, and use it as D in the calculation.

[0113] The following example illustrates the above situation:

[0114] If the inner diameter of the pipe at point A is 820 mm, and the inner diameter of the pipe at point B is 820 mm or 529 mm, then the inner diameter of the pipe section AB is 820 mm;

[0115] If the data of A and B in the Mul matrix are as shown in Table 3, that is, the pipes where A and B are located have two inner diameters, then the inner diameter of the pipe section AB is 325 mm;

[0116] Table 3 Inner diameter data 1

[0117] Point Number Inner diameter 1 / mm Inner diameter 2 / mm Point A 1482 1020 325 Point B 637 630 325

[0118] If the data of A and B in the Mul matrix are as shown in Table 4, that is, the pipes where A and B are located have two inner diameters, and the values ​​are the same, then check the positions of A and B in the pipe network data table N, and get that the inner diameter of the pipe section is 159 mm, then use 159 mm as the inner diameter of the AB section pipe for calculation;

[0119] Table 4 Inner diameter data 2

[0120] Point Number Inner diameter 1 / mm Inner diameter 2 / mm Point A 487 159 219 Point B 637 159 219

[0121] In this embodiment, in order to verify the effectiveness of the algorithm, artificial water release is used to simulate pipeline leakage. Figure 3 A point is selected on the regional pipeline for water release, and the partial time delay standard library established is shown in Table 3. The data in the table means the time (unit: second) when the corresponding sensor detects the negative pressure wave signal if a leak occurs at the discrete point.

[0122] Table 3 Delay standard library (partial)

[0123]

[0124] Step 205: The time calculated in step 204 is formed into a data column. When a leak actually occurs at a certain location, the arrival time of the negative pressure wave measured by each sensor is compared with the delay standard library. The point corresponding to the data column in the standard library that is closest to the measured propagation delay is determined as the leak point, and the leak point position is output.

[0125] In the embodiment, the latitude and longitude coordinates of the actual water release point are (117.23804°E, 39.07384°N), and the measured time is compared with the delay standard library to obtain the point with the smallest sum of squared differences in the 25 data columns. The output prediction leakage point result is shown in Table 4, and it can be seen that the actual water release point is in the prediction result.

[0126] Table 4 Prediction results of position

[0127]

[0128] In summary, the fluid pipe network leakage monitoring method based on shortest path planning designed in the embodiment of the present invention can realize leakage monitoring and positioning of the fluid pipe network.

[0129] Unless otherwise specified, the models of the components in the embodiments of the present invention are not limited, and any device that can perform the above functions may be used.

[0130] Those skilled in the art will appreciate that the accompanying drawing is only a schematic diagram of a preferred embodiment, and the serial numbers of the embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0131] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for monitoring fluid pipe network leakage based on shortest path planning, characterized in that: The method comprises the following steps: Establish a negative pressure wave propagation model in a single straight pipeline; map key points of the pipeline, obtain the longitude and latitude coordinates of several points, make a pipeline network distribution data table, establish an adjacency matrix based on the connection relationship between different points, and make a pipeline network diagram in MATLAB; Points are taken at fixed intervals in all pipe sections of the pipe network to subdivide the entire pipe network into a number of discrete points; Solve the shortest path from each discrete point to each deployed sensor, and then calculate the arrival time of the negative pressure wave signal detected by each deployed pressure sensor when a leak occurs at each discrete point; The arrival time is formed into a data column, and a time delay standard library is established based on it. When a leak actually occurs at a certain location, the arrival time of the negative pressure wave measured by each sensor is compared with the time delay standard library, and the coordinates of the leak point are screened; The pipeline network distribution data table is prepared, and an adjacency matrix is ​​established according to the connection relationship between different points. The pipeline network diagram is made in MATLAB as follows: Suppose there are n distinct points, numbered 1, 2, 3, ... n, then the adjacency matrix C has n rows and n columns. Let the element in the i-th row and j-th column of the adjacency matrix be c ij , then: The adjacency matrix C is established by searching the matrix N row by row and column by column. When searching in each row, starting from the second pair of coordinate points, the adjacency matrix elements corresponding to each coordinate point and its previous point are set to 1. Define the coordinate matrix Coordinate. Every time a longitude and latitude coordinate that does not appear is searched, add it to the matrix Coordinate. The number of its rows in the matrix is ​​the number of the point. In the process of establishing the adjacency matrix C, the repeated points are merged. Let the point searched in a certain cycle be P1, and the point searched before point P1 be P0. The existence of points P0 and P1 is analyzed: (1) Points P0 and P1 are the coordinate points that appear for the first time. Let the point numbers of points P0 and P1 be a and b, then c ab =1; (2) Point P0 is not the first coordinate point to appear. The point with the same coordinate value is numbered k. However, point P1 is the first coordinate point to appear. The point number of point P1 is b. Then c kb =1; (3) Point P1 is not the first coordinate point to appear. The point with the same coordinate value is numbered t. However, point P0 is the first coordinate point to appear. The point number of point P0 is a. Then c at =1; (4) Point P0 is not the first coordinate point to appear, and the point with the same coordinate value is numbered k. Point P1 is not the first coordinate point to appear, and the point with the same coordinate value is numbered t. Then c kt =1.

2. A method for monitoring fluid pipe network leakage based on shortest path planning according to claim 1, characterized in that: The method of forming a data column from the arrival time and establishing a delay standard library is as follows: The line connecting two adjacent points in the shortest path is regarded as a straight pipe section. The propagation time is obtained by taking the pipe length and the wave speed as the quotient. Multiple propagation times are accumulated and the arrival time of the negative pressure wave signal detected by each pressure sensor is calculated. All arrival times form a data column, and a time delay standard library is established based on this.

3. The method for monitoring fluid pipe network leakage based on shortest path planning according to claim 1, characterized in that: Two adjacent path points in the shortest path are A and B. The situations of A and B are analyzed as follows: (1) A and B are not multi-pass locations, and the inner diameter of the pipe where A or B is located is directly used as D for calculation; (2) If A or B is a multi-pass position, but B or A corresponds to only one pipe inner diameter, the pipe inner diameter corresponding to B or A shall be used as D for calculation; (3) A and B are both multi-pass locations, and they have the same set of pipe diameters. This value is used as D in the calculation; (4) If both A and B are multi-pass locations and have two or more sets of identical pipe diameters, search for the pipelines where points A and B are located in the original pipeline data table, obtain the pipe diameter information, and use it as D in the calculation.

4. The method for monitoring fluid pipe network leakage based on shortest path planning according to claim 1, characterized in that: The solution of the shortest path is: (1) Initialize the number of nodes, node access flag variables, and traverse the cache queue; (2) The starting node is set to a discrete point denoted as u1, and the ending node is set to a sensor position point numbered as v1 in the segmentation area. The starting node number is added to the traversal cache queue, and the access flag variable corresponding to the point is set to 1; (3) Define the first point of the cache queue as m, traverse all points in the network in order of number, record the traversed point number as s, check the adjacency matrix, add all nodes connected to the first point of the queue to the end of the queue in sequence, and set the corresponding access flag variable to 1 for each visit; store the first and last elements of the queue after each visit in the matrix O; (4) If the sensor position v1 does not appear in the cache queue queue, the first element of the cache queue queue is set to empty and the remaining elements are moved forward by one position; (5) Repeat step (3) until the sensor position v1 appears in the cache queue queue. Through the path point information stored in the matrix O, trace the source level by level to obtain the complete shortest path from the discrete point u1 to the sensor position v1. (6) By traversing the starting vertex and the end point, changing the position of the starting node and the end point, and repeating the above steps, all required shortest paths are obtained.

5. The method for monitoring fluid pipe network leakage based on shortest path planning according to claim 1, characterized in that: The method further includes: determining the repeated coordinate points, specifically: When searching the pipeline coordinate data one by one, determine whether the longitude and latitude of the location are consistent with the existing coordinates in the coordinate matrix. For consistent longitude data, record the row number of the existing coordinates in the Coordinate matrix and store it in the vector ko; for consistent latitude data, record the row number of the existing coordinates in the Coordinate matrix and store it in the vector ki; if the intersection of vectors ko and ki is not empty, it indicates that the coordinate point currently being searched is a duplicate point and a merge operation is required.

Images