A Clustering and Screening Method and System for Typical Monitoring Areas of Urban Non-Point Source Pollution

Abstract

The present invention discloses a clustering and screening method and system for typical monitoring areas of urban non-point source pollution. First, according to the catchment sub-areas, land use, and terrain of the target area, calculate the area, proportion of construction land, and slope of the catchment sub-areas; then perform normalization processing on the area, proportion of construction land, and slope of the catchment sub-areas, and initialize the clustering centers according to the spatial point density and the distribution of the proportion of construction land; then traverse the distances from all catchment sub-area points to each clustering center point, divide each catchment sub-area point into the clustering cluster with the closest distance, update the clustering centers, and repeat this step until the positions of the new clustering centers no longer change or converge to the set range; finally, in combination with the center points obtained by clustering, find the catchment sub-areas with the smallest variance from the clustering centers in each clustering cluster as the finally selected typical monitoring areas. The present invention improves the representativeness and stability of the selection of typical monitoring areas, and provides technical support for the scientific monitoring of urban non-point source pollution.

Description

Technical Field

[0001] The present invention relates to pollution load monitoring, and in particular to a method and system for clustering and screening typical monitoring areas of urban non-point source pollution. Background Art

[0002] With the rapid development of urbanization, the ecological environment problems caused by urban non-point source pollution generated by rainfall runoff scouring have become increasingly prominent, and have become the key to restricting the "quality improvement" of the urban water environment. Urban non-point source pollution has the characteristics of random, intermittent emissions, and large spatio-temporal variation. Conducting urban non-point source pollution monitoring can provide important basic data references for estimating urban non-point source pollution loads, and is of great significance for urban non-point source pollution prevention and control and water environment governance.

[0003] At present, the main idea of conducting urban non-point source pollution monitoring is to analyze the spatial runoff generation and concentration trends of non-point source pollution based on data such as terrain, underground pipe networks, and rivers, delimit sub-catchment areas of the city, and estimate the non-point source pollution loads of each catchment area to achieve the estimation of the non-point source pollution load of the entire city. Obviously, it is time-consuming and laborious to conduct non-point source pollution monitoring on all catchment areas of the city, and it is not economically feasible. The artificial selection of typical monitoring areas has no quantitative index reference, has the problems of randomness and poor representativeness, resulting in great uncertainty in the estimation results of urban non-point source pollution loads.

[0004] In addition, due to differences in population density, underlying surface, drainage pipe networks, and functional zoning during the process of urbanization development, there are significant spatial differences in urbanization development. For example, areas with low urbanization intensity are mainly mountainous areas in the city, with relatively large catchment area, the smallest proportion of construction land, and the largest slope; areas with medium urbanization intensity generally have both mountainous areas and urban construction areas, with moderate catchment area, proportion of construction land, and slope; while areas with high urbanization intensity are mainly urban built-up areas, with high development intensity, mostly located in flat areas, with a relatively large proportion of construction land and a small slope, and due to the artificial cutting of the runoff trend by urban construction, their catchment area is generally small. The selection of typical monitoring areas for urban non-point source pollution needs to consider this objectively existing difference in urbanization development, so there are often more than one typical monitoring area. How to scientifically and reasonably screen typical monitoring areas for non-point source pollution in highly urbanized areas has become an urgent problem to be solved in urban non-point source pollution monitoring. Summary of the Invention

[0005] Object of the Invention: The object of the present invention is to provide a method and system for clustering and screening typical monitoring areas of urban non-point source pollution, to overcome the problems of randomness and poor representativeness existing in artificial subjective selection, and to provide technical support for scientific monitoring of urban non-point source pollution.

[0006] Technical solution: To achieve the above-mentioned invention objective, a clustering and screening method for typical monitoring areas of urban non-point source pollution provided by the present invention includes:

[0007] (1) Calculate the area, construction land proportion, and slope of the catchment area according to the catchment area division, land use, and terrain of the target area;

[0008] (2) Normalize the area, construction land proportion, and slope of the catchment area to form spatial point coordinates, and each catchment area corresponds to a point in space; count the number N of catchment area points within the set spatial distance radius range of each catchment area point, and screen out the catchment area points with N not less than the set threshold to obtain a high-density catchment area set H; sort according to the construction land proportion of the catchment area points in set H, and select k catchment area points as the initial clustering centers, where k is the set number of monitoring areas;

[0009] (3) Traverse the Euclidean distances from all catchment area points to the k clustering center points, divide each catchment area point into the nearest clustering cluster, and update the clustering center of each clustering cluster. Repeat this step until the position of the new clustering center no longer changes or converges to the set range, and finally obtain k clustering centers;

[0010] (4) Combine the k center points obtained by clustering, and find the catchment area with the smallest variance from the clustering center in each clustering cluster as the finally screened typical monitoring area.

[0011] Preferably, in step (1), the area, construction land proportion, and slope of each catchment area are statistically calculated in GIS, where the construction land proportion is the proportion of the sum of the roof and road areas to the catchment area, and the slope is the average slope of all grids in the catchment area after calculating the slope according to the terrain DEM data.

[0012] Preferably, in step (2), the maximum values of the area, construction land proportion, and slope of the catchment area are respectively used for normalization, and the normalized values are between 0 and 1.

[0013] Preferably, in step (2), the method for initializing the clustering center is:

[0014] First, calculate the distance d(U i (A' i , C' i , S' i ) and U j (A' j , C' j , S' j ) of any two catchment area points U i , U j ), and the spatial distance radius With the threshold where m is the total number of catchment sub - areas in the target area; A' i 、A' j 、C' i 、C' j 、S' i 、S' j respectively represent the area, the proportion of construction land, and the slope of the i - th and j - th catchment sub - area points after normalization;

[0015]

[0016] Then, count the number of catchment sub - areas N within the spatial distance radius r for each catchment sub - area point, and select the catchment sub - area points with N≥M as the high - density catchment sub - area set H;

[0017] Suppose there are h catchment sub - area points in the high - density catchment sub - area set H. Sort the points in set H in descending order according to the proportion of construction land. Then, the initialized 1st, 2nd, 3rd, …, k - 1th, and kth cluster centers are the catchment sub - area points with sorting equal to 1, h respectively; where, floor represents rounding down.

[0018] Preferably, the step (3) specifically includes:

[0019] (3 - 1) Traverse the Euclidean distances from all catchment sub - area points to the k cluster centers, and divide each catchment sub - area point into the cluster with the closest distance; where, the Euclidean distance d i ,C' i ,S' i ) of the i - th catchment sub - area point (A' j ,y j ,z j ) to the j - th cluster center (x j ,y j ,z j ) is calculated as follows: i,j The calculation formula is as follows:

[0020]

[0021] (3 - 2) Update the cluster centers; Suppose there are n catchment sub - area points in the cluster of a certain cluster center (x, y, z). The calculation formula for the new cluster center position is as follows:

[0022]

[0023]

[0024]

[0025] where, (new_x, new_y, new_z) represents the spatial coordinates of the new cluster center;

[0026] (3-3) Repeat steps (3-1) to (3-2) until the position of the new cluster center no longer changes or converges to the set range, and finally obtain k cluster centers; among them, the judgment condition for the convergence of the cluster center is as follows:

[0027]

[0028] Among them, δ is the convergence threshold.

[0029] Preferably, in step (4), assume that there are n catchment area points in a certain cluster, then the distance variance D of the i-th catchment area point among them i is calculated as follows:

[0030]

[0031] Among them, d i represents the distance from the i-th catchment area to its cluster center, and represents the average value of the distances from all catchment areas in the cluster to the cluster center.

[0032] Based on the same inventive concept, a clustering and screening system for typical monitoring areas of urban non-point source pollution provided by the present invention includes:

[0033] A data acquisition module, configured to calculate the area, construction land proportion, and slope of the catchment area according to the catchment area, land use, and terrain of the target area;

[0034] A cluster center initialization module, configured to normalize the area, construction land proportion, and slope of the catchment area to form spatial point coordinates, and each catchment area corresponds to a point in space; count the number N of catchment areas within the set spatial distance radius range of each catchment area point, screen out the catchment area points with N not less than the set threshold to obtain a high-density catchment area set H; sort according to the construction land proportion of the catchment area points in the set H, and select k catchment area points as the initial cluster centers, where k is the set number of monitoring areas;

[0035] A clustering module, configured to traverse the Euclidean distances from all catchment area points to the k cluster center points, divide each catchment area point into the cluster with the closest distance, update the cluster center of each cluster, and repeat this step until the position of the new cluster center no longer changes or converges to the set range, and finally obtain k cluster centers;

[0036] And a monitoring area determination module, configured to combine the k center points obtained by clustering, find the catchment area with the smallest variance from the cluster center in each cluster as the finally screened typical monitoring area.

[0037] Based on the same inventive concept, a computer system provided by the present invention includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the computer program is loaded into the processor, the steps of the clustering and screening method for typical monitoring areas of urban non-point source pollution are implemented.

[0038] Based on the same inventive concept, a computer-readable storage medium provided by the present invention stores a computer program. When the computer program is executed by a processor, the steps of the clustering and screening method for typical monitoring areas of urban non-point source pollution are implemented.

[0039] Beneficial effects: Compared with the prior art, the significant advantages of the present invention are as follows:

[0040] (1) In previous related studies, the selection of typical monitoring areas mostly relied on subjective human judgment, which had problems of randomness and poor representativeness, resulting in great uncertainty in the estimation results of urban non-point source pollution loads. The present invention uses the catchment area, the proportion of construction land, and the slope to measure the spatial difference degree of different catchment areas, and can quantitatively cluster and screen typical monitoring areas of urban non-point source pollution, improving the representativeness and pertinence of urban non-point source pollution monitoring.

[0041] (2) The present invention improves the clustering center initialization method, focuses on considering the representativeness of the proportion of construction land for urbanization development, and finally the selected typical monitoring areas can represent different degrees of urbanization development. At the same time, the present invention can independently define the number of clusters, and can maintain relatively stable clustering results at different numbers of clusters. Moreover, as the number of clusters increases, more information on the differences in urbanization development will be obtained for the typical monitoring areas obtained by clustering. Description of the Drawings

[0042] Figure 1 is a schematic flowchart of an embodiment of the present invention;

[0043] Figure 2 is a schematic diagram of land use and catchment area division of this embodiment;

[0044] Figure 3 is a schematic diagram of the topographic spatial distribution of this embodiment;

[0045] Figure 4 is a schematic diagram of the numerical spatial distribution of each catchment area obtained by statistics;

[0046] Figure 5 is a schematic diagram of the initial clustering centers corresponding to k = 3 and k = 5;

[0047] Figure 6 is a flowchart of the clustering algorithm for typical monitoring areas;

[0048] Figure 7It is a schematic diagram of the clustering results of catchment areas when k = 3 and k = 5;

[0049] Figure 8 It is a schematic diagram of the spatial distribution of typical monitoring areas when k = 3 and k = 5. Specific implementation manners

[0050] The technical solutions of the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0051] As Figure 1 shown, the typical monitoring area clustering and screening method for urban non-point source pollution disclosed in the embodiment of the present invention includes the following steps:

[0052] (1) Calculate the area, construction land proportion and slope of the catchment area according to the catchment area, land use and terrain of the target area. Specifically, it includes:

[0053] (1-1) Collect the catchment area, land use and terrain data of the target area, where the land use types include roofs, roads, green spaces and water bodies, etc.

[0054] (1-2) In GIS, count the area A, construction land proportion C and slope S of each catchment area. Then any catchment area U can be represented by the spatial point coordinates (A, C, S) composed of the area, construction land proportion and slope. Among them, the construction land proportion is the proportion of the sum of the roof and road areas to the catchment area. After calculating the slope according to the terrain DEM data, the average slope of all grids in the catchment area is taken.

[0055] The catchment area, land use and terrain of the target area collected in this embodiment are as Figure 2 and Figure 3 shown. The total area of the target area is 263.83 km 2 , and a total of 59 catchment areas (m = 59) are divided. Among them, roofs and roads are mainly distributed in flat areas, with proportions of 28.31% and 10.98% respectively, while green spaces are mainly distributed in mountainous areas with higher terrain, with a proportion of up to 57.83%. The area, construction land proportion and slope of each catchment area are counted in GIS, and a spatial coordinate system is constructed as Figure 4 shown.

[0056] (2) Perform normalization processing on the area, construction land proportion and slope of the catchment area, and initialize k clustering centers according to the set number of monitoring areas. Specifically, it includes:

[0057] (2-1) Normalization processing: To avoid the influence of different dimensions on the distance calculation result, use the maximum value of each dimension to normalize the 3D data of the catchment area into numbers between 0 and 1. The calculation formula is as follows:

[0058] A' i = A i / A max

[0059] C' i = C i / C max

[0060] S' i = S i / S max

[0061] Among them, A i , C i , S i respectively represent the area, the proportion of construction land, and the slope of the i-th catchment sub-region; A max , C max , S max respectively represent the maximum values of the area, the proportion of construction land, and the slope in the catchment sub-region; A' i , C' i , S' i respectively represent the area, the proportion of construction land, and the slope of the i-th catchment sub-region after normalization.

[0062] (2-2) Initialization of clustering centers: Assume that there are m catchment sub-regions to be clustered into k categories (2 ≤ k < m), then k clustering centers need to be initialized. To avoid the problem of unstable clustering results caused by randomly initializing clustering centers in traditional methods, the improved idea of initializing clustering centers is as follows:

[0063] First, calculate the distance d(U i (A' i , C' i , S' i ) and U j (A' j , C' j , S' j ) of any two catchment sub-region points, as well as the spatial distance radius i , U j ) and the threshold where m is the total number of catchment sub-regions in the target area; A' Among them, A' i , A' j , C' i , C' j , S' i , S' j respectively represent the area, the proportion of construction land, and the slope of the i-th and j-th catchment sub-region points after normalization;

[0064]

[0065] Then, count the number of catchment areas N within the spatial distance radius r for each catchment area point, and select the catchment area points with N≥M as the high-density catchment area set H.

[0066] Suppose there are h catchment area points in the high-density catchment area set H. Sort the points in set H in descending order according to the proportion of construction land. Then, the initialized 1st, 2nd, 3rd, …, k-1th, and kth cluster centers are the catchment area points with sorting equal to 1, h respectively; where, floor represents rounding down.

[0067] In this embodiment, the maximum values of the area, proportion of construction land, and slope of each catchment area are counted, where A max is 32.55 km 2 , C max is 0.77, and S max is 21.81°.

[0068] Taking k = 3 and 5 as examples, 3 and 5 cluster centers are initialized respectively. It is calculated that r = 0.171. The initialized cluster centers are as Figure 5 shown.

[0069] (3) Traverse the Euclidean distances from all catchment areas to the k cluster center points, divide each catchment area into the cluster with the nearest distance, update the cluster center of each cluster, and repeat this step until the position of the new cluster center no longer changes or converges to the set range, and finally obtain k cluster centers. Specifically, it includes:

[0070] (3-1) Traverse the Euclidean distances from all catchment areas to the k cluster center points, and divide each catchment area into the cluster with the minimum distance; where, the Euclidean distance calculation formula from the i-th catchment area to the j-th cluster center is as follows:

[0071]

[0072] where, d i,j represents the Euclidean distance from the i-th catchment area to the j-th cluster center, and (x j , y j , z j ) represents the spatial coordinates of the j-th cluster center;

[0073] (3-2) Calculate the average value of the three coordinate axes in each cluster as the new cluster center; where, assuming that there are n catchment areas in the cluster of a certain cluster center (x, y, z), the new cluster center position calculation formula is as follows:

[0074]

[0075]

[0076]

[0077] Among them, (new_x, new_y, new_z) represents the spatial coordinates of the new clustering center;

[0078] (3-3) Repeat steps (3-1) to (3-2) until the position of the new clustering center no longer changes or converges to the set range, and finally obtain k new clustering centers; among them, the judgment conditions for the convergence of the new clustering center are as follows:

[0079]

[0080] Among them, δ is a convergence threshold that can be defined artificially.

[0081] The flow of the clustering algorithm in this embodiment is as Figure 6 shown. The defined convergence threshold δ is 0.0001. Through the iteration of the clustering algorithm, 59 catchment areas are finally clustered into 3 categories and 5 categories, as Figure 7 shown. The area, construction land proportion, and slope information of the three center points obtained by clustering are shown in Tables 1 and 2 below.

[0082] Table 1 Clustering center results for k = 3

[0083]

[0084]

[0085] Table 2 Clustering center results for k = 5

[0086]

[0087] (4) Combine the k center points obtained by clustering, and find the catchment area with the smallest variance of the distance to the clustering center in each clustering cluster as the finally selected typical monitoring area. Suppose there are n catchment area points in a certain clustering cluster, then the distance variance D of the i-th catchment area point in it i is calculated as follows:

[0088]

[0089] Among them, d i represents the distance from the i-th catchment area to its clustering center, represents the mean value of the distances from all catchment areas in the clustering cluster to the clustering center; the smallest distance variance D i means the closest to the average level of the distance. Selecting the catchment area with the smallest D i is more representative.

[0090] Combined with the clustering centers obtained in step (3) of this embodiment, the catchment area information with the smallest variance from the clustering centers in each clustering cluster is found, as shown in Tables 3 and 4. The spatial distributions of typical monitoring areas in the two cases are as Figure 8 shown, and subsequent urban non-point source pollution monitoring can be carried out for the typical areas selected by clustering.

[0091] Table 3 Clustering screening results of typical monitoring areas when k = 3

[0092]

[0093] Table 4 Clustering screening results of typical monitoring areas when k = 5

[0094]

[0095] Based on the same inventive concept, a clustering screening system for typical monitoring areas of urban non-point source pollution disclosed in an embodiment of the present invention includes: a data acquisition module, configured to calculate the area, construction land proportion, and slope of the catchment area according to the catchment area, land use, and terrain of the target area; a clustering center initialization module, configured to normalize the area, construction land proportion, and slope of the catchment area to form spatial point coordinates, and each catchment area corresponds to a point in space; count the number N of catchment areas within the set spatial distance radius range of each catchment area point, screen out the catchment area points with N not less than the set threshold to obtain a high-density catchment area set H; sort according to the construction land proportion of the catchment area points in the set H, and select k catchment area points as the initial clustering centers; a clustering module, configured to traverse the Euclidean distances from all catchment area points to the k clustering center points, divide each catchment area point into the nearest clustering cluster, update the clustering center of each clustering cluster, and repeat this step until the position of the new clustering center no longer changes or converges to the set range, and finally obtain k clustering centers; and a monitoring area determination module, configured to combine the k center points obtained by clustering, and find the catchment area with the smallest variance from the clustering center in each clustering cluster as the finally selected typical monitoring area.

[0096] For the specific working processes of the above-described modules, reference can be made to the corresponding processes in the foregoing method embodiments, which will not be elaborated here. The division of the modules is only a logical function division, and there may be other division methods in actual implementation. For example, multiple modules can be combined or integrated into another system.

[0097] Based on the same inventive concept, a computer system disclosed in an embodiment of the present invention includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the computer program is loaded into the processor, the steps of the clustering screening method for typical monitoring areas of urban non-point source pollution are implemented.

[0098] Based on the same inventive concept, a computer-readable storage medium disclosed in an embodiment of the present invention stores a computer program, and when the computer program is executed by a processor, the steps of the typical monitoring area clustering and screening method for urban non-point source pollution are implemented.

[0099] Those skilled in the art can understand that the technical solution of the present invention, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer system (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the embodiment of the present invention. The storage medium includes: various media that can store computer programs such as USB flash drives, mobile hard disks, read-only memory ROM, random access memory RAM, magnetic disks, or optical discs.

Claims

1. A clustering and screening method for typical monitoring areas of urban non-point source pollution, characterized in that, the method includes: (1) Calculate the area, construction land proportion, and slope of the catchment area according to the catchment area division, land use, and terrain of the target area; (2) Normalize the area, proportion of construction land, and slope of each catchment sub - area to form spatial point coordinates, with each catchment sub - area corresponding to a point in space; count the number N of catchment sub - area points within the set spatial distance radius range for each catchment sub - area point, and filter out the catchment sub - area points where N is not less than the set threshold to obtain the high - density catchment sub - area set H; sort the catchment sub - area points in set H according to the proportion of construction land, and select k catchment sub - area points as the initial clustering centers, where k is the set number of monitoring areas; assume there are h catchment sub - area points in the high - density catchment sub - area set H, and the initialized 1st, 2nd, 3rd, …, k - 1th, kth clustering centers are respectively the catchment sub - area points with sorting equal to 1, h; where floor represents rounding down. (3) Traverse the Euclidean distances from all catchment area points to k clustering centers, divide each catchment area point into the nearest clustering cluster, update the clustering center of each clustering cluster, and repeat this step until the position of the new clustering center no longer changes or converges to a set range, and finally obtain k clustering centers; (4) Combine the k centers obtained by clustering, and find the catchment area with the smallest variance from the clustering center in each clustering cluster as the finally selected typical monitoring area.

2. The clustering and screening method for typical monitoring areas of urban non-point source pollution according to claim 1, characterized in that: In the step (1), Statistically calculate the area A, construction land proportion C, and slope S of each catchment area in GIS. Among them, the construction land proportion is the proportion of the sum of the roof and road areas to the catchment area. The slope is the average value of the slopes of all grids in the catchment area after calculating the slope according to the terrain DEM data.

3. The clustering and screening method for typical monitoring areas of urban non-point source pollution according to claim 1, characterized in that: In the step (2), the maximum values of the area, construction land proportion, and slope of the catchment area are respectively used for normalization, and the normalized values are between 0 and 1.

4. The clustering and screening method for typical monitoring areas of urban non-point source pollution according to claim 1, characterized in that: In the step (2), the method for initializing the clustering center is: First, calculate the distance between any two catchment sub - area points U i (A' i , C' i , S' i ) and U j (A' j , C' j , S' j ), denoted as d(U i , U j ), as well as the spatial distance radius and the threshold where m is the total number of catchment sub - areas in the target area; A' i , A' j , C' i , C' j , S' i , S' j respectively represent the area, construction land proportion, and slope of the i - th and j - th catchment sub - area points after normalization; Then count the number N of catchment areas within the spatial distance radius r of each catchment area point, and screen the catchment area points with N≥M as the high-density catchment area set H; Suppose there are h catchment area points in the high-density catchment area set H. Sort the points in set H in descending order according to the proportion of construction land. Then the 1st, 2nd, 3rd, …, k-1th, and kth clustering centers after initialization are the catchment area points with sorting equal to 1, h respectively.

5. The clustering and screening method for typical monitoring areas of urban non-point source pollution according to claim 1, characterized in that: The step (3) specifically includes: (3-1) Traverse the Euclidean distances from all catchment points to the k clustering centers, and divide each catchment point into the clustering cluster with the closest distance; among them, the Euclidean distance d of the i-th catchment point (A' i , C' i , S' i ) to the j-th clustering center (x j , y j , z j ) is calculated as follows: i,j The calculation formula is as follows: (3-2) Update the clustering center; assume that there are n catchment area points in the clustering cluster of a certain clustering center (x, y, z), then the calculation formula for the position of the new clustering center is as follows: Among them, (new_x, new_y, new_z) represents the spatial coordinates of the new clustering center; (3-3) Repeat steps (3-1) to (3-2) until the position of the new clustering center no longer changes or converges to a set range, and finally obtain k clustering centers; among them, the judgment condition for the convergence of the clustering center is as follows: Among them, δ is the convergence threshold.

6. The clustering and screening method for typical monitoring areas of urban non-point source pollution according to claim 1, characterized in that: In step (4), assume that there are n catchment area points in a certain clustering cluster, then the distance variance D of the i-th catchment area point among them i is calculated as follows: Among them, d i represents the distance from the i-th catchment subarea to its clustering center, and represents the average value of the distances from all catchment subareas in the clustering cluster to the clustering center.

7. A clustering and screening system for typical monitoring areas of urban non-point source pollution, characterized in that, including: A data acquisition module for calculating the area, construction land proportion, and slope of the catchment area according to the catchment area division, land use, and terrain of the target area; The clustering center initialization module is used to normalize the area, construction land proportion, and slope of the catchment sub - areas to form spatial point coordinates, with each catchment sub - area corresponding to a point in space; count the number N of catchment sub - areas within the set spatial distance radius for each catchment sub - area point, screen out the catchment sub - area points with N not less than the set threshold to obtain the high - density catchment sub - area set H; sort the catchment sub - area points in set H according to the construction land proportion, and select k catchment sub - area points as the initial clustering centers, where k is the number of set monitoring areas; assume there are h catchment sub - area points in the high - density catchment sub - area set H, and the 1st, 2nd, 3rd, …, k - 1th, and kth initialized clustering centers are respectively the catchment sub - area points with sorting equal to 1, h; where floor represents rounding down. A clustering module, which is used to traverse the Euclidean distances from all catchment division points to k clustering center points, divide each catchment division point into the clustering cluster with the closest distance, update the clustering center of each clustering cluster, and repeat this step until the position of the new clustering center no longer changes or converges to a set range, and finally obtain k clustering center points; And a monitoring area determination module, which is used to combine the k center points obtained by clustering, find the catchment division with the smallest variance from the clustering center in each clustering cluster, and use it as the finally selected typical monitoring area.

8. The typical monitoring area clustering and screening system for urban non-point source pollution according to claim 7, characterized in that: For the clustering center initialization module, the method for initializing the clustering center is: First, calculate the distance between any two catchment sub - area points U i (A' i , C' i , S' i ) and U j (A' j , C' j , S' j ), denoted as d(U i , U j ), and the spatial distance radius and the threshold where m is the total number of catchment sub - areas in the target area; A' i , A' j , C' i , C' j , S' i , S' j respectively represent the area, the proportion of construction land, and the slope of the i - th and j - th catchment sub - area points after normalization; Then count the number N of catchment division points within the spatial distance radius r of each catchment division point, and screen out the catchment division points with N≥M as the high-density catchment division set H; Suppose there are h catchment area points in the high-density catchment area set H. Sort the points in set H in descending order according to the proportion of construction land. Then the first, second, third, …, k-1, and kth cluster centers after initialization are the catchment area points with sorting equal to 1, h, respectively.

9. A computer system, including a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the computer program is loaded into the processor, the steps of the typical monitoring area clustering and screening method for urban non-point source pollution according to any one of claims 1-6 are implemented.

10. A computer-readable storage medium, the computer-readable storage medium stores a computer program, characterized in that, When the computer program is executed by the processor, the steps of the typical monitoring area clustering and screening method for urban non-point source pollution according to any one of claims 1-6 are implemented.

Images