Multi-lid scenario runoff regulation effect identification method based on two-order clustering of som and k-means

Abstract

This application provides a method for identifying the runoff regulation effect under multiple LID scenarios based on two-order clustering using SOM and K-means. The method includes: acquiring basic data of the target area; calculating runoff regulation data for each sub-catchment of the target area based on the basic data; inputting the runoff regulation data into a SOM neural network to obtain a reasonable range of cluster numbers and the final cluster centers for each cluster; then using the results of SOM clustering as the cluster numbers and initial cluster centers for K-means clustering to obtain an optimized number of clusters. This allows for spatial clustering analysis of the runoff control index reduction rate under different LID regulation scenarios, and visualization of the results to compare the effects of design schemes for different control objectives. This application can reduce dimensionality and display spatially under multiple sub-catchments, multiple hydrological control target indicators, multiple modification scenario design schemes, and multiple rainfall scenarios, enabling effect comparison of design schemes for different control objectives and providing a reference for selecting schemes to adapt to different construction requirements.

Description

Technical Field

[0001] This invention belongs to the field of environmental information management technology, and relates to a method for identifying the effect of runoff regulation, particularly a method for identifying the effect of runoff regulation in multiple LID scenarios based on two-order clustering of SOM and K-means. Background Technology

[0002] In recent years, global warming and rapid urbanization have led to frequent natural disasters, with urban flooding and water shortages becoming increasingly serious problems. Therefore, building sponge cities can reduce the impact of urban flooding, waterlogging, and other natural disasters.

[0003] A sponge city, like a sponge, possesses excellent "elasticity" in adapting to environmental changes and responding to natural disasters. When it rains, it can absorb, store, infiltrate, and purify water. When necessary, it can "release" the stored water for use. Building sponge cities fully utilizes natural ecological functions and artificial intervention to effectively control rainwater runoff, achieving a city development model of natural accumulation, natural infiltration, and natural purification. This is beneficial for restoring urban water ecology, conserving water resources, enhancing urban flood control capabilities, expanding effective investment in public goods, improving the quality of new urbanization, and promoting harmonious development between humans and nature. The greatest significance of building sponge cities lies in solving a series of problems caused by paved roads, particularly urban flooding, reducing groundwater, and mitigating the urban heat island effect. Building sponge cities can also protect and restore the urban ecological environment, providing a better living environment.

[0004] LID (Low Impact Development) technology, also known as stormwater management or non-point source pollution control technology, controls stormwater runoff and pollution through decentralized, small-scale source control, bringing developed areas closer to the natural hydrological cycle. LID technologies include: urban natural drainage systems, rain gardens, eco-retention swales, green streets, permeable pavements, eco-roofs, and rainwater harvesting systems. LID technologies can effectively utilize landscape space to treat non-point source pollution and control stormwater runoff. They can also reduce stormwater runoff by 30-99%, delay peak runoff by 5-20 minutes, effectively remove pollutants such as phosphorus, nitrogen, grease, and heavy metals from stormwater runoff, reduce acid rain, save energy, reduce rainwater reuse costs, beautify the environment, create comfortable living spaces, and alleviate pressure on municipal drainage systems.

[0005] Currently, in the process of building sponge cities with LID technology at its core, it is required to reasonably delineate control zones, clarify the current rainwater control capacity of each zone, and formulate rainwater runoff control targets for each zone. However, there is currently no method for visually classifying and identifying the runoff control effects of different LID facilities on different underlying surfaces. Summary of the Invention

[0006] In view of the shortcomings of the prior art described above, the purpose of this application is to provide a method for identifying the runoff control effect of multiple LID scenarios based on SOM and K-means two-order clustering, in order to solve the problem that there is no method for visual classification and identification of the runoff control effect of different LID facilities for different underlying surfaces in the process of implementing the runoff control effect identification technology of multiple LID scenarios.

[0007] To achieve the above and other related objectives, in a first aspect, this application provides a method for identifying the runoff regulation effect in multiple LID scenarios based on two-stage clustering of SOM and K-means, comprising the following steps: acquiring basic data of the target area; calculating runoff regulation data of each sub-catchment of the target area based on the basic data; performing two-stage clustering on the runoff regulation data to obtain the correlation between different variables at different levels, so as to compare the effects of design schemes for different control objectives.

[0008] In one implementation of the first aspect, the basic data includes any one or more combinations of: target area elevation, underlying surface, drainage system and facility data, regional water system data, and hydrological and rainfall data.

[0009] In one implementation of the first aspect, calculating the runoff control data of each sub-catchment of the target area based on the basic data includes the following steps: normalizing the basic data to obtain storm runoff management data; inputting the storm runoff management data into a storm flood management model to calculate the impervious surface ratio of the target area, and dividing the target area into several sub-catchments and the impervious surface ratio of the sub-catchments; calculating the runoff control data of each sub-catchment based on the sub-catchments and the impervious surface ratio of the sub-catchments; wherein the runoff control data includes: total outflow runoff, peak velocity reduction rate, runoff coefficient reduction rate, outflow peak velocity reduction rate, and total outflow runoff reduction rate.

[0010] In one implementation of the first aspect, performing two-stage clustering on the runoff regulation data to obtain the correlation between different variables at different levels, in order to compare the effectiveness of design schemes for different control objectives, includes the following steps: clustering the runoff regulation data to obtain initial cluster centers and initial K values; performing K-means clustering based on the initial cluster centers and initial K values ​​to obtain the correlation between different variables at different levels, in order to compare the effectiveness of design schemes for different control objectives.

[0011] In one implementation of the first aspect, clustering the runoff regulation data to obtain initial cluster centers and initial K values ​​includes the following steps: inputting the runoff regulation data into a self-organizing map neural network model for clustering to obtain a self-organizing map neural network topology; obtaining runoff regulation self-organizing mapping results for different LID scenarios in each sub-catchment area based on the self-organizing map neural network topology; and calculating the cluster centers and initial K values ​​of the self-organizing map neural network based on the runoff regulation self-organizing mapping results.

[0012] In one implementation of the first aspect, the process of re-clustering based on the initial cluster centers and initial K values ​​to obtain the correlation between different variables at different planes, in order to compare the effects of design schemes for different control objectives, includes the following steps: using the initial cluster centers and initial K values ​​as initial values ​​to perform K-means clustering to obtain the optimal number of clusters; dividing the neuron classification clusters of different LID scenarios into neuron node training graphs according to the optimal number of clusters; and judging the correlation between variables based on the neuron node training graphs.

[0013] In one implementation of the first aspect, inputting the runoff regulation data into a self-organizing map neural network model for clustering to obtain a self-organizing map neural network topology includes the following steps: storing the runoff regulation data in an array, obtaining the input vector and weight vector of the runoff regulation data; normalizing the input vector and the weight vector; and calculating the Euclidean distance between the normalized input vector and each weight vector.

[0014] In one implementation of the first aspect, the input vector formula is:

[0015] X (n) =(x1(n),x2(n),...,x n (n),)

[0016] The formula for the weight vector is:

[0017] W i(t) =(W i1 (t),W i2 (t),...,W im (t),),i=1,2,3...,m

[0018] Where n represents the number of samples in the input vector; m represents the number of sample indices in the weight vector.

[0019] The normalization calculation formula is as follows:

[0020]

[0021]

[0022] in, Represented as the normalized current input vector; It is represented as the normalized initial weight vector of the i-th neuron.

[0023] Secondly, this application provides a multi-LID scenario runoff regulation effect identification system based on SOM and K-means two-order clustering, including: an acquisition module for acquiring basic data of the target area; a data processing module for calculating runoff regulation data of each sub-catchment of the target area based on the basic data; and a clustering and discrimination module for performing two-order clustering on the runoff regulation data to obtain the correlation between different variables at different levels, so as to realize the comparison of the effects of design schemes for different control targets.

[0024] Finally, this application provides a device for identifying the runoff regulation effect of multiple LID scenarios based on SOM and K-means two-order clustering, comprising: a processor and a memory. The memory is used to store a computer program; the processor is connected to the memory and is used to execute the computer program stored in the memory, so that the device for identifying the runoff regulation effect of multiple LID scenarios based on SOM and K-means two-order clustering performs the method for identifying the runoff regulation effect of multiple LID scenarios based on SOM and K-means two-order clustering.

[0025] As described above, the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention has the following beneficial effects:

[0026] This application provides a method for identifying the runoff regulation effect under multiple LID scenarios based on SOM and K-means two-order clustering. Under different control objectives, this method allows for the selection of corresponding runoff control indicators to identify the strength of regional background rainfall-runoff control capabilities. Based on this, spatial control of the regional low-impact development (LID) construction layout can be implemented. For larger areas with numerous control units, the SOM and K-means two-order clustering method leverages the dimensionality reduction and spatial visualization advantages of SOM. Under multiple sub-catchments, multiple hydrological control objectives, multiple modification scenario design schemes, and multiple rainfall scenarios, the four-dimensional scale is reduced to two dimensions, and spatial visualization is used to compare the effects of design schemes for different control objectives, providing a reference for selecting schemes that meet different construction requirements. Attached Figure Description

[0027] Figure 1The diagram shown is a flowchart of an embodiment of the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention.

[0028] Figure 2 The diagram shows an application scenario of the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention in one embodiment.

[0029] Figure 3A The diagram shown is a flowchart of step S12 in the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention.

[0030] Figure 3B The diagram shows the target sub-catchment area in the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention.

[0031] Figure 4A The results shown are the U-matrix self-organizing mapping results of the rainfall-runoff control index after the SOM of the present invention is modified for different LID scenarios in each sub-catchment area.

[0032] Figure 4B The image shows the self-organized mapping results of the SOM (Self-Organized Mapping Method) of the present invention for the reduction of total runoff in rainfall-runoff control indicators after different LID (Limited Irrigation Detection) scenarios in each sub-catchment area.

[0033] Figure 4C The image shows the self-organized mapping results of the SOM (Self-Organized Mapping Method) for reducing the peak velocity of rainfall-runoff control indicators after different LID (Limited Irrigation Detection) scenarios in each sub-catchment area.

[0034] Figure 4D The image shows the self-organized mapping results of the SOM (Self-Organized Mapping Method) for reducing the runoff coefficient, a control index for rainfall-runoff control, after different LID (Limited Irrigation Detection) scenarios in each sub-catchment area, according to the present invention.

[0035] Figure 4E The diagram shows the self-organized mapping results of the SOM (Self-Organized Mapping Method) of the rainfall-runoff control index, peak flow velocity reduction (outflow), after the modification of different LID scenarios in each sub-catchment area according to the present invention.

[0036] Figure 4F The image shows the self-organized mapping results of the SOM (Self-Organized Mapping Method) of the present invention for the reduction (outflow) of total runoff in rainfall-runoff control indicators after different LID (Low Id) scenarios in each sub-catchment area.

[0037] Figure 5 The diagram shown is a flowchart of step S13 in the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention.

[0038] Figure 6 The image shown is a Davies-Bouldwin index diagram in the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention.

[0039] Figure 7A This diagram illustrates the sample size of the neuron clustering in the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention.

[0040] Figure 7B The diagram shows the distribution of neurons for runoff control efficiency indicators under different LID (Low Idling) modification scenarios according to the present invention.

[0041] Figure 8 The diagram shown is a schematic representation of the principle structure of the multi-LID scenario runoff regulation effect identification system based on SOM and K-means two-order clustering of the present invention in one embodiment.

[0042] Figure 9 The diagram shown is a schematic representation of the principle structure of the multi-LID scenario runoff regulation effect identification device based on SOM and K-means two-order clustering of the present invention in one embodiment.

[0043] Component designation explanation

[0044] 81 Acquisition Module

[0045] 82 Data Processing Module

[0046] 83 Clustering and Discriminant Module

[0047] 91 processor

[0048] 92 Memory

[0049] Steps S11 to S14 Detailed Implementation

[0050] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0051] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0052] The method for identifying the runoff regulation effect based on SOM and K-means two-order clustering provided in this application will be described in detail below with reference to the accompanying drawings in the embodiments of this application.

[0053] Please see Figure 1 and Figure 2 The figures show a flowchart of an embodiment of the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention, and an application scenario diagram of an embodiment of the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention. Figure 1 and Figure 2 As shown in the figure, this embodiment provides a method for identifying the runoff regulation effect in multiple LID scenarios based on two-order clustering of SOM and K-means.

[0054] The method for identifying the runoff regulation effect in multiple LID scenarios based on SOM and K-means two-order clustering specifically includes the following steps:

[0055] S11, Obtain basic data for the target area. Please refer to [link / reference]. Figure 3A The diagram shows a flowchart of step S11 in the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention. Figure 3A As shown, step S11 includes the following steps:

[0056] In this embodiment, basic data within the region is collected. This basic data includes, but is not limited to: target area elevation, underlying surface, drainage system and facility data, regional water system data, and hydrological and rainfall data.

[0057] Elevation refers to the distance from a point along the plumb line to the absolute datum, called absolute elevation, or simply elevation. The distance from a point along the plumb line to an assumed leveling datum is called assumed elevation. The elevation datum is the starting point for calculating all leveling elevations in the national unified elevation control network; it includes a leveling datum and a permanent leveling origin. Methods of elevation measurement include: leveling, trigonometric leveling, GNSS (Global Navigation Satellite System) elevation measurement, and physical elevation measurement.

[0058] The underlying surface is the interface between the atmosphere and the solid ground or liquid water surface below it. It is the main source of atmospheric heat and water vapor, and also the boundary surface for lower atmospheric movement. The underlying surface can also be described as the characteristics of the Earth's surface, such as the distribution of land and sea, topographic relief and surface roughness, vegetation, soil moisture, snow cover, etc., and its influence on climate is significant. Underlying surface factors include: topography (elevation, windward and leeward positions), location relative to land and sea, ocean currents, and vegetation cover. In terms of the scale of underlying surface differences and their role in climate formation, the difference between land and sea is the most fundamental, primarily affecting temperature, moisture, and circulation.

[0059] Data on drainage systems and facilities includes: the number of drainage pipe networks, the number of drainage wells, and the diameter of drainage pipe networks.

[0060] The regional water system data includes three layers: water system surface, water system line, and water system point. The water system surface includes lakes, reservoirs, double-line rivers, and ditches; the water system line includes single-line rivers, ditches, and river structure lines; and the water system point includes springs and wells.

[0061] Hydrological and rainfall data include: water level, precipitation, flow rate, sediment, water quality, and other data.

[0062] In summary, the basic data in this embodiment were all obtained through methods that can be implemented in the relevant technical fields.

[0063] S12, Based on the aforementioned basic data, calculate the runoff regulation data for each sub-catchment of the target area. Please refer to [link / reference]. Figure 3A and Figure 3B Figure 4 shows a flowchart of step S12 in the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention, and a schematic diagram of the target area sub-catchment in the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention. As shown in Figure 4, step S12 includes the following steps:

[0064] S121, The basic data is normalized to obtain rainstorm runoff management data.

[0065] In this embodiment, the acquired basic data undergoes secondary processing.

[0066] Specifically, different types of basic data are standardized to generate basic data of the same type, thereby obtaining the data required for storm runoff management modeling, i.e., storm runoff management data. For example, if the basic data is drainage facility data in CAD format, which is non-standardized data, it needs to be converted to generate a data type suitable for storm flood management models.

[0067] S122, the storm runoff management data is input into the storm flood management model to calculate the impervious surface ratio of the target area, and the target area is divided into several sub-catchments and the impervious surface ratio of each sub-catchment. Please continue reading. Figure 3B .

[0068] In this embodiment, stormwater runoff management data is input into the SWMM (storm water management model) to generalize the drainage system network of the target area. The impervious surface ratio of the target area is calculated using the spatial analysis tool ArcGIS. Then, several sub-catchments are divided, and the impervious surface ratio of each sub-catchment is calculated.

[0069] The SWMM model is a dynamic precipitation-runoff simulation model, primarily used to simulate a single precipitation event or long-term water quantity and quality simulation in a city. Its runoff module comprehensively processes precipitation, runoff, and pollution loads occurring in each sub-basin. Its confluence module transmits water through pipe networks, channels, water storage and treatment facilities, pumps, and regulating gates. This model can track and simulate the water quality and quantity of runoff generated in each sub-basin at any time step at different time steps, as well as the flow rate, depth, and water quality in each pipe and channel.

[0070] Specifically, let's take a target area as an example. Referring to Table 1, the stormwater runoff management data for the target area—that is, the processed basic data of the target area (such as data on different rainfall intensities and different urban construction)—is input into the SWMM model. This divides the target area into 30 smaller areas, and then calculates five indicators for each of these 30 smaller areas based on different precipitation conditions. The first smaller area is a sub-catchment. Therefore, the target area (containing 30 sub-catchments) generates 150 data points. Then, the spatial analysis tool ArcGIS is used to calculate the impervious surface ratio of these data belonging to the same sub-catchment to obtain the impervious surface ratio of each sub-catchment.

[0071] S123, calculate the runoff regulation data for each sub-catchment based on the sub-catchment area and its impermeable surface ratio. Please refer to [link / reference]. Figure 4A , 4B4C, 4D, 4E, and 4F respectively represent the following: the self-organized mapping results of the rainfall-runoff control index U-matrix of the SOM for runoff regulation data after modification under different LID scenarios in each sub-catchment area; the self-organized mapping results of the SOM for the reduction of total runoff volume of the rainfall-runoff control index after modification under different LID scenarios in each sub-catchment area; the self-organized mapping results of the SOM for the reduction of peak velocity of the rainfall-runoff control index after modification under different LID scenarios in each sub-catchment area; the self-organized mapping results of the SOM for the reduction of runoff coefficient of the rainfall-runoff control index after modification under different LID scenarios in each sub-catchment area; the self-organized mapping results of the SOM for the reduction of peak flow velocity (outflow) of the rainfall-runoff control index after modification under different LID scenarios in each sub-catchment area; and the self-organized mapping results of the SOM for the reduction of total runoff volume (outflow) of the rainfall-runoff control index after modification under different LID scenarios in each sub-catchment area.

[0072] In this embodiment, multiple LID design scenarios are set according to the target area and the area and impermeability of each sub-catchment. Based on the different LID design scenarios, the runoff regulation capacity of each sub-catchment is simulated and calculated using the SWMM model.

[0073] Table 1: Comparison of Data from Various LID Design Scenarios

[0074]

[0075] Specifically, the LID design scenario includes six scenarios as an example. Please refer to Table 1. The preferred LID design scenario is six scenarios. As mentioned above, 30 sub-catchments have been divided in the target area. The data for each sub-catchment includes five runoff control data points: total outflow runoff, peak velocity reduction rate, runoff coefficient reduction rate, outflow peak velocity reduction rate, and total outflow runoff reduction rate. Based on the six different LID design scenarios, the runoff control data for each sub-catchment is simulated and calculated using the SWMM model. In this embodiment, the runoff reduction effect is calculated using the simulation results (including 900 sample data points) of the five runoff control data points from the sub-catchments divided in the 30 SWMM models.

[0076] The sample data quantity = number of sub-catchment areas * number of runoff regulation data * number of design scenarios.

[0077] S13, perform two-stage clustering on the runoff regulation data to obtain the correlation between different variables at different levels, so as to compare the effects of design schemes for different control objectives. Please refer to... Figure 5The diagram shows a flowchart of step S13 in the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention. Figure 5 As shown, step S13 includes the following steps:

[0078] S131, the runoff regulation data is clustered to obtain initial cluster centers and initial K values. (See also...) Figure 6 , Figure 7A and Figure 7B The figures are respectively shown as follows: the Davies-Bouldwin index diagram in the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention; the sample size diagram of the neuron cluster in the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering of the present invention; and the neuron distribution diagram of the runoff control efficiency index under different LID modification scenario modes of the present invention.

[0079] In this embodiment, a Self-organizing Map (SOM) neural network is used to perform cluster analysis on five hydrological parameters of each sub-catchment. The runoff regulation data is input into the SOM neural network model for clustering to obtain a SOM neural network topology map. Based on the SOM neural network topology map, the runoff regulation self-organizing mapping results for different LID scenarios in each sub-catchment are obtained. Based on the runoff regulation self-organizing mapping results, the cluster centers and initial K values ​​of the SOM neural network are calculated.

[0080] Specifically, the runoff regulation data is stored in an array to obtain the input vector and weight vector of the runoff regulation data; the input vector and weight vector are normalized; the Euclidean distance between the normalized input vector and each weight vector is calculated. A self-organizing mapping result is obtained, that is, the four-dimensional scale data is reduced to two dimensions and represented spatially; based on the runoff regulation self-organizing mapping result, the cluster centers and initial K values ​​of the self-organizing mapping neural network are calculated.

[0081] For example, by inputting the above parameters into the SOM neural network model, clustering is performed according to the similarity of the distribution of each parameter to obtain the SOM neural network topology. Each cluster pattern represents the runoff regulation capability of each LID scenario in each sub-catchment area. The self-organizing mapping results of rainfall-runoff control indicators for different LID scenarios in each sub-catchment area are obtained, and the cluster centers and the number of clusters after clustering are calculated.

[0082] The specific calculation process involved is as follows:

[0083] 1. Store the runoff regulation data in an array; let the sample size be n, the number of sample indicators be m, and the sample vector be m-dimensional. The calculation formula is as follows:

[0084] The formula for the input vector is:

[0085] X (n) =(x1(n),x2(n),...,x n (n),)

[0086] The formula for the weight vector is:

[0087] W i(t) =(W i1 (t),W i2 (t),...,W im (t),),i=1,2,3...,m

[0088] Where n represents the number of samples in the input vector; m represents the number of sample indices in the weight vector.

[0089] 2. Normalize the input vector and weight vector.

[0090] The normalization calculation formula is as follows:

[0091]

[0092]

[0093] in, Represented as the normalized current input vector; It is represented as the normalized initial weight vector of the i-th neuron.

[0094] 3. For the input vector Xi, calculate its Euclidean distance to each weight matrix.

[0095] The Euclidean distance calculation formula is as follows:

[0096]

[0097] Where, d k It is represented as the Euclidean distance between the input vector and the weight matrix.

[0098] 4. Randomly select the initial weight vector W j Based on this, an initial winning neighborhood kj(0) is established, and an initial value η(0) is assigned to the learning rate.

[0099] 5. Update similarity matching.

[0100] Right now:

[0101] Wj (n+1)=W j (n)+η(n)h j ,d k (n)(x(n),w j (n))

[0102] Where η(n) represents the learning rate parameter; d k (n) represents the neighborhood function of the winning neuron i(x).

[0103] 6. When the results in the feature mapping no longer change significantly, stop training to obtain the initial cluster centers and initial K values ​​of the SOM neural network class.

[0104] At this point, the cluster centers and number of clusters obtained after clustering are actually a data range, and the process of deriving an optimal value from this data range is crucial. For example, after SOM processing, the point with the minimum geometric distance between different neurons is obtained; this point represents the optimal point and the optimal number of neurons. Figure 6 As can be seen from this, the effect is best when the number of cluster points is 10.

[0105] Therefore, after importing 900 sample data into the SOM neural network in step S12, performing standardization processing, and conducting self-organizing cluster analysis, the final number of neuron nodes is determined to be 60 according to the formula for calculating the number of neurons.

[0106] S132, K-means clustering is performed based on the initial cluster centers and initial K values ​​to obtain the correlation between different variables on different planes, so as to compare the effects of design schemes for different control objectives.

[0107] In this embodiment, the initial cluster centers and the initial K value are used as initial values ​​for K-means clustering to obtain the optimal number of clusters; the neuron classification clusters for different LID scenarios are divided according to the optimal number of clusters to form a neuron node training graph; and the correlation between variables is determined based on the neuron node training graph.

[0108] Specifically, the cluster centers and the number of clusters after SOM neural network clustering are used as the initial cluster centers and initial K value for K-means clustering. The K-means clustering algorithm is executed until convergence, and the optimal number of clusters is selected by combining the DB index.

[0109] Neuron classification clusters for different LID scenarios are divided according to the optimal number of clusters. The training graph composed of neuron nodes linked by their output vectors is represented graphically in a 2D plane (U-matrix, Variableplanes), where each variable is indicated by a different color to indicate the variable distribution value on different graph regions.

[0110] The method for identifying the runoff regulation effect under multiple LID scenarios based on SOM and K-means two-order clustering provided in this application is used to guide and evaluate the runoff control index reduction rate under different LID regulation scenarios.

[0111] The scope of protection of the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering described in this application is not limited to the execution order of the steps listed in this embodiment. Any solution implemented by adding, subtracting or replacing steps in the prior art based on the principle of this application is included within the scope of protection of this application.

[0112] This embodiment also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the following... Figure 1 The proposed method for identifying the runoff regulation effect in multiple LID scenarios based on two-order clustering of SOM and K-means.

[0113] At any possible level of technical detail, this application can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this application.

[0114] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example, (but not limited to) electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0115] The computer-readable program described herein can be downloaded from a computer-readable storage medium to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives computer-readable program instructions from the network and forwards these instructions to a computer-readable storage medium in the respective computing / processing device. The computer program instructions used to perform the operations of this application may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, integrated circuit configuration data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and procedural programming languages ​​such as "C" or similar programming languages. Computer-readable program instructions may execute entirely on a user's computer, partially on a user's computer, as a standalone software package, partially on a user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing state information from the computer-readable program instructions. This electronic circuitry can execute the computer-readable program instructions to implement various aspects of this application.

[0116] This application also provides a multi-LID scenario runoff regulation effect identification system based on SOM and K-means two-order clustering. The multi-LID scenario runoff regulation effect identification system based on SOM and K-means two-order clustering can implement the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering described in this application. However, the implementation device of the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering described in this application includes, but is not limited to, the structure of the multi-LID scenario runoff regulation effect identification system based on SOM and K-means two-order clustering listed in this embodiment. All structural modifications and substitutions of the prior art made according to the principles of this application are included within the protection scope of this application.

[0117] The following will describe in detail the multi-LID scenario runoff regulation effect identification system based on SOM and K-means two-order clustering provided in this embodiment, with reference to the illustrations.

[0118] This embodiment provides a multi-LID scenario runoff regulation effect identification system based on SOM and K-means two-order clustering, including:

[0119] Please see Figure 8 The diagram shows a schematic representation of the principle structure of the multi-LID scenario runoff regulation effect identification system based on SOM and K-means two-order clustering in one embodiment of the present invention. Figure 8 As shown, the multi-LID scenario runoff regulation effect identification system based on SOM and K-means two-order clustering includes: an acquisition module 81, a data processing module 82, and a clustering and discrimination module 83.

[0120] The acquisition module 81 is used to acquire basic data of the target area.

[0121] In this embodiment, basic data within the region is collected. This basic data includes, but is not limited to: target area elevation, underlying surface, drainage system and facility data, regional water system data, and hydrological and rainfall data.

[0122] Among them, underlying surface factors include: topography (elevation, windward and leeward), location relative to land and sea, ocean currents, vegetation cover, etc.

[0123] Data on drainage systems and facilities includes: the number of drainage pipe networks, the number of drainage wells, and the diameter of drainage pipe networks.

[0124] The regional water system data includes three layers: water system surface, water system line, and water system point. The water system surface includes lakes, reservoirs, double-line rivers, and ditches; the water system line includes single-line rivers, ditches, and river structure lines; and the water system point includes springs and wells.

[0125] Hydrological and rainfall data include: water level, precipitation, flow rate, sediment, water quality, and other data.

[0126] In summary, the basic data in this embodiment were all obtained through methods that can be implemented in the relevant technical fields.

[0127] The data processing module 82 is connected to the acquisition module 81 and is used to calculate the runoff regulation data of each sub-catchment of the target area based on the basic data.

[0128] In this embodiment, runoff regulation data for each sub-catchment of the target area are calculated based on the basic data.

[0129] The basic data is normalized to obtain storm runoff management data.

[0130] In this embodiment, the acquired basic data undergoes secondary processing.

[0131] Specifically, different types of basic data are standardized to generate basic data of the same type, thereby obtaining the data required for storm runoff management modeling, namely: storm runoff management data.

[0132] The storm runoff management data is input into the storm flood management model to calculate the impervious surface ratio of the target area, and the target area is divided into several sub-catchments and the impervious surface ratio of the sub-catchments.

[0133] In this embodiment, stormwater runoff management data is input into the SWMM model to generalize the drainage system network of the target area. The impervious surface ratio of the target area is calculated using the spatial analysis tool ArcGIS. Then, several sub-catchments are divided, and the impervious surface ratio of each sub-catchment is calculated.

[0134] Runoff regulation data for each sub-catchment is calculated based on the sub-catchment area and its impermeable surface ratio.

[0135] In this embodiment, multiple LID design scenarios are set according to the target area and the area and impermeability of each sub-catchment. Based on the different LID design scenarios, the runoff regulation capacity of each sub-catchment is simulated and calculated using the SWMM model.

[0136] Specifically, we will illustrate this using six LID design scenarios as an example. Please refer to Table 1. The six preferred LID design scenarios are as follows. As mentioned above, 30 sub-catchments have been identified within the target area. Data for each sub-catchment includes five runoff control metrics: total outflow runoff, peak velocity reduction rate, runoff coefficient reduction rate, outflow peak velocity reduction rate, and total outflow runoff reduction rate. Based on the six different LID design scenarios, the runoff control data for each sub-catchment is simulated and calculated using the SWMM model.

[0137] The simulation results of the regulation data (including 900 sample data) were used to calculate the runoff reduction effect. The number of sample data points = number of sub-catchments * number of runoff regulation data points * number of design scenarios.

[0138] The prediction module 83 is used to perform two-stage clustering on the runoff regulation data to obtain the correlation between different variables at different levels, so as to compare the effects of design schemes for different control objectives.

[0139] The runoff regulation data is subjected to two-stage clustering to obtain the correlation between different variables at different levels, so as to compare the effects of design schemes for different control objectives.

[0140] The runoff regulation data are clustered to obtain initial cluster centers and initial K values.

[0141] In this embodiment, the SOM neural network is used to perform cluster analysis on five hydrological parameters of each sub-catchment. The runoff regulation data is input into the self-organizing map neural network model for clustering to obtain a self-organizing map neural network topology. Based on the self-organizing map neural network topology, the runoff regulation self-organizing mapping results for different LID scenarios in each sub-catchment are obtained. Based on the runoff regulation self-organizing mapping results, the cluster centers and initial K values ​​of the self-organizing map neural network are calculated.

[0142] Specifically, the runoff regulation data is stored in an array to obtain the input vector and weight vector of the runoff regulation data; the input vector and weight vector are normalized; the Euclidean distance between the normalized input vector and each weight vector is calculated. A self-organizing mapping result is obtained, that is, the four-dimensional scale data is reduced to two dimensions and represented spatially; based on the runoff regulation self-organizing mapping result, the cluster centers and initial K values ​​of the self-organizing mapping neural network are calculated.

[0143] The specific calculation process involved is as follows:

[0144] 1. Store the runoff regulation data into an array; let the number of samples be n, the number of sample indicators be m, and the sample vector be m-dimensional.

[0145] 2. Normalize the input vector and weight vector.

[0146] 3. For the input vector Xi, calculate its Euclidean distance to each weight matrix.

[0147] 4. Randomly select an initial weight vector, and based on this, establish an initial winning neighborhood and assign an initial value to the learning rate.

[0148] 5. Update similarity matching.

[0149] 6. When the results in the feature mapping no longer change significantly, stop training to obtain the initial cluster centers and initial K values ​​of the SOM neural network class.

[0150] At this point, the cluster centers and the number of clusters obtained after clustering are actually a data range, and the process of obtaining an optimal value from this data range is then carried out.

[0151] K-means clustering is performed based on the initial cluster centers and initial K values ​​to obtain the correlation between different variables on different planes, so as to compare the effects of design schemes for different control objectives.

[0152] In this embodiment, the initial cluster centers and the initial K value are used as initial values ​​for K-means clustering to obtain the optimal number of clusters; the neuron classification clusters for different LID scenarios are divided according to the optimal number of clusters to form a neuron node training graph; and the correlation between variables is determined based on the neuron node training graph.

[0153] Specifically, the cluster centers and the number of clusters after SOM neural network clustering are used as the initial cluster centers and initial K value for K-means clustering. The K-means clustering algorithm is executed until convergence, and the optimal number of clusters is selected by combining the DB index. Neuron classification clusters for different LID scenarios are divided according to the optimal number of clusters. The training graph composed of neuron nodes linked by their output vectors is represented graphically in a 2D plane (U-matrix, variable planes), where each variable is indicated by a different color to indicate the variable distribution value on different graph regions.

[0154] A multi-LID scenario runoff regulation effect identification system based on SOM and K-means two-order clustering is built. This system can reduce the four-dimensional scale to two dimensions and use spatial display to compare the effects of design schemes for different control objectives in multiple sub-catchments, multiple hydrological control target indicators, multiple modification scenario design schemes, and multiple rainfall scenarios, providing a reference for the selection of schemes to meet different construction requirements.

[0155] It should be noted that the division of the various modules in the above system is merely a logical functional division. In actual implementation, they can be fully or partially integrated into a single physical entity, or they can be physically separated. Furthermore, these modules can be implemented entirely in software through processing element calls; they can be fully implemented in hardware; or some modules can be implemented by processing element calls to software, while others are implemented in hardware. For example, module x can be a separate processing element, or it can be integrated into a chip within the system. Alternatively, it can be stored as program code in the system's memory, and its function can be called and executed by a processing element. The implementation of other modules is similar. Moreover, these modules can be fully or partially integrated together, or they can be implemented independently. The processing element mentioned here can be an integrated circuit with signal processing capabilities. During implementation, each step of the above method or each of the above modules can be completed through integrated logic circuits in the hardware of the processor element or through software instructions.

[0156] These modules can be one or more integrated circuits configured to implement the above methods, such as one or more Application Specific Integrated Circuits (ASICs), one or more digital signal processors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs). Alternatively, when a module is implemented using processing element scheduler code, the processing element can be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. Furthermore, these modules can be integrated together as a system-on-a-chip (SOC).

[0157] Please see Figure 9 The diagram shows a schematic representation of the principle structure of the multi-LID scenario runoff regulation effect identification device based on SOM and K-means two-order clustering of the present invention in one embodiment. Figure 9As shown, this embodiment provides a multi-LID scenario runoff regulation effect identification device based on SOM and K-means two-order clustering. The multi-LID scenario runoff regulation effect identification device based on SOM and K-means two-order clustering includes: a processor 91 and a memory 92; the memory 92 is used to store computer programs; the processor 91 is connected to the memory 92 and is used to execute the computer programs stored in the memory 92, so that the multi-LID scenario runoff regulation effect identification device based on SOM and K-means two-order clustering performs the various steps of the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering as described above.

[0158] Preferably, the memory may include random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device.

[0159] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0160] In summary, the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering provided in this application has the following beneficial effects:

[0161] The multi-LID scenario runoff control effect identification method based on SOM and K-means two-order clustering provided in this application can identify the strength of regional background rainfall runoff control capacity under different control objectives. This method allows for the selection of corresponding runoff control indicators, enabling spatial control of regional low-impact development (LID) construction layouts. For larger regions with numerous control units, the multi-LID scenario runoff control effect identification method leverages the dimensionality reduction and spatial display advantages of SOM. Under multiple sub-catchments, multiple hydrological control objectives, multiple modification scenario design schemes, and multiple rainfall scenarios, it reduces the four-dimensional scale to two dimensions and uses spatial display to compare the effects of design schemes for different control objectives, providing a reference for selecting schemes that meet different construction requirements.

[0162] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A method for identifying the runoff regulation effect in multiple LID scenarios based on two-order clustering of SOM and K-means, characterized in that, Includes the following steps: Obtain basic data for the target area; The calculation of runoff control data for each sub-catchment of the target area based on the aforementioned basic data includes: normalizing the basic data to obtain storm runoff management data; inputting the storm runoff management data into a storm flood management model to calculate the impervious surface ratio of the target area, and dividing the target area into several sub-catchments and the impervious surface ratio of each sub-catchment; calculating the runoff control data for each sub-catchment based on the sub-catchments and their impervious surface ratios; wherein the runoff control data includes: total outflow runoff, peak velocity reduction rate, runoff coefficient reduction rate, outflow peak velocity reduction rate, and total outflow runoff reduction rate. The runoff regulation data is subjected to two-order clustering to obtain the correlation between different variables at different levels, so as to compare the effects of design schemes for different control objectives; that is: The process involves clustering the runoff regulation data to obtain initial cluster centers and an initial K value. This includes: inputting the runoff regulation data into a self-organizing map neural network model for clustering to obtain a self-organizing map neural network topology; obtaining runoff regulation self-organizing mapping results for different LID scenarios in each sub-catchment area based on the self-organizing map neural network topology; and calculating the cluster centers and initial K value of the self-organizing map neural network based on the runoff regulation self-organizing mapping results. K-means clustering is performed based on the initial cluster centers and initial K values ​​to obtain the correlation between different variables at different planes, so as to compare the effects of design schemes for different control objectives; including: using the initial cluster centers and initial K values ​​as initial values ​​to perform K-means clustering to obtain the optimal number of clusters; dividing the neuron classification clusters of different LID scenarios into neuron node training graphs according to the optimal number of clusters; and judging the correlation between variables based on the neuron node training graphs.

2. The method for identifying the runoff regulation effect in multiple LID scenarios based on SOM and K-means two-order clustering as described in claim 1, characterized in that, The basic data includes any one or more combinations of the following: target area elevation, underlying surface, drainage system and facility data, regional water system data, and hydrological and rainfall data.

3. The method for identifying the runoff regulation effect in multiple LID scenarios based on SOM and K-means two-order clustering as described in claim 1, characterized in that, The process of inputting the runoff regulation data into a self-organizing map neural network model for clustering to obtain a self-organizing map neural network topology includes the following steps: Store the runoff regulation data into an array, and obtain the input vector and weight vector of the runoff regulation data; The input vector and the weight vector are normalized. Calculate the Euclidean distance between the normalized input vector and each of the weight vectors.

4. The method for identifying the runoff regulation effect in multiple LID scenarios based on SOM and K-means two-order clustering as described in claim 3, characterized in that, The formula for the input vector is: X (n) =(x1(n),x2(n),…,x n (n),) The formula for the weight vector is: W i(t) =(W i1 (t),W i2 (t),...,W im (t),),i=1,2,3...,m Where n represents the number of samples in the input vector; m represents the number of sample indices in the weight vector; The normalization calculation formula is as follows: in, Represented as the normalized current input vector; It is represented as the normalized initial weight vector of the i-th neuron.

5. A multi-LID scenario runoff regulation effect identification system based on SOM and K-means two-order clustering, characterized in that, include: The acquisition module is used to acquire basic data for the target area; The data processing module is used to calculate runoff control data for each sub-catchment of the target area based on the basic data; including: normalizing the basic data to obtain storm runoff management data; inputting the storm runoff management data into a storm flood management model to calculate the impervious surface ratio of the target area, and dividing the target area into several sub-catchments and the impervious surface ratio of the sub-catchments; calculating runoff control data for each sub-catchment based on the sub-catchments and the impervious surface ratio of the sub-catchments; wherein, the runoff control data includes: total outflow runoff, peak velocity reduction rate, runoff coefficient reduction rate, outflow peak velocity reduction rate, and total outflow runoff reduction rate; The clustering and discrimination module is used to perform two-order clustering on the runoff regulation data to obtain the correlation between different variables at different levels, so as to compare the effects of design schemes for different control objectives; that is: The process involves clustering the runoff regulation data to obtain initial cluster centers and an initial K value. This includes: inputting the runoff regulation data into a self-organizing map neural network model for clustering to obtain a self-organizing map neural network topology; obtaining runoff regulation self-organizing mapping results for different LID scenarios in each sub-catchment area based on the self-organizing map neural network topology; and calculating the cluster centers and initial K value of the self-organizing map neural network based on the runoff regulation self-organizing mapping results. K-means clustering is performed based on the initial cluster centers and initial K values ​​to obtain the correlation between different variables at different planes, so as to compare the effects of design schemes for different control objectives; including: using the initial cluster centers and initial K values ​​as initial values ​​to perform K-means clustering to obtain the optimal number of clusters; dividing the neuron classification clusters of different LID scenarios into neuron node training graphs according to the optimal number of clusters; and judging the correlation between variables based on the neuron node training graphs.

6. A device for identifying the runoff regulation effect in multiple LID scenarios based on two-order clustering of SOM and K-means, characterized in that, include: Processor and memory; The memory is used to store computer programs; The processor is connected to the memory and is used to execute the computer program stored in the memory so that the multi-LID scenario runoff regulation effect identification device based on SOM and K-means two-order clustering performs the multi-LID scenario runoff regulation effect identification method based on SOM and K-means two-order clustering as described in any one of claims 1 to 4.

Images