Modularity maximization-based polder area river water quality monitoring partition dynamic defining method

Abstract

The invention discloses a modularity maximization-based polder area river water quality monitoring partition dynamic defining method, relates to the technical field of water environment monitoring and intelligent water affair technologies, and solves the technical problems of insufficient data representativeness, partition static stiffness and verification means deficiency in polder area river water quality monitoring. According to the technical scheme, the method is characterized in that a hydraulic model and a modularity algorithm are combined, partition boundaries better fit actual water flow paths, and precision is improved; redundant monitoring points are reduced, the number of distributed points is reduced by 20%, and the cost is reduced; the dynamic adjustment mechanism deals with sudden hydrological changes such as extreme rainfall and pump gate regulation and control, and the real-time performance is enhanced.

Description

Technical Field

[0001] The present application relates to the technical fields of water environment monitoring and intelligent water conservancy, and particularly relates to a dynamic definition method for the water quality monitoring zones of polder rivers based on modularity maximization. Background Art

[0002] With the acceleration of the urbanization process and the intensification of climate change, the water quality monitoring of polder rivers faces increasingly complex challenges. There are significant deficiencies in the existing technologies for the design of water quality monitoring networks and the division of drainage areas, which are specifically manifested as follows:

[0003] (1) Unscientific layout of monitoring points: Traditional methods rely on uniform point layout or empirical judgment, without considering the dynamic changes in the water flow direction in the polder area, resulting in data that cannot reflect the true local water quality conditions.

[0004] (2) Static zoning: Most of the existing drainage area divisions are based on fixed terrain or pipe network topologies and cannot adapt to the changes in water flow paths caused by pump and sluice regulation, rainfall, etc.

[0005] (3) Lack of verification means: There is a lack of quantitative indicators to evaluate the representativeness of monitoring points and the rationality of zoning, making it difficult to optimize the point layout scheme.

[0006] And there are three core contradictions in the current polder water quality monitoring technology:

[0007] (1) The contradiction between dynamic hydrological conditions and static zoning models: Traditional methods cannot capture the sudden changes in flow direction caused by pump and sluice regulation, extreme rainfall, etc., resulting in the "lag failure" of the monitoring network.

[0008] (2) The contradiction between data volume requirements and cost control: High-precision monitoring relies on dense point layout, but the construction and maintenance costs far exceed the affordability of local finances.

[0009] (3) The contradiction between complex river network structures and simplified analysis tools: Existing algorithms (such as Dijkstra shortest path, Kriging interpolation) are difficult to handle complex correlation networks driven by hydraulic weights.

[0010] With the development of intelligent water conservancy and digital twin technologies, the industry urgently needs a monitoring zoning method that is low-cost, adaptive, and can be quantitatively verified. It should be able to respond to changes in hydrological conditions in real time, adjust the coverage of the monitoring network, and quantify the data representativeness through hydrodynamic models and statistical indicators, and meet lightweight calculations, adapt to the resource limitations of edge computing devices, and achieve minute-level response. Summary of the Invention

[0011] This application provides a dynamic boundary definition method for water quality monitoring in polder river channels based on modularity maximization. Its technical objective is to achieve adaptive adjustment of the boundary of the water quality monitoring area in the polder river channel, accurately quantify the coverage range of monitoring points, and ensure dynamic verification of the spatial representativeness of water quality data.

[0012] The above technical objective of this application is achieved through the following technical solutions:

[0013] A dynamic boundary definition method for water quality monitoring in polder river channels based on modularity maximization includes:

[0014] Step S1: Construct a river network topological model with the start and end sections of the river reach and the river confluence points as nodes and the river channels as edges. Define the edge weights of the river network topological model through head loss, and construct a dynamic partition model driven by hydraulic correlation according to the edge weights.

[0015] Step S2: Use the modularity maximization algorithm to divide the partitions to obtain the final partitions.

[0016] Step S3: Calculate the pollutant concentration in the final partitions through a water quality diffusion model, and take the highest point of the pollutant concentration as the water quality monitoring point.

[0017] Step S4: Conduct a correlation analysis on the water quality within the partitions and adjacent partitions according to the pollutant concentration, and further divide the partitions according to the correlation analysis.

[0018] Furthermore, in the above step S1, the head loss is expressed as:

[0019]

[0020] where h f,ij represents the head loss of the river channel section e ij ; Q ij represents the river channel flow of the river channel section e ij , that is, the amount of water passing through the river channel cross-section per unit time; A ij represents the river channel cross-sectional area of the river channel section e ij , A ij = bottom width × water depth + slope area; R ij represents the hydraulic radius of the river channel section e ij ; P ij represents the wetted perimeter of the river channel section e ij , that is, the perimeter of the water flow in contact with the river channel; n ij represents the Manning roughness coefficient of the river channel section e ij , reflecting the roughness of the inner wall of the river channel; v ij represents the water flow velocity of the river channel section e ij ; s ij represents the river channel section eij The length of the river channel; e ij Denote node v i and node v j The river channel segment between them, v i and v j Both represent the confluence points of the river channels;

[0021] Then the edge weight is expressed as:

[0022]

[0023] Among them, s ij Denote the river channel segment e ij The length of.

[0024] Furthermore, in step S1, the construction of the dynamic partition model driven by hydraulic correlation according to the edge weight includes:

[0025] Set each node as an independent initial partition, and then calculate the initial modularity Q of each initial partition, expressed as:

[0026]

[0027] Among them, m represents the total network weight, and Respectively represent the in-degree weight and out-degree weight of node v i , and Respectively represent the in-degree weight and out-degree weight of node v j , δ(c i , c j ) represents the indicator function. If node v i and v j Belong to the same partition, then δ(c i , c j ) = 1, otherwise δ(c i , c j ) = 0.

[0028] Furthermore, in the said step S2, using the modularity maximization algorithm to divide the partition to obtain the final partition, includes:

[0029] Step S21: Calculate the modularity gain ΔQ when each node moves to the partition where the neighbor node is located. If ΔQ > 0, find the node and partition that make ΔQ the largest, move the node to that partition, and update the partition; the modularity gain ΔQ is expressed as:

[0030]

[0031] Among them, ∑ in represents the sum of the weights of the internal edges of the partition; ∑ tot represents the sum of the weights of all edges of the partition; k i represents the weighted degree of node v i ; k i,in represents the sum of the connection weights between node v i and the internal nodes of the partition;

[0032] Step S22: Merge all the updated partitions into a supernode, construct a new river network topology model, and repeatedly optimize the new river network topology model according to Step S21 until the modularity Q converges to obtain the final partition.

[0033] Furthermore, in the said Step S3, the water quality diffusion model is expressed as:

[0034]

[0035] Among them, C is the pollutant concentration, representing the content of pollutants in the water body; t is the time, representing the length of the simulated time; v is the flow velocity; x is the spatial coordinate, representing the position in the direction of the river channel length; D is the diffusion coefficient, representing the diffusion ability of pollutants in the water body.

[0036] Furthermore, in the said Step S4, the variance σ 2 of the pollutant concentrations of all water quality monitoring points within the same partition is expressed as:

[0037]

[0038] Among them, σ 2 is the variance of the pollutant concentration, representing the degree of dispersion of the concentrations of water quality monitoring points within the partition; C p represents the pollutant concentration of the p-th water quality monitoring point; represents the average value of the concentrations of water quality monitoring points within the partition, N represents the number of water quality monitoring points within the partition;

[0039] When σ 2 ≥2.0, the partition is further subdivided, including: extracting the current partition, and applying the modularity maximization algorithm to the current partition again for secondary partitioning until the variance σ 2 of the pollutant concentrations of all water quality monitoring points within the same partition <2.0.

[0040] Furthermore, in the said Step S4, for adjacent partitions A and B, the correlation coefficient r of their pollutant concentrations is expressed as:

[0041]

[0042] Among them, CA,p Denote the pollutant concentration at the p-th water quality monitoring point in sub-region A; C B,q Denote the pollutant concentration at the q-th water quality monitoring point in sub-region B; Denote the average value of the pollutant concentrations at all water quality monitoring points within sub-region A; Denote the average value of the pollutant concentrations at all water quality monitoring points within sub-region B;

[0043] When r ≥ 0.8 for adjacent sub-regions, merge the adjacent sub-regions.

[0044] Furthermore, 0.01 ≤ n ij ≤ 0.05.

[0045] Furthermore, the river confluence point is at least a double-fork confluence point.

[0046] The beneficial effects of this application are as follows: The dynamic boundary definition method for water quality monitoring sub-regions in polder river networks based on modularity maximization in this application constructs a river network topological model with the start and end sections of river reaches and river confluence points as nodes and rivers as edges, defines the edge weights of the river network topological model through head loss, constructs a dynamic sub-region model driven by hydraulic correlation according to the edge weights; uses the modularity maximization algorithm to divide the sub-regions to obtain the final sub-regions; conducts correlation analysis on the water quality within and adjacent to the sub-regions according to the final sub-regions, and further divides the sub-regions based on the correlation analysis. By combining the hydraulic model with the modularity algorithm, the sub-region boundaries are more in line with the actual water flow paths, improving the accuracy; reducing redundant monitoring points, with the number of layout points reduced by 20%, reducing the cost; the dynamic adjustment mechanism can cope with sudden hydrological changes such as extreme rainfall and pump gate regulation, enhancing the real-time performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic diagram of the structure of the initial river network topological model in the embodiment of this application;

[0048] Figure 2 It is a schematic diagram of the structure of the river network topological model after sub-region optimization in the embodiment of this application;

[0049] Figure 3 It is a schematic diagram of the optimal water quality monitoring points obtained according to the method of this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] The technical solution of this application will be described in detail below with reference to the accompanying drawings.

[0051] The dynamic boundary definition method for water quality monitoring sub-regions in polder river networks based on modularity maximization in this application includes:

[0052] Step S1: Construct a river network topology model with the starting and ending segments of the river reach and the river channel intersection points as nodes and the river channels as edges. Define the edge weights of the river network topology model through head loss, and construct a dynamic partitioning model driven by hydraulic correlation based on the edge weights. In the embodiment of the present application, the initial river network topology model is as shown in Figure 1 shown.

[0053] Specifically, take the river channel intersection points (such as three-way junctions, four-way junctions) as nodes, denoted as V = {v1, v2, …, v N}, where N is the total number of nodes. Each node v i includes attributes such as coordinates (x i , y i ), flow rate Q i , water depth h i , degree d i , etc. The degree d i represents the number of river channels intersecting at the river channel intersection point.

[0054] Take the river channel segments as edges, denoted as E = {e ij}, where e ij represents the river channel segment between nodes v i and v j . The edge weight w ij is defined by the reciprocal of the head loss, and the edge weight is used to characterize the hydraulic correlation strength.

[0055] The head loss is expressed as:

[0056]

[0057] where h f,ij represents the head loss of the river channel segment e ij ; Q ij represents the river channel flow rate (m ij / s) of the river channel segment e 3 , that is, the amount of water passing through the river channel cross-section per unit time; A ij represents the river channel cross-section area (m ij ) of the river channel segment e 2 , A ij = bottom width × water depth + slope area; R ij represents the hydraulic radius (m) of the river channel segment e ij ; P ij represents the wetted perimeter of the river channel segment e ij , that is, the perimeter of the water flow in contact with the river channel; n ij represents the Manning roughness coefficient of the river channel segment e ij , reflecting the roughness of the inner wall of the river channel, and the value range is usually from 0.01 (smooth cement) to 0.05 (natural river channel); v ij represents the river channel segment eij The water flow velocity (m / s); s ij Indicates the river channel section e ij The length of; e ij Indicates the node v i And the node v j The river channel section between, v i And v j Both represent river channel confluence points.

[0058] Then the edge weight is expressed as:

[0059]

[0060] Among them, s ij Indicates the river channel section e ij The length of.

[0061] The construction of the dynamic partitioning model driven by hydraulic correlation according to the edge weight includes:

[0062] Set each node as an independent initial partition, and then calculate the initial modularity Q of each initial partition, which is expressed as:

[0063]

[0064] Among them, m represents the total network weight, And Respectively represent the in-degree weight and out-degree weight of the node v i The in-degree weight and out-degree weight of, And Respectively represent the in-degree weight and out-degree weight of the node v j The in-degree weight and out-degree weight of, δ(c i , c j ) represents the indicator function. If the nodes v i And v j Belong to the same partition, then δ(c i , c j ) = 1, otherwise δ(c i , c j ) = 0.

[0065] Step S2: Use the modularity maximization algorithm to partition the partition to obtain the final partition.

[0066] Furthermore, the step S2 includes:

[0067] Step S21: Calculate the modularity gain ΔQ when each node moves to the partition where the neighbor node is located. If ΔQ>0, find the node and partition that make ΔQ the largest, move the node to that partition, and update the partition; the modularity gain ΔQ is expressed as:

[0068]

[0069] Among them, ∑ in represents the sum of the weights of the internal edges of the partition; ∑ tot represents the sum of the weights of all edges of the partition; k i represents the weighted degree of node v i ; k i,in represents the sum of the connection weights between node v i and the internal nodes of the partition.

[0070] Step S22: Merge all the updated partitions into a supernode, construct a new river network topology model, and repeatedly optimize the new river network topology model according to Step S21 until the modularity Q converges to obtain the final partition.

[0071] Furthermore, if the change in Q is less than 0.001, then Q is considered to have converged.

[0072] Step S3: Calculate the pollutant concentration of the final partition through a water quality diffusion model, and take the highest point of the pollutant concentration as the water quality monitoring point.

[0073] Furthermore, the water quality diffusion model is expressed as:

[0074]

[0075] Among them, C is the pollutant concentration (mg / L), representing the content of pollutants in the water body; t is the time (s), representing the length of the simulation time; v is the flow velocity (m / s), calculated by the Manning formula; x is the spatial coordinate (m), representing the position in the river length direction; D is the diffusion coefficient (m 2 / s), representing the diffusion ability of pollutants in the water body.

[0076] Step S4: Conduct a correlation analysis on the water quality within and adjacent to the partition according to the pollutant concentration, and further divide the partition according to the correlation analysis.

[0077] Furthermore, in Step S4, the concentration variance σ 2 of all water quality monitoring points within the same partition is expressed as:

[0078]

[0079] Among them, σ 2 is the pollutant concentration variance, representing the dispersion degree of the pollutant concentrations of the water quality monitoring points within the partition; C p represents the pollutant concentration of the p-th water quality monitoring point; represents the average value of the pollutant concentrations of the water quality monitoring points within the partition, N represents the number of water quality monitoring points in the partition.

[0080] When σ 2 ≥ 2.0, the partition is further subdivided, including: extracting the current partition, and applying the modularity maximization algorithm to the current partition again for secondary partitioning until the variance σ of the pollutant concentrations of all water quality monitoring points in the same partition 2 < 2.0.

[0081] Furthermore, analyze the relationship between the modularity Q and σ through the Spearman rank correlation coefficient ρ, expressed as: 2

[0082]

[0083] Among them, ρ is the Spearman rank correlation coefficient (range [-1, 1]), indicating the correlation between the modularity Q and the concentration variance σ 2 ; the closer ρ is to -1, the lower the water quality variance corresponding to the high modularity partition (strong negative correlation); d p represents the rank difference between the modularity and the concentration variance, and the calculation formula is N is the number of partitions.

[0084] Furthermore, in step S4, for adjacent partitions A and B, the correlation coefficient r of their pollutant concentrations is expressed as:

[0085]

[0086] Among them, C A,p represents the pollutant concentration of the p-th water quality monitoring point in partition A; C B,q represents the pollutant concentration of the q-th water quality monitoring point in partition B; represents the average value of the pollutant concentrations of all water quality monitoring points in partition A; represents the average value of the pollutant concentrations of all water quality monitoring points in partition B.

[0087] The closer r is to 1, the higher the water quality correlation between adjacent partitions; the closer r is to -1, the more the water quality of adjacent partitions changes in the opposite direction; when r = 0, it means that there is no correlation between the changes of adjacent partitions s.

[0088] When r ≥ 0.8 for adjacent partitions, merge the adjacent partitions.

[0089] The following are specific embodiments:

[0090] (1) Input data

[0091] River channel data: 133 river channels in Wusongwei, Kunshan City, with the cross-section slope standardized (taking 0.2), the river width taking the average width, and the bank height based on the designed bank height. ​

[0092] The Manning coefficient n = 0.03, and the initial flow rate Q = 3.2 m 3 / s.

[0093] (2) Calculation process

[0094] 1) Construct a river network topological model

[0095] Taking the start and end sections and intersections of river reaches as nodes and the river channels as edges, a river network topological model is constructed. Among them, each node has a unique node_id and coordinates (x, y), and each river channel has a start_node, an end_node, and attributes width (width), bank_height (bank height), and slope (slope). There are 249 nodes and 288 edges in total.

[0096] Node example:

[0097] v1 = {node id : 1, x: 585562.5375, y: 3469150.8367}

[0098] v2 = {node id : 2, x: 585939.3901, y: 3469032.1087}

[0099] v3 = {node id : 3, x: 585952.2348, y: 3469386.1371}

[0100] Edge example:

[0101] e 12 = {start_node: 1, end_node: 2, width: 31.47, bank_height: 4.0, slope: 0.2}

[0102] e 23 = {start_node: 2, end_node: 3, width: 30.76, bank_height: 4.0, slope: 0.2}

[0103] Record the node and river reach edge information as G = (V, e); where V is the node set: V = {v1, v2, v3,...}, and E is the edge set: E = {e 12 , e 23 ,...}.

[0104] 2) Calculate the edge weights

[0105] Taking e 12 as an example, the cross-sectional area A of the water flow:

[0106] A = width × depth = 31.47 × 2.4 = 75.528 m 2

[0107] The water depth depth = 2.4 meters.

[0108] Wetted perimeter P:

[0109] P = 2 × depth + width = 2 × 2.4 + 31.47 = 36.27 m

[0110] Hydraulic radius R:

[0111]

[0112] Flow velocity V:

[0113]

[0114] Calculation:

[0115]

[0116] Channel length L:

[0117]

[0118] Head loss h f :

[0119]

[0120] where g = 9.81 m / s 2 is the acceleration due to gravity.

[0121] Edge weight w 12 :

[0122] w 12 = h f ≈ 13.7

[0123] 3) Calculate the initial modularity

[0124] The modularity Q measures the quality of the network partitioning, and the calculation formula is: where m is the total network weight: k i is the weighted degree of node i: k i = Σ j w ij ; δ(c i , c j ) is the indicator function, which takes 1 if nodes i and j belong to the same partition, and 0 otherwise.

[0125] Assume that the initial partition is that each node is independently partitioned, then δ(ci , c j ) = 1 if and only if i = j, otherwise 0.

[0126] Initial modularity = -0.007415266420419451.

[0127] 4) Modularity optimization

[0128] Maximize the modularity Q by continuously adjusting the partitioning of nodes.

[0129] Initially, each node belongs to an independent partition: partition = {v1: 1, v2: 2, v3: 3, …}.

[0130] For each node i, calculate the modularity gain ΔQ when moving it to the partition where its neighbor nodes are located.

[0131] Find the node and partition that maximize ΔQ, move the node to that partition, and update the partition.

[0132] Repeat updating the partition until the modularity Q converges (i.e., the change in ΔQ is less than the threshold 0.0001).

[0133] Merge each partition into a supernode, the edge weight between supernodes is the sum of the weights of all edges between the two partitions in the original network, and use the merged supernode network to recalculate the modularity Q.

[0134] According to the Louvain algorithm, 16 partitions are finally divided, modularity Q = 0.8183. The partition diagram is as Figure 2 shown.

[0135] 5) Determine water quality monitoring points

[0136] Based on the water quality monitoring data of Wusongwei, there are only two places where the ammonia nitrogen in the water quality of the inner - river channels exceeds the standard (≥2.0 mg / L); for other river nodes, low - concentration data (≤0.5 mg / L) are assumed, and the worst - quality points in each partition are calculated as monitoring points using the advection - diffusion model.

[0137] 6) Water quality correlation analysis

[0138] Add pollutants (ammonia nitrogen 5 mg / L) at the 16 monitoring points, and after simulating for 24 hours, σ in each partition 2 The results are shown in Table 1:

[0139] Table 1

[0140] Partition number <![CDATA[σ 2 (mg 2 / L 2 )]]> Partition number <![CDATA[σ 2 (mg 2 / L 2 )]]> Partition 0 0.0167 Partition 8 0.0067 Partition 1 0.0017 Partition 9 0.0141 Partition 2 0.0052 Partition 10 0.008 Partition 3 0.0017 Partition 11 0.0001 Partition 4 0.0204 Partition 12 0.0144 Partition 5 0.0064 Partition 13 0.0001 Partition 6 0.0476 Partition 14 0.0019 Partition 7 0.0077 Partition 15 0.0002

[0141] As can be seen from Table 1, σ in each partition 2 < 2 mg / L.

[0142] The correlation coefficients of pollutant concentrations in adjacent partitions are shown in Table 2 as follows:

[0143] Table 2

[0144]

[0145]

[0146] As can be seen from Table 1, the correlation coefficient r of pollutant concentrations in adjacent partitions is less than 0.8.

[0147] The output result is the location of the river channel monitoring point, as Figure 3 shown.

[0148] The above are exemplary embodiments of the present application, and the protection scope of the present application is defined by the claims and their equivalents.

Claims

1. A dynamic definition method for the water quality monitoring zoning of polder rivers based on maximizing modularity, characterized in that Including: Step S1: Construct a river network topology model with the start and end sections of the river reach and the river channel confluence points as nodes and the river channels as edges. Define the edge weights of the river network topology model through head loss, and construct a dynamic partition model driven by hydraulic correlation according to the edge weights. Step S2: Use the modularity maximization algorithm to divide the partitions to obtain the final partitions. Step S3: Calculate the pollutant concentration of the final partitions through a water quality diffusion model, and take the highest point of the pollutant concentration as the water quality monitoring point. Step S4: Conduct a correlation analysis on the water quality within and adjacent to the partitions according to the pollutant concentration, and further divide the partitions based on the correlation analysis.

2. The dynamic zoning method for monitoring the water quality of river channels in polder areas as described in claim 1, characterized in that In the said Step S1, the head loss is expressed as: Among them, h f,ij represents the head loss of the river channel segment e ij ; Q ij represents the river channel flow of the river channel segment e ij , that is, the amount of water passing through the river channel cross-section per unit time; A ij represents the river channel cross-sectional area of the river channel segment e ij , A ij = bottom width × water depth + slope area; R ij represents the hydraulic radius of the river channel segment e ij ; P ij represents the wetted perimeter of the river channel segment e ij , that is, the perimeter of the water flow in contact with the river channel; n ij represents the Manning roughness coefficient of the river channel segment e ij , reflecting the roughness of the inner wall of the river channel; v ij represents the water flow velocity of the river channel segment e ij ; s ij represents the length of the river channel segment e ij ; e ij represents the river channel segment between node v i and node v j , v i and v j both represent river channel confluence points; Then the edge weight is expressed as: Among them, s ij represents the length of river section e ij .

3. The dynamic definition method for the water quality monitoring sub-region of the river in the polder area as described in claim 2, wherein In Step S1, the construction of the dynamic partition model driven by hydraulic correlation according to the edge weights includes: Set each node as an independent initial partition, and then calculate the initial modularity Q of each initial partition, expressed as: Among them, m represents the total network weight, and respectively represent the in-degree weight and out-degree weight of node v i . and respectively represent the in-degree weight and out-degree weight of node v j . δ(c i , c j ) represents the indicator function. If nodes v i and v j belong to the same partition, then δ(c i , c j ) = 1; otherwise, δ(c i , c j ) = 0.

4. The dynamic definition method for the water quality monitoring zones of the river channels in the polder area according to claim 3, characterized in that, In the said Step S2, using the modularity maximization algorithm to divide the partitions to obtain the final partitions includes: Step S21: Calculate the modularity gain ΔQ when each node moves to the partition where the neighbor node is located. If ΔQ>0, find the node and partition that make ΔQ the largest, move the node to that partition, and update the partition. The modularity gain ΔQ is expressed as: Among them, ∑ in represents the sum of the weights of the internal edges of the partition; ∑ tot represents the sum of the weights of all edges of the partition; k i represents the weighted degree of node v i ; k i,in represents the sum of the connection weights between node v i and the internal nodes of the partition; Step S22: Merge all the updated partitions into super nodes, construct a new river network topology model, and repeat the optimization of the new river network topology model according to Step S21 until the modularity Q converges to obtain the final partitions.

5. The dynamic delimitation method for water quality monitoring zones of river channels in polder areas as described in claim 4, characterized in that, In the said Step S3, the water quality diffusion model is expressed as: Where, C is the pollutant concentration, representing the content of pollutants in the water body; t is the time, representing the length of the simulation time; v is the flow velocity; x is the spatial coordinate, representing the position in the river channel length direction; D is the diffusion coefficient, representing the diffusion ability of pollutants in the water body.

6. The dynamic zoning definition method for water quality monitoring of river channels in polder areas as described in claim 5, wherein In the step S4, the variance σ of the pollutant concentrations of all water quality monitoring points within the same partition 2 is expressed as: Among them, σ 2 is the variance of pollutant concentration, representing the degree of dispersion of the concentrations of water quality monitoring points within the sub-region; C p represents the pollutant concentration at the p-th water quality monitoring point; represents the average value of the concentrations of water quality monitoring points within the sub-region, represents the number of water quality monitoring points within the sub-region; When σ 2 ≥ 2.0, the partition is further subdivided, including: extracting the current partition, and applying the modularity maximization algorithm to the current partition again for secondary partitioning until the variance σ 2 < 2.0 for the pollutant concentrations of all water quality monitoring points within the same partition.

7. The dynamic delimitation method for the water quality monitoring zones of the river channels in the polder area according to claim 6, wherein, In the said Step S4, for adjacent partitions A and B, the correlation coefficient r of their pollutant concentrations is expressed as: Among them, C A,p represents the pollutant concentration at the p-th water quality monitoring point in zone A; C B,q represents the pollutant concentration at the q-th water quality monitoring point in zone B; represents the average value of the pollutant concentrations at all water quality monitoring points within zone A; represents the average value of the pollutant concentrations at all water quality monitoring points within zone B; When r≥0.8 for adjacent partitions, merge the adjacent partitions.

8. The dynamic definition method for water quality monitoring zoning in polder river channels according to claim 7, characterized in that 0.01≤n ij ≤0.05。 9. The dynamic definition method for the water quality monitoring sub-region of the river in the polder area as described in claim 1, characterized in that, The river channel confluence point is at least a double-fork confluence point.

Images