Water quality monitoring method, system and device for environmental protection

Abstract

The invention relates to the technical field of water quality information monitoring, in particular to a water quality monitoring method, system and device for environmental protection. According to the method, an aggregation hierarchical clustering algorithm is utilized, monitoring nodes are clustered based on pollution degree differences among the monitoring nodes, and after a tree diagram is obtained, the tree diagram is segmented to obtain a combination of clustering clusters segmented each time. For each time of segmentation, representative nodes are screened out based on the pollution degree difference between the nodes, the feature significance of the representative nodes obtained through current segmentation is evaluated in combination with distribution of the representative nodes in the drainage pipe network and the influence degree of key areas, the reasonable degree is obtained, the optimal segmentation process is screened out, and key monitoring nodes are determined. Based on the hierarchical clustering algorithm, the key monitoring node position with remarkable characteristics can be determined based on the dynamic characteristics of the water area of the current drainage pipe network, and effective pollution information in the current drainage pipe network can be extracted based on the key monitoring node position.

Description

Technical Field

[0001] The present invention relates to the technical field of water quality information monitoring, and in particular to a water quality monitoring method, system and device for environmental protection. Background Art

[0002] For water pollution monitoring in large-scale drainage network waters, it is necessary to increase the density of monitoring points, optimize the location distribution of monitoring points, and establish a hierarchical monitoring network for effective monitoring to ensure that the pollution location and pollution information in the waters can be determined in a timely manner, and pollutants can be tracked to determine the origin and transmission path.

[0003] To build a monitoring network covering a large area of ​​water, existing technologies typically place sensors at fixed intervals or at pre-set key locations. This sensor monitoring network, designed to account for the dynamic nature of the water, results in redundant and complex information, a large number of monitoring locations, and difficulty effectively identifying pollution signatures within the water. Summary of the Invention

[0004] In order to solve the technical problems in the prior art of sensor monitoring networks for large water areas, such as excessive and redundant monitoring nodes, a large number of monitoring locations, and difficulty in effectively determining pollution characteristics in the water area, the present invention aims to provide a water quality monitoring method, system, and device for environmental protection. The technical solutions adopted are as follows: The present invention proposes a water quality monitoring method for environmental protection, the method comprising: Obtaining the pollution level at each monitoring node of the drainage network within a preset time period, the pollution level being obtained by statistics of information from multiple sensors deployed at the monitoring nodes; Performing hierarchical clustering on the monitoring nodes according to the differences in pollution levels between them to obtain a dendrogram of the hierarchical clustering; in the dendrogram, by changing the position of the horizontal dividing line, obtaining the combination of clusters for each division; For each split, the representativeness of the nodes is obtained based on the difference in pollution levels between the nodes in the clusters, and the representative nodes of each cluster are screened based on the representativeness. For each representative node, it is determined whether the node location of the cluster to which the representative node belongs coincides with the key area of ​​the drainage network, and the degree of influence of each representative node on the key area is obtained based on the coincidence situation. The degree of preference of each split is obtained based on the influence of each representative node, the distribution uniformity of the representative nodes, and the distribution concentration of the representative node and its neighboring nodes. The rationality of each segmentation is obtained according to the degree of preference and the representativeness of the representative nodes, the optimal segmentation process is screened out according to the rationality, and the representative nodes corresponding to the optimal segmentation process are used as key monitoring nodes.

[0005] Furthermore, the pollution degree includes: for each monitoring node, obtaining the COD pollution data obtained by the optical sensor of the monitoring node at a moment, the turbidity pollution data monitored by the suspended matter monitoring sensor, and the temperature data monitored by the temperature sensor, and performing weighted summation on the COD pollution data, turbidity pollution data and temperature data to obtain the pollution degree of each monitoring node at each moment.

[0006] Furthermore, obtaining the combination of clusters for each segmentation includes: Set the initial horizontal dividing line. In the dendrogram, gradually raise the initial horizontal dividing line to obtain the horizontal dividing line for each division. The number of clusters obtained by different divisions gradually increases.

[0007] Furthermore, the representative acquisition method includes: For a target node in a cluster, the pollution degree difference between the target node and other nodes in the same cluster at each moment is obtained, and all pollution degree differences obtained for the target node are accumulated to obtain the representativeness of the target node.

[0008] Furthermore, the method for determining whether the node location of the cluster to which the representative node belongs coincides with the key area of ​​the drainage network includes: For a target representative node, the non-representative nodes in the same cluster as the target representative node are the same cluster nodes, the key area closest to the target representative node is the target key area, and the representative nodes in the target key area are the representative nodes to be analyzed; The average distance between the nodes in the same cluster and the representative node to be analyzed is obtained, and the maximum width of the target key area is obtained. If the average distance is less than or equal to the maximum width, it is considered that the node position of the cluster to which the target representative node belongs coincides with the key area of ​​the drainage network; otherwise, it is judged that they do not coincide.

[0009] Furthermore, obtaining the degree of influence of each representative node on the key area according to the overlap situation includes: If it is determined that the node position of the cluster to which the target representative node belongs coincides with the key area of ​​the drainage network, then the average distance negative correlation mapping is added with a preset first parameter as the influence degree; If it is determined that the node location of the cluster does not coincide with the key area of ​​the drainage network, the degree of influence is set to the preset first parameter.

[0010] Furthermore, the method for obtaining the preference degree includes: In any segmentation process, the distance between each representative node and the nearest representative node is taken as the proximity distance, and the difference between the proximity distance of each representative node and the average proximity distance is taken as the proximity distance deviation; the distance between each node and the nearest representative node is taken as the distribution distance; the proximity distance deviations of all representative nodes and the distribution distances of all nodes are counted to obtain the node distribution characteristics; the preference degree is obtained according to the node distribution characteristics and the average influence degree of all representative nodes.

[0011] Furthermore, the method for obtaining the reasonable degree includes: For any segmentation process, the average representativeness of the representative nodes is obtained, and the product of the average representativeness and the normalized preference degree is used as the reasonableness.

[0012] The present invention also proposes a water quality monitoring system for environmental protection, comprising a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the computer program, it implements any one of the steps of the water quality monitoring method for environmental protection.

[0013] The present invention also proposes a water quality monitoring device for environmental protection, the device comprising: An experimental information acquisition module is used to obtain the pollution level at each monitoring node of the drainage network within a preset time period. The pollution level is obtained by statistically analyzing information from multiple sensors deployed at the monitoring nodes. The hierarchical clustering segmentation module is used to aggregate the monitoring nodes into hierarchical clusters according to the differences in pollution levels between the monitoring nodes, and obtain a dendrogram of the aggregated hierarchical clusters; in the dendrogram, by changing the position of the horizontal dividing line, the combination of clusters of each segmentation is obtained; The segmentation result analysis module is used to obtain the representativeness of the nodes for each segmentation based on the difference in pollution levels between the nodes in the clusters, and screen out the representative nodes of each cluster based on the representativeness; for each representative node, determine whether the node location of the cluster to which the representative node belongs coincides with the key area of ​​the drainage network, and obtain the degree of influence of each representative node on the key area based on the coincidence; and obtain the degree of preference for each segmentation based on the degree of influence of each representative node, the distribution uniformity of the representative nodes, and the distribution concentration of the representative node and its neighboring nodes; The key monitoring node determination module is used to obtain the rationality of each segmentation according to the preferred degree and the representativeness of the representative node, screen out the optimal segmentation process according to the rationality, and use the representative node corresponding to the optimal segmentation process as the key monitoring node.

[0014] The present invention has the following beneficial effects: The present invention first sets a time period, sets a plurality of monitoring nodes in the drainage network under the time period, and then identifies the key monitoring nodes through a data analysis method. The present invention utilizes an aggregated hierarchical clustering algorithm to cluster the monitoring nodes based on the difference in pollution levels between the monitoring nodes, obtains a dendrogram, and then segments it to obtain a combination of cluster clusters for each segmentation, that is, each segmentation can be regarded as changing the scale of the clustering result, and determining the monitoring nodes with the most obvious characteristics between different scales for monitoring. For each segmentation, the present invention first screens out representative nodes based on the difference in pollution levels between the nodes, and then combines the distribution of the representative nodes in the drainage network and the degree of influence of the key areas to evaluate the characteristic significance of the representative nodes obtained by the current segmentation, and then obtains a reasonable degree, screens out the optimal segmentation process and determines the key monitoring nodes. Based on the hierarchical clustering algorithm, the present invention can determine the position of the key monitoring nodes with significant characteristics based on the dynamic characteristics of the current drainage network water area through the integrated analysis of the monitoring nodes, and can extract effective pollution information in the current drainage network based on the position of the key monitoring nodes. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0016] Figure 1 This is a flow chart of a water quality monitoring method for environmental protection provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0017] To further illustrate the technical means and effects of the present invention to achieve the intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail a water quality monitoring method, system, and device for environmental protection proposed by the present invention, including its specific implementation, structure, features, and effects. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable form.

[0018] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0019] The following describes in detail a method, system and device for water quality monitoring for environmental protection provided by the present invention with reference to the accompanying drawings.

[0020] See also Figure 1 , which shows a flow chart of a water quality monitoring method for environmental protection provided by one embodiment of the present invention, the method comprising: Step S1: obtaining the pollution degree at each monitoring node of the drainage network within a preset time period, wherein the pollution degree is obtained by statistics of information from multiple sensors deployed at the monitoring nodes.

[0021] In the embodiment of the present invention, the target area is a large drainage network water area, in which monitoring nodes have been deployed using the existing monitoring node deployment method. When constructing the monitoring node network, the embodiment of the present invention sets monitoring nodes at intervals at the drainage outlet. Starting from the drainage outlet, the monitoring nodes are set and numbered in the reverse direction at fixed intervals of 200 meters, thereby obtaining a monitoring node network. The embodiment of the present invention sets the information collection interval at the monitoring node location to 30 seconds, and sets the time period for data analysis to 250 minutes, that is, 500 pollution level data are collected at each monitoring node location.

[0022] It should be noted that in this embodiment of the present invention, optical sensors, suspended solids monitoring sensors, and temperature sensors are deployed at monitoring nodes. The pollution level is determined based on the data generated by these three types of sensors. That is, within the preset time period of this embodiment of the present invention, each monitoring node will receive a time series of pollution levels, with each moment corresponding to a pollution level.

[0023] Preferably, in an embodiment of the present invention, for each monitoring node, COD pollution data obtained by the optical sensor, turbidity pollution data monitored by the suspended matter monitoring sensor, and temperature data monitored by the temperature sensor are obtained at a certain moment in time. The COD pollution data, turbidity pollution data, and temperature data are weighted and summed to obtain the pollution level of each monitoring node at each moment. As an example, in a specific implementation of the present invention, the weight of the COD pollution data is set to 0.6, the weight of the turbidity pollution data is set to 0.3, and the weight of the temperature data is set to 0.1.

[0024] It should be noted that, at the monitoring node position of the embodiment of the present invention, after each sensor collects data, range normalization processing can be performed based on the maximum and minimum values ​​under the corresponding dimension to achieve de-dimensionalization and normalization of the data.

[0025] Step S2: Performing aggregate hierarchical clustering on the monitoring nodes according to the differences in pollution levels between the monitoring nodes to obtain an aggregate hierarchical clustering dendrogram; in the dendrogram, by changing the position of the horizontal dividing line, obtaining the combination of clusters for each division.

[0026] The present invention takes into account the presence of numerous redundant nodes in existing monitoring node networks, which often represent identical pollution information. Therefore, in order to efficiently obtain characteristic data from the current drainage network, it is necessary to focus on monitoring node information with significant characteristics. Therefore, based on the concept of aggregated hierarchical clustering, the present invention performs cluster analysis on monitoring nodes to produce an aggregated hierarchical clustering dendrogram. A dendrogram is a common graphical representation method for visualizing tree-like clustering results. It displays the hierarchical relationships between clustering results, including the order and distance (dissimilarity) between samples. In a dendrogram, individual measurements are represented as nodes at the bottom, while hierarchical clusters are formed by merging individual nodes or existing lower-level clusters. A dendrogram can help visually assign related objects to clusters. Horizontal lines in the dendrogram indicate the locations where clusters are merged, while vertical lines indicate the distances between lower-level clusters when they merge into new higher-level clusters. The dendrogram can be cut using horizontal split lines at any height along the vertical axis to obtain the desired number of clusters, with lower heights corresponding to more clusters. This means that each split can yield multiple clusters.

[0027] Through hierarchical clustering, different levels can be considered clustering results of different scales. Higher levels have higher clustering scales, resulting in fewer clusters; lower levels have more specific clustering scales, resulting in more clusters. The present invention aims to screen out effective key monitoring nodes from multiple monitoring nodes. Therefore, it is not possible to analyze clustering results at only one scale. Therefore, the present invention changes the position of the horizontal dividing line in the dendrogram to obtain a combination of clusters for each segmentation. The results of each segmentation are then analyzed in subsequent steps to determine the optimal segmentation process.

[0028] It should be noted that, in the aggregation hierarchical clustering process, the difference distance of the pollution degree sequence between the monitoring nodes can be used as the dissimilarity, and the difference distance can be calculated using the Euclidean distance, which is a technique well known to those skilled in the art and will not be described in detail.

[0029] Preferably, in one embodiment of the present invention, the position of the initial horizontal dividing line is set at a position where the dissimilarity is 0.85. In the dendrogram, the initial horizontal dividing line is gradually raised to obtain a horizontal dividing line for each segmentation, and the number of clusters obtained from different segmentations gradually increases. It should be noted that setting the initial horizontal dividing line can reduce the amount of algorithm calculation, which can be regarded as setting the maximum number of key monitoring nodes. The specific position of the initial horizontal dividing line can be set according to the specific needs of the implementer and is not limited in the embodiment of the present invention.

[0030] Step S3: For each segmentation, the representativeness of the node is obtained according to the difference in pollution levels between the nodes in the clusters, and the representative nodes of each cluster are screened out according to the representativeness; for each representative node, it is determined whether the node position of the cluster to which the representative node belongs coincides with the key area of ​​the drainage network, and the degree of influence of each representative node by the key area is obtained according to the coincidence; the preferred degree of each segmentation is obtained according to the degree of influence of each representative node, the uniformity of the distribution of the representative nodes, and the distribution concentration of the representative nodes and the adjacent nodes.

[0031] Because the present invention aims to select monitoring nodes with the most obvious characteristics as key monitoring nodes, each segmentation process first determines the representativeness of the nodes based on the differences in pollution levels between nodes in each cluster. Based on this representativeness, the representative node of each cluster is selected. The representative node is the node with the most significant characteristics in that cluster.

[0032] Preferably, in the embodiment of the present invention, considering that each monitoring node can obtain a pollution level difference at each moment within a time period, a representative acquisition method includes: For a target node within a cluster, the difference in pollution levels between the target node and other nodes in the same cluster at each moment is obtained. All pollution level differences obtained for the target node are accumulated to determine the representativeness of the target node. In this embodiment of the present invention, the pollution level difference is the difference between the pollution level of the target node and the pollution levels of other nodes. In other words, the greater the pollution level difference, the more severe the pollution level of the target node and the more significant the pollution characteristics within the cluster. In other words, within each cluster, the most representative monitoring node is selected as the representative node.

[0033] After determining the representative nodes under each segmentation process, the distribution of the representative nodes in the drainage network should be analyzed. The distribution of the representative nodes under a certain segmentation process may be too concentrated, and there are also certain key areas in the drainage network. Key areas include drinking water sources, industrial discharge outlets, agricultural runoff areas, etc. The pollutant diffusion rate in these key areas is relatively fast, and the discharge frequency and water flow are also relatively fast. Therefore, it is necessary to determine the impact of the key areas and the representative nodes based on their relative distribution. The present invention aims to plan a group of representative node combinations that are relatively evenly distributed, have more surrounding monitoring nodes, and can cover key areas as key monitoring nodes. Even distribution can ensure that the key monitoring nodes cover a larger range of the drainage network; more surrounding monitoring nodes indicate that the distribution of the key monitoring nodes is more reasonable, and they are nodes with significant characteristics in a certain area; the more the key monitoring nodes cover the key areas, the more effective monitoring of the key areas can be achieved. Therefore, the embodiment of the present invention determines for each representative node whether the node location of the cluster to which the representative node belongs overlaps with a key area of ​​the drainage network. Based on the overlap, the degree of influence of each representative node on the key area is determined. The preferred degree of each segmentation is determined based on the influence of each representative node, the distribution uniformity of the representative nodes, and the distribution concentration of the representative node and its neighboring nodes. In other words, a greater preferred degree indicates a more effective division of the representative node set during the segmentation process.

[0034] Preferably, in an embodiment of the present invention, the method for determining whether the node location of the cluster to which the representative node belongs coincides with a key area of ​​the drainage network includes: For a target representative node, the non-representative nodes in the same cluster as the target representative node are the same cluster nodes, the key area closest to the target representative node is the target key area, and the representative nodes in the target key area are the representative nodes to be analyzed.

[0035] The average distance between the nodes in the same cluster and the representative node to be analyzed is obtained. It should be noted that if the target key area includes multiple representative nodes to be analyzed, only the representative node to be analyzed with the closest distance is calculated.

[0036] The smaller the average distance is, the more likely it is that the nodes in the same cluster of the target representative node are close to the target key area, which means that the representative node is more affected by the key area.

[0037] The maximum width of the target key area is obtained. If the average distance is less than or equal to the maximum width, it is considered that the node position of the cluster to which the target representative node belongs coincides with the key area of ​​the drainage network; otherwise, it is judged that they do not coincide.

[0038] Furthermore, the degree of influence of each representative node on the key area is obtained based on the overlap, including: If it is judged that the node position of the cluster to which the target representative node belongs coincides with the key area of ​​the drainage network, the preset first parameter is added to the negative correlation mapping of the average distance as the degree of influence; if it is judged that the node position of the cluster to which the target representative node belongs does not coincide with the key area of ​​the drainage network, the degree of influence is set to the preset first parameter.

[0039] That is, the degree of influence when the nodes overlap is always greater than when they don't overlap. Furthermore, the smaller the average distance, the closer the target representative node's co-cluster nodes are to the target key area, and the greater the impact of the target key area on the target representative node. Therefore, a negative correlation mapping of the average distance is necessary. It should be noted that in this embodiment of the present invention, the first parameter is set to 1. Considering that the distance between the co-cluster nodes and the representative node to be analyzed cannot be 0, the inverse of the average distance can be directly used as the negative correlation mapping result.

[0040] Preferably, in an embodiment of the present invention, the method for obtaining the preference degree includes: In any segmentation process, the distance between each representative node and the nearest representative node is used as the proximity distance, and the difference between each representative node's proximity distance and the average proximity distance is used as the proximity distance deviation. If the proximity distance deviation of each representative node is small, it indicates that the representative nodes are relatively evenly distributed in the drainage network.

[0041] The distance between each node and the nearest representative node is used as the distribution distance. If the distribution distance of all nodes is small, it means that the representative node's position has a greater reference value and is more reasonable.

[0042] The node distribution feature is obtained by calculating the neighborhood distance deviations of all representative nodes and the distribution distances of all nodes. In this embodiment of the present invention, the neighborhood distance deviations of all representative nodes and the distribution distances of all nodes are accumulated. These two sums are multiplied together, and the inverse of the product is used as the node distribution feature. In other words, a larger node distribution feature indicates a more reasonable selection of representative nodes during the segmentation process, and a higher degree of preference.

[0043] The preference level is obtained based on the node distribution characteristics and the average influence of all representative nodes. In the embodiment of the present invention, the node distribution characteristics are quantified, and the influence level is also quantified through an assignment method. The greater the node distribution characteristics, the greater the preference level should be; the greater the average influence level, the greater the preference level should be. Therefore, the product of the two is used as the preference level.

[0044] Step S4: obtaining the rationality of each segmentation according to the preferred degree and the representativeness of the representative nodes, screening out the optimal segmentation process according to the rationality, and taking the representative nodes corresponding to the optimal segmentation process as key monitoring nodes.

[0045] After determining the optimality of each segmentation, we can further combine the representativeness of the representative nodes in each segmentation process to determine the rationality of each segmentation. Specifically, the optimality represents the rationality of the distribution of the representative nodes, while the representativeness represents the rationality of the feature significance of the representative nodes. Therefore, based on the rationality, we can select the optimal segmentation process. Specifically, the segmentation process with the highest rationality is selected as the optimal segmentation process, and the corresponding representative node of the optimal segmentation process is selected as the key monitoring node.

[0046] In embodiments of the present invention, once key monitoring nodes are identified, the number of other monitoring nodes can be directly reduced, reducing monitoring costs. When tracing pollution information at key monitoring nodes, a cellular automaton model can be established to recreate the migration of pollutants within the river, achieving the purpose of tracing the source. Cellular automata (CA) are grid dynamics models in which time, space, and state are discrete, and spatial interactions and temporal causality are localized. They are capable of simulating the spatiotemporal evolution of complex systems. The specific methods are well known to those skilled in the art and will not be elaborated on here.

[0047] Preferably, in an embodiment of the present invention, for any segmentation process, the average representativeness of the representative nodes is obtained, and the product of the average representativeness and the normalized preference degree is used as the reasonableness.

[0048] In summary, the present invention utilizes an aggregated hierarchical clustering algorithm to cluster monitoring nodes based on the difference in pollution levels between the monitoring nodes, obtains a dendrogram, and then segments it to obtain a combination of clusters for each segmentation. For each segmentation, representative nodes are screened out based on the difference in pollution levels between the nodes, and the characteristic significance of the representative nodes obtained by the current segmentation is evaluated in combination with the distribution of the representative nodes in the drainage network and the degree of influence of the key areas. The reasonableness is obtained, the optimal segmentation process is screened out, and the key monitoring nodes are determined. Based on the hierarchical clustering algorithm, the present invention can determine the location of key monitoring nodes with significant characteristics based on the dynamic characteristics of the current drainage network water area through the integrated analysis of the monitoring nodes, and can extract effective pollution information in the current drainage network based on the location of the key monitoring nodes.

[0049] Based on the same inventive concept, the present invention also proposes a water quality monitoring system for environmental protection, including a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the computer program, it implements any one of the steps of the water quality monitoring method for environmental protection.

[0050] Based on the same inventive concept, a water quality monitoring device for environmental protection is provided, comprising: An experimental information acquisition module is used to obtain the pollution level at each monitoring node of the drainage network within a preset time period. The pollution level is obtained by statistically analyzing information from multiple sensors deployed at the monitoring nodes. The hierarchical clustering segmentation module is used to aggregate the monitoring nodes into hierarchical clusters according to the differences in pollution levels between the monitoring nodes, and obtain a dendrogram of the aggregated hierarchical clusters; in the dendrogram, by changing the position of the horizontal dividing line, the combination of clusters of each segmentation is obtained; The segmentation result analysis module is used to obtain the representativeness of the nodes for each segmentation based on the difference in pollution levels between the nodes in the clusters, and screen out the representative nodes of each cluster based on the representativeness; for each representative node, determine whether the node location of the cluster to which the representative node belongs coincides with the key area of ​​the drainage network, and obtain the degree of influence of each representative node on the key area based on the coincidence; and obtain the degree of preference for each segmentation based on the degree of influence of each representative node, the distribution uniformity of the representative nodes, and the distribution concentration of the representative node and its neighboring nodes; The key monitoring node determination module is used to obtain the rationality of each segmentation according to the preferred degree and the representativeness of the representative node, screen out the optimal segmentation process according to the rationality, and use the representative node corresponding to the optimal segmentation process as the key monitoring node.

[0051] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0052] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A water quality monitoring method for environmental protection, characterized in that: The method comprises: Obtaining the pollution level at each monitoring node of the drainage network within a preset time period, the pollution level being obtained by statistics of information from multiple sensors deployed at the monitoring nodes; Performing hierarchical clustering on the monitoring nodes according to the differences in pollution levels between them to obtain a dendrogram of the hierarchical clustering; in the dendrogram, by changing the position of the horizontal dividing line, obtaining the combination of clusters for each division; For each split, the representativeness of the nodes is obtained based on the difference in pollution levels between the nodes in the clusters, and the representative nodes of each cluster are screened based on the representativeness. For each representative node, it is determined whether the node location of the cluster to which the representative node belongs coincides with the key area of ​​the drainage network, and the degree of influence of each representative node on the key area is obtained based on the coincidence situation. The degree of preference of each split is obtained based on the influence of each representative node, the distribution uniformity of the representative nodes, and the distribution concentration of the representative node and its neighboring nodes. The rationality of each segmentation is obtained according to the degree of preference and the representativeness of the representative nodes, the optimal segmentation process is screened out according to the rationality, and the representative nodes corresponding to the optimal segmentation process are used as key monitoring nodes.

2. The water quality monitoring method for environmental protection according to claim 1, characterized in that: The pollution degree includes: for each monitoring node, obtaining the COD pollution data obtained by the optical sensor, the turbidity pollution data monitored by the suspended matter monitoring sensor, and the temperature data monitored by the temperature sensor at the monitoring node at a certain moment, and performing weighted summation on the COD pollution data, turbidity pollution data and temperature data to obtain the pollution degree of each monitoring node at each moment.

3. The water quality monitoring method for environmental protection according to claim 1, characterized in that: The method of obtaining a combination of clusters for each segmentation includes: Set the initial horizontal dividing line. In the dendrogram, gradually raise the initial horizontal dividing line to obtain the horizontal dividing line for each division. The number of clusters obtained by different divisions gradually increases.

4. The water quality monitoring method for environmental protection according to claim 1, characterized in that: The representative acquisition methods include: For a target node in a cluster, the pollution degree difference between the target node and other nodes in the same cluster at each moment is obtained, and all pollution degree differences obtained for the target node are accumulated to obtain the representativeness of the target node.

5. The water quality monitoring method for environmental protection according to claim 1, characterized in that: Methods for determining whether the node location of the cluster to which the representative node belongs coincides with the key area of ​​the drainage network include: For a target representative node, the non-representative nodes in the same cluster as the target representative node are the same cluster nodes, the key area closest to the target representative node is the target key area, and the representative nodes in the target key area are the representative nodes to be analyzed; The average distance between the nodes in the same cluster and the representative node to be analyzed is obtained, and the maximum width of the target key area is obtained. If the average distance is less than or equal to the maximum width, it is considered that the node position of the cluster to which the target representative node belongs coincides with the key area of ​​the drainage network; otherwise, it is judged that they do not coincide.

6. The water quality monitoring method for environmental protection according to claim 5, characterized in that: The method of obtaining the degree of influence of each representative node on the key area according to the overlap situation includes: If it is determined that the node position of the cluster to which the target representative node belongs coincides with the key area of ​​the drainage network, then the average distance negative correlation mapping is added with a preset first parameter as the influence degree; If it is determined that the node location of the cluster does not coincide with the key area of ​​the drainage network, the degree of influence is set to the preset first parameter.

7. The water quality monitoring method for environmental protection according to claim 1, characterized in that: The method for obtaining the preference degree includes: In any segmentation process, the distance between each representative node and the nearest representative node is taken as the proximity distance, and the difference between the proximity distance of each representative node and the average proximity distance is taken as the proximity distance deviation; the distance between each node and the nearest representative node is taken as the distribution distance; the proximity distance deviations of all representative nodes and the distribution distances of all nodes are counted to obtain the node distribution characteristics; the preference degree is obtained according to the node distribution characteristics and the average influence degree of all representative nodes.

8. The water quality monitoring method for environmental protection according to claim 1, characterized in that: Methods for obtaining the reasonable degree include: For any segmentation process, the average representativeness of the representative nodes is obtained, and the product of the average representativeness and the normalized preference degree is used as the reasonableness.

9. A water quality monitoring system for environmental protection, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the water quality monitoring method for environmental protection as described in any one of claims 1 to 8 are implemented.

10. A water quality monitoring device for environmental protection, characterized in that: The device comprises: An experimental information acquisition module is used to obtain the pollution level at each monitoring node of the drainage network within a preset time period. The pollution level is obtained by statistically analyzing information from multiple sensors deployed at the monitoring nodes. The hierarchical clustering segmentation module is used to aggregate the monitoring nodes into hierarchical clusters according to the differences in pollution levels between the monitoring nodes, and obtain a dendrogram of the aggregated hierarchical clusters; in the dendrogram, by changing the position of the horizontal dividing line, the combination of clusters of each segmentation is obtained; The segmentation result analysis module is used to obtain the representativeness of the nodes for each segmentation based on the difference in pollution levels between the nodes in the clusters, and screen out the representative nodes of each cluster based on the representativeness; for each representative node, determine whether the node location of the cluster to which the representative node belongs coincides with the key area of ​​the drainage network, and obtain the degree of influence of each representative node on the key area based on the coincidence; and obtain the degree of preference for each segmentation based on the degree of influence of each representative node, the distribution uniformity of the representative nodes, and the distribution concentration of the representative node and its neighboring nodes; The key monitoring node determination module is used to obtain the rationality of each segmentation according to the preferred degree and the representativeness of the representative node, screen out the optimal segmentation process according to the rationality, and use the representative node corresponding to the optimal segmentation process as the key monitoring node.

Images