Urban drainage green measure vertical layout control optimization method and device, equipment and medium

Abstract

The application relates to a method, device, equipment and medium for optimizing vertical layout control of urban drainage green measures, which comprises the following steps: determining a plurality of layout evaluation indexes according to a region to be laid with green measures, waterlogging water volume, green measure storage capacity and population information of a city to be managed for drainage by using a GIS geographic analysis technology; determining a layout priority coefficient of each green measure type in each catchment area based on a fuzzy comprehensive evaluation method according to the plurality of layout evaluation indexes and the weight corresponding to each layout evaluation index; obtaining a total outflow of green measures and a Pareto optimal solution set of total cost of green measure layout of all catchment areas in the city to be managed for drainage by using a preset green measure vertical collaborative optimization algorithm according to the layout priority coefficient and green measure parameters to be laid; and determining an optimal green measure layout scheme of the region to be laid with green measures in each catchment area. The application can significantly improve the flood control efficiency of the urban drainage system.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of drainage control, and particularly relates to a method for optimizing vertical layout control of urban drainage green measures, a corresponding device, an electronic device, and a computer readable storage medium. BACKGROUND

[0002] Urban stormwater management has traditionally relied on drainage networks and grey infrastructure, with the goal of rapidly removing stormwater from urban areas through centralized systems. However, the rapid pace of urbanization has led to a significant increase in impervious surfaces, which has resulted in shorter stormwater collection times and increased peak flow. Existing drainage systems often cannot be rapidly expanded or updated, exacerbating flood risk. To address the environmental impacts of urbanization, sustainable low-impact development (LID) strategies have been introduced. As stormwater management space becomes increasingly scarce in cities, research has shifted towards improving the multi-dimensional design of green measures and optimizing their allocation strategies to maximize the feasibility and effectiveness of these spaces in stormwater management.

[0003] Early studies were relatively general and rough in terms of land use classification and sub-basin feature description. Ahiablame and Shakya (2016) used a reclassification technique for land use and found that different levels of low-impact development (LID) implementation resulted in significant differences in runoff control rates (ranging from 3% to 47%). In later stages, remote sensing technology and high-precision satellite images have been used for more detailed land use classification. Randall et al. (2019) used WorldView-3 images and GEOBIA to classify land use and model various low-impact development (LID) deployment scenarios. Recently, to further improve the accuracy and efficiency of land use classification, semantic segmentation algorithms based on convolutional neural networks (CNN), such as FCN (Long et al., 2015), U-Net (Yang et al., 2021), SegNet (Son et al., 2022), and Deeplabv3+ (Rahman et al., 2021), have been used to automatically identify detailed land use conditions. These advanced algorithms can achieve pixel-level segmentation and capture more detailed land use features in satellite images. However, there is currently no method specifically for land use detection and feasibility assessment for vertical low-impact development (LID) allocation.

[0004] (Jia et al., 2012; Kong et al., 2017; Xie et al., 2017) introduced independent green modules, such as green roofs, rain gardens, and permeable pavement, in modeling tools, but these green modules can only be laid out and connected in two-dimensional scale. (Liu et al., 2015) used a series of models to more accurately simulate these water flow dynamics, (Gao et al., 2019) showed that integrated measures are superior to single measures in terms of runoff control, and (Zhang et al., 2021) showed that cascading chains of green measures are more effective in reducing runoff than parallel connections, but they mainly focused on planar layout and ignored the potential of vertical space combination.

[0005] To reasonably evaluate local conditions and the applicability of various factors (Hou et al., 2019; Huang et al., 2024), the hydrological and geological characteristics of the soil, as well as related parameters such as soil type, slope, and hydraulic conductivity, are considered to affect the feasibility of green measure distribution. Heidari (2022) further considered the priorities of different stakeholders such as municipal authorities, builders, and planners, and evaluated ten indicators using a multi-criteria decision method to determine the most effective green measures. In terms of green measure distribution, more comprehensive methods are being developed, but evaluation is mainly carried out in planar layout. The suitability of green measures in the vertical direction has not been integrated, and existing technologies mainly have the following technical defects, which include:

[0006] First, the data precision is insufficient: traditional land use classification methods such as remote sensing visual interpretation have low precision and are difficult to support precise layout of green measures.

[0007] Second, there is no dedicated feasibility evaluation method: there is no land use detection and feasibility evaluation method specifically for vertical low-impact development (LID) distribution.

[0008] Third, planar layout is limited: existing methods mainly focus on two-dimensional planar optimization, which cannot fully utilize urban vertical space resources such as building tops and underground pipe networks, resulting in the potential of flood control not being fully tapped.

[0009] Fourth, the priority evaluation is single: existing evaluation systems mainly rely on a single indicator such as runoff reduction rate, and lack comprehensive priority evaluation in multiple dimensions such as land use, waterlogging risk, and population density.

[0010] In summary, to adapt to the existing technology which mainly focuses on two-dimensional planar optimization and cannot fully utilize urban vertical space resources, resulting in the potential of flood control not being fully tapped, the applicant made corresponding explorations to solve this problem. SUMMARY

[0011] The application aims to solve the above problems and provide a vertical layout control optimization method for urban drainage green measures, a corresponding device, electronic equipment and a computer readable storage medium.

[0012] To meet the various purposes of the application, the application adopts the following technical solutions:

[0013] A vertical layout control optimization method for urban drainage green measures is proposed to adapt to one of the purposes of the application, comprising:

[0014] In response to an instruction for layout control optimization of green measures in a city to be drained and managed, a preset semantic segmentation model is used to perform land identification segmentation on a remote sensing image of the city to be drained and managed, and a preset vertical feasibility evaluation module is called to perform green measure vertical layout feasibility evaluation on the remote sensing image of the city to be drained and managed, so as to determine the green measure layout area in each catchment area, wherein the city to be drained and managed includes a plurality of catchment areas, and each catchment area includes a plurality of green measure layout areas;

[0015] A vertical water dynamic coupling simulation model integrating gray measures and green measures is called, and a preset rainfall is input into the vertical water dynamic coupling simulation model to determine the waterlogging accumulation and green measure storage capacity of the city to be drained and managed.

[0016] A GIS geographic analysis technique is used to determine a plurality of layout evaluation indexes according to the green measure layout area, the waterlogging accumulation, the green measure storage capacity and the population information of the city to be drained and managed, and a fuzzy comprehensive evaluation method is used to determine the layout priority coefficient of each green measure type in each catchment area according to the plurality of layout evaluation indexes and the weight corresponding to each layout evaluation index.

[0017] The green measure layout area corresponding to the city to be drained and managed is obtained, and a preset green measure vertical coordination optimization algorithm is used to obtain the Pareto optimal solution set of the total outflow of green measures and the total cost of green measure layout in all catchment areas of the city to be drained and managed according to the layout priority coefficient and the green measure layout area, and determine the optimal green measure layout scheme of the green measure layout area in each catchment area.

[0018] Optionally, the step of calling the vertical water dynamic coupling simulation model integrating gray measures and green measures and inputting the preset rainfall into the vertical water dynamic coupling simulation model to determine the waterlogging accumulation and green measure storage capacity of the city to be drained and managed comprises:

[0019] determine the rainfall intensity corresponding to the to-be-drainage-managed city according to a preset rainfall intensity calculation formula, and determine the rainfall amount corresponding to the to-be-drainage-managed city according to a fourth product between the rainfall intensity corresponding to the to-be-drainage-managed city and a preset rainfall duration;

[0020] input the rainfall amount into a preset vertical water dynamic coupling simulation model to determine the waterlogging accumulation amount and the green measure lag storage amount corresponding to the to-be-drainage-managed city, wherein the waterlogging accumulation amount includes the green measure lag storage amount and the residual flood flow, and the green measure lag storage amount includes the aboveground lag storage amount and the underground lag storage amount.

[0021] Optionally, based on the fuzzy comprehensive evaluation method, the step of determining the layout priority coefficient of each green measure type in each catchment area according to the plurality of layout evaluation indexes and the weight corresponding to each layout evaluation index includes:

[0022] obtain a plurality of layout evaluation indexes and the weight corresponding to each layout evaluation index, wherein the layout evaluation indexes include one or any number of the following: land use nature, vertical layout suitability, building density, population density, and waterlogging risk;

[0023] based on the fuzzy comprehensive evaluation method, determine a plurality of evaluation grades of the layout evaluation indexes in each catchment area according to the difference between the maximum value and the minimum value of the layout evaluation indexes, normalize each layout evaluation index into a membership value in the interval [0, 1] using a preset membership function, and determine the normalized membership value of the layout evaluation indexes in each catchment area;

[0024] construct a membership matrix corresponding to each catchment area according to the normalized membership value of the layout evaluation indexes in each catchment area and the plurality of evaluation grades of the layout evaluation indexes in each catchment area;

[0025] determine the importance weight of each layout evaluation index using the analytic hierarchy process, and determine the layout priority coefficient of each green measure type in each catchment area according to a third product between the membership matrix and the importance weight of each layout evaluation index.

[0026] Optionally, the step of determining the total outflow of the green measure includes:

[0027] obtain the number of catchment areas of the to-be-drainage-managed city, the number of green measure types, the unit time outflow of the green measure of each catchment area, and the area of the green measure of each catchment area;

[0028] calculating a first product between each green measure type unit time outflow of each catchment area and a green measure area of each green measure type of each catchment area to determine a green measure outflow water volume of each green measure type of each catchment area;

[0029] determining a total green measure outflow of all catchment areas of the city to be managed according to the green measure outflow water volume of each green measure type of each catchment area, the number of catchment areas of the city to be managed, and the number of green measure types.

[0030] Optionally, the step of determining the total green measure layout cost comprises:

[0031] obtaining the number of catchment areas of the city to be managed, the number of green measure types, a layout priority coefficient of each type of green measure in each catchment area, a unit area cost of each type of green measure, and a layout area of each type of green measure in each catchment area;

[0032] calculating a second product between the layout priority coefficient of each type of green measure in each catchment area, the unit area cost of each type of green measure, and the layout area of each type of green measure in each catchment area;

[0033] determining a total green measure layout cost of all catchment areas of the city to be managed according to the number of catchment areas of the city to be managed, the number of green measure types, and the second product.

[0034] Optionally, the step of determining the optimal green measure layout scheme of the green measure area to be laid out in each catchment area comprises:

[0035] calling a preset green measure vertical coordination optimization algorithm to determine the optimal green measure layout scheme of the green measure area to be laid out in each catchment area according to the layout priority coefficient and the green measure parameter to be laid out, taking the total green measure outflow of all catchment areas and the total green measure layout cost as double optimization objectives, taking the layoutable area of different green measure types in each catchment area as a constraint condition, and using a Gaussian regression model to probabilistically model a nonlinear update between each green measure parameter to be laid out of each catchment area and a response value of an objective function to predict an expected mean and a variance of an objective function value of a potential sample point corresponding to the green measure area to be laid out, wherein the green measure parameter to be laid out represents a green measure sample point corresponding to each green measure type to be laid out of each catchment area.

[0036] The preset expected hyper-volume improvement algorithm is used to calculate a collection score of each potential sample point according to a probability distribution of the Gaussian regression model, and a sample point with the highest score is selected as a target for next function evaluation;

[0037] The above steps are repeated to iteratively generate a set of Pareto optimal solutions satisfying the constraint condition by quantifying an expected gain of the existing Pareto front hyper-volume measure of the new sample, so as to determine an optimal green measure layout scheme of the green measure region to be laid out in each catchment area.

[0038] Optionally, the green measure types include green roof, rainwater garden and permeable pavement.

[0039] The green measure region to be laid out includes one or any combination of commercial building, office building, hospital, school, hotel, traffic building, sports building and residential area.

[0040] The green measure vertical coordination optimization algorithm is a multi-objective Bayesian optimization algorithm.

[0041] The optimal green measure layout scheme represents that the green measure type and the corresponding optimal green measure layout area of the green measure region to be laid out in each catchment area are obtained by solving the Pareto front of the total green measure outflow and the total green measure layout cost in the drainage managed city.

[0042] Another object of the present application is to provide a drainage green measure vertical layout control optimization device for a city, comprising:

[0043] The green measure region to be laid out is determined by a preset semantic segmentation model for land identification and segmentation of a remote sensing image of the drainage managed city, and a preset vertical feasibility evaluation module is called to perform a green measure vertical layout feasibility evaluation on the remote sensing image of the drainage managed city, so as to determine the green measure region to be laid out in each catchment area, wherein the drainage managed city includes a plurality of catchment areas, and each catchment area includes a plurality of green measure regions to be laid out.

[0044] The feasibility data is determined by calling a vertical hydrodynamic coupling simulation model fusing the gray measure and the green measure, and inputting a preset rainfall into the vertical hydrodynamic coupling simulation model, so as to determine the corresponding waterlogging accumulation and green measure storage capacity of the drainage managed city.

[0045] The layout priority determination module is configured to determine a plurality of layout evaluation indexes according to the region to be laid green measures, the waterlogging water volume, the green measure storage capacity and the population information of the city to be drained based on GIS geographic analysis technology, and determine a layout priority coefficient of each green measure type in each catchment area based on a fuzzy comprehensive evaluation method according to the plurality of layout evaluation indexes and the weight corresponding to each layout evaluation index.

[0046] The optimal scheme determination module is configured to obtain a green measure parameter corresponding to the region to be laid green measures of the city to be drained, and obtain a Pareto optimal solution set of a total outflow of green measures and a total cost of green measure layout of all catchment areas in the city to be drained based on a preset green measure vertical coordination optimization algorithm according to the layout priority coefficient and the green measure parameter, and determine an optimal green measure layout scheme of the region to be laid green measures in each catchment area.

[0047] Another object of the present application is to provide an electronic device comprising a central processing unit and a memory, wherein the central processing unit is configured to invoke a computer program stored in the memory to execute the steps of the city drainage green measure vertical layout control optimization method.

[0048] Another object of the present application is to provide a computer readable storage medium storing a computer program implemented according to the city drainage green measure vertical layout control optimization method in the form of computer readable instructions, wherein the computer program is invoked by a computer to execute the steps included in the corresponding method when running.

[0049] Compared with the prior art, the present application can make full use of the vertical space resources of the city, and can not fully exploit the potential of flood control and regulation, and the present application includes but is not limited to the following beneficial effects:

[0050] Firstly, the present application can significantly improve the flood control efficiency of the city drainage system. By constructing a vertical runoff transfer model of above-ground and underground and building multi-layer linkage, the city rainwater flow process can be more comprehensively simulated, and the limitation of traditional two-dimensional plane model can be broken through, thereby significantly improving the storage capacity. This multi-layer linkage modeling method can accurately predict the runoff after rainfall, help to develop more effective drainage schemes, and reduce the risk of urban waterlogging and flood;

[0051] Secondly, the present application can optimize resource allocation and measure implementation efficiency. By comprehensively considering a plurality of indexes such as land use nature, vertical layout suitability, waterlogging risk, building density, population density and the like, the priority can be quantified by the fuzzy evaluation method, and the region needing to be processed in priority can be accurately identified, which not only helps to improve the scientificity of resource allocation, but also improves the efficiency and effect of measure implementation.

[0052] Thirdly, the application can balance flood control and economic cost, adopts a multi-objective Bayesian optimization algorithm to generate a Pareto optimal solution, and can balance flood control efficiency (such as minimizing total outflow) and economic cost (such as minimizing layout cost). This method can provide multiple optimization solutions for decision makers to flexibly select the best solution according to actual budget and demand, so as to ensure the best flood control effect under limited resources;

[0053] Fourthly, the application predicts the effect (such as total outflow and layout cost) of the green measures to be laid out by using a Gaussian process regression model, and selects an optimal region for green measure layout by using an expected hyper volume improvement algorithm. This dynamic optimization method can adjust and update the green measure layout scheme in real time, thereby realizing the flexibility and adaptability of urban drainage management.

[0054] Fifthly, the application can reduce the operation cost and environmental impact of the drainage system. By optimizing the layout of green measures, the outflow of the drainage system can be effectively reduced, the probability of waterlogging and flood occurrence can be reduced, and the construction and maintenance cost of the drainage system can be reduced by reducing excessive infrastructure construction (such as expensive underground drainage pipe network). At the same time, the vertical layout optimization of green measures such as rainwater gardens and permeable pavement can help to improve the urban ecological environment, increase urban green space, and improve the sustainability of the city.

[0055] Further, through this multi-level and multi-objective optimization model, various resources of the urban drainage system can be effectively integrated, green drainage measures can be scientifically and reasonably laid out, flood control efficiency can be significantly improved, economic cost can be reduced, and urban environmental quality can be improved, thereby providing strong support for the sustainable development of the city. BRIEF DESCRIPTION OF DRAWINGS

[0056] The above and / or additional aspects and advantages of the application will become apparent and more readily appreciated from the following description, taken in conjunction with the following drawings, in which:

[0057] Figure 1 FIG. 1 is a flowchart of a method for controlling and optimizing the vertical layout of green measures in urban drainage according to an embodiment of the application;

[0058] Figure 2 FIG. 2 is an exemplary architecture used by the method for controlling and optimizing the vertical layout of green measures in urban drainage according to an embodiment of the application;

[0059] Figure 3 FIG. 3 is an exemplary network architecture of a Deeplabv3+ model according to an embodiment of the application;

[0060] Figure 4A schematic diagram of a preliminary land use type segmented image and a fine land use type segmented image in an embodiment of the present application;

[0061] Figure 5 A flowchart of a process for simulating the runoff transfer and interaction calculation between the multi-layer green measures and the underground pipe network in an embodiment of the present application;

[0062] Figure 6 A schematic diagram of a vertical water dynamic coupling simulation model fusing the gray infrastructure and the green measures in an embodiment of the present application;

[0063] Figure 7 A schematic diagram of determining the aboveground storage SS in an embodiment of the present application;

[0064] Figure 8 A schematic diagram of determining the underground storage US in an embodiment of the present application;

[0065] Figure 9 A flowchart of determining the layout priority coefficient of each type of green measure in each catchment area in an embodiment of the present application;

[0066] Figure 10 A comprehensive display diagram of the layout priority of each green measure type of each catchment area of the to-be-drained managed city in an embodiment of the present application;

[0067] Figure 11 A principle block diagram of a vertical layout control optimization device of the urban drainage green measures in an embodiment of the present application;

[0068] Figure 12 A structural schematic diagram of a computer device in an embodiment of the present application. DETAILED DESCRIPTION

[0069] Embodiments of the present application are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference signs represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the accompanying drawings are exemplary, only for explaining the present application, and cannot be interpreted as a limitation on the present application.

[0070] It will be further understood that the terms "comprises" and / or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. As used herein, "connected," "coupled," and / or "coupling," can include both direct connections and / or indirect connections (i.e., via one or more other elements). As used herein, "connection" or "coupling" can include a wireless connection or a wireless coupling. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0071] Those skilled in the art will appreciate that unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the specification and relevant art and should not be interpreted in an overly legal sense unless expressly so defined herein.

[0072] Those skilled in the art will understand that, as used herein, the terms "client," "terminal," and "terminal device" include both devices that are solely wireless signal receivers and devices that have both receiving and transmitting hardware that can communicate bi-directionally over a bi-directional communication link. Such devices can include cellular or other communication devices with single-line or multiple-line displays, or no display, Personal Communications Service (PCS) devices that can combine a voice and / or data processor, a PDA that can include a radio frequency receiver and a pager, Internet and / or Intranet access, a Web browser, a calendar, and / or a GPS receiver, a conventional laptop and / or palmtop computer and other devices that have a radio frequency receiver. As used herein, the terms "client," "terminal," and "terminal device" can be portable, transportable, installed in a vehicle (aeronautical, maritime, and / or land), or adapted and / or configured for local and / or distributed operation on Earth and / or any other location in space. As used herein, the terms "client," "terminal," and "terminal device" can also be a communication terminal, an Internet terminal, a music / video playing terminal, such as a PDA, a Mobile Internet Device (MID), and / or a mobile phone with music / video playing function, a smart television, a set-top box, and / or the like.

[0073] As used herein, the terms "server," "client," "service node," and the like refer to hardware that has the equivalent capability of a personal computer, i.e., an electronic device having a central processing unit (including an arithmetic unit and a controller), a memory, an input device, and an output device, and the like necessary components disclosed by the Von Neumann principle. A computer program is stored in the memory, the central processing unit loads the program stored in the external memory into the memory and runs it, executes the instructions in the program, and interacts with the input and output devices, thereby completing a specific function.

[0074] It should be noted that the concept of "server" in the present application can also be extended to the case of a server cluster. According to the principle of network deployment understood by those skilled in the art, the servers should be logically divided, and in physical space, these servers can be independent of each other but can be called through an interface, or can be integrated into a physical computer or a computer cluster. Those skilled in the art should understand this variation and should not be restricted by the implementation of the network deployment of the present application.

[0075] One or more technical features of the present application, unless explicitly specified, can be deployed on a server and accessed by a client remotely calling an online service interface provided by the server, or can be directly deployed and run on a client to implement access.

[0076] The neural network model referred to or possibly referred to in the present application, unless explicitly specified, can be deployed on a remote server and remotely called by a client, or can be deployed on a client with sufficient device capability for direct calling. In some embodiments, when it runs on a client, its corresponding intelligence can be obtained through transfer learning to reduce the requirement for client hardware running resources and avoid excessive occupation of client hardware running resources.

[0077] The various data involved in the present application, unless explicitly specified, can be stored remotely on a server or stored locally on a terminal device, as long as it is suitable for being called by the technical solutions of the present application.

[0078] Those skilled in the art should know that the various methods of the present application, although based on the same concept and described to present commonality between them, are independently executable unless otherwise specified. Similarly, for each embodiment disclosed in the present application, it is based on the same inventive concept, so the same concept is understood to be equivalent, and although the concept is expressed differently, it is only for convenience and is appropriately transformed.

[0079] Unless it is explicitly stated that the embodiments disclosed in the present application are mutually exclusive, the technical features involved in each embodiment can be combined flexibly to construct new embodiments, as long as such combination does not deviate from the spirit of the present application and can meet the needs of the prior art or solve some deficiencies in the prior art. For this variation, those skilled in the art should know.

[0080] Please refer to Figure 1 and Figure 2 The vertical layout control optimization method of the urban drainage green measure of the present application includes, in one embodiment thereof:

[0081] In response to an instruction of optimizing control of green measures to be arranged in the city to be managed for drainage, a preset semantic segmentation model is used to perform land identification segmentation on the remote sensing image of the city to be managed for drainage, and a preset vertical feasibility evaluation module is called to perform vertical arrangement feasibility evaluation of the green measures on the remote sensing image of the city to be managed for drainage, so as to determine the region of the green measures to be arranged in each catchment area, wherein the city to be managed for drainage includes a plurality of catchment areas, and the catchment area includes a plurality of regions of the green measures to be arranged;

[0082] The urban drainage green measure vertical arrangement control system in the terminal device can respond to an instruction of optimizing control of green measures to be arranged in the city to be managed for drainage, use a preset semantic segmentation model to perform land identification segmentation on the remote sensing image of the city to be managed for drainage, and call a preset vertical feasibility evaluation module to perform vertical arrangement feasibility evaluation of the green measures on the remote sensing image of the city to be managed for drainage, so as to determine the region of the green measures to be arranged in each catchment area, wherein the city to be managed for drainage includes a plurality of catchment areas, and the catchment area includes a plurality of regions of the green measures to be arranged, and the basic network architecture of the semantic segmentation model is a Deeplabv3+ model or the like; the region of the green measures to be arranged includes one or any combination of a commercial building, an office building, a hospital, a school, a hotel, a traffic building, a sports building, and a residential community.

[0083] In some embodiments, the remote sensing image of the city to be managed for drainage is segmented based on a semantic segmentation model, and land use of the city to be managed for drainage is automatically identified and geometrically quantified, wherein the basic network architecture of the semantic segmentation model is a Deeplabv3+ model or the like.

[0084] This embodiment takes the Deeplabv3+ model as an example, and does not limit the present application. A data set for training the Deeplabv3+ model, processing prior knowledge masks, and a study area needs to be collected in advance. The training of the Deeplabv3+ model depends on the land cover data set LoveDA provided by a first university. The data set contains remote sensing images and corresponding label images. The LoveDA data set covers a total of 5987 images in multiple cities in a certain country. Each image has a resolution of 0.3 meters and a size of 1024x1024 pixels. The study area is located in the south-central part of a city, covers an area of about 5012.19 hectares, has an elevation of 1497 to 1882 meters, and the elevation decreases from south to northeast.

[0085] Further, in order to train and verify the segmentation model, image samples of urban areas are selected from the data set. The land use types represented by these samples are carefully divided into seven categories: vacant land, building, road, water system, wasteland, green land and farmland. Prior knowledge masks combine the data of Open Street Map and the data set of the second university. OSM is mainly used to refine and supplement the information of roads, water systems and buildings, and the data set of the second university focuses on the improvement of forest and green land data. In addition, the high-resolution satellite map used for research area recognition segmentation is derived from Google Maps, and the image resolution is 0.275x0.275 meters.

[0086] In the training process of the Deeplabv3+ model, the "imageDataAugmenter" function of Matlab is used to implement two data enhancement techniques: random reflection and random rotation. After data enhancement, the size of the training set and the validation set is expanded to 2082 and 462 images respectively. The model training uses the stochastic gradient descent optimizer (SGDM) with momentum, and the momentum value is set to 0.9. In terms of training parameters, the initial learning rate is set to 0.001, the maximum number of iterations is 30, and the mini-batch size is 1. The experiment is completed on a Windows 10 system equipped with an Intel Core i5-10200H CPU and an NVIDIA GeForce RTX 2060 GPU (6GB video memory).

[0087] In this embodiment, the Deeplabv3+ algorithm is used to automatically recognize and segment urban land use, and obtain land use type, area and spatial distribution information.

[0088] Please refer to Figure 3 , the Deeplabv3+ model is composed of an encoder and a decoder. The encoder mainly includes a deep convolutional neural network (DCNN) using a ResNet-18 feature extraction network and an atrous spatial pyramid pooling module (ASPP) for fusing multi-scale information. The features extracted by ResNet-18 are sent to ASPP, which captures semantic information of objects at different scales through parallel convolution layers, and then performs feature splicing and fusion. In the decoder, the expanded 4 times high-level feature map is spliced with the reduced dimension low-level feature map, and then the features are fused, and the subsequent output segmentation image is output.

[0089] Based on the constructed Deeplabv3+ model, the data set is constructed, the model is trained and tested, and after the performance indicators of the model meet the standards, the remote sensing image of the study area is input, and the preliminary segmentation image of the land use type of the study area is obtained. The local initial segmentation result of the drainage management urban land use type is as follows Figure 4As shown in (a) of FIG. 1, it can be seen that the Deeplabv3+ model can accurately identify most land types, but is insufficient in sensitivity when dealing with satellite maps with complex details and chaotic edges, and the boundaries of the identified and segmented images are relatively fuzzy. By utilizing the prior knowledge mask (OSM mask data) obtained through previous processing, the preliminary segmented image is corrected and supplemented in detail to obtain a more refined land use type segmentation image, and a local map of the obtained result is as shown in (b) of FIG. 1. Figure 4

[0090] Further, based on GIS attribute extraction and spatial connection analysis, the vertical layout feasibility of green measures is evaluated, and the building and its surrounding land type information related to the vertical layout are obtained.

[0091] The fine processed image is imported into GIS software, from which building information is extracted, and a POI data set containing building attributes is synchronously imported. Through geographic registration, the specific types of buildings are identified, including 8 types of commercial buildings, office buildings, hospitals, schools, hotels, transportation buildings, sports buildings and residential communities. On this basis, 5 types of buildings suitable for implementing green measures, including commercial buildings, office buildings, hospitals, schools and hotels, are selected, and a buffer zone is created as the layout area of the green measures based on the boundaries of the buildings. Finally, the fine land use segmentation image is cropped according to the contour of the buffer zone to obtain the type distribution and area of the land use around the buildings, and spatial extraction and data statistics are performed.

[0092] In further embodiments, three performance indicators are used to evaluate the performance of the Deeplabv3+ model, and the performance indicators used are described in detail as follows, which include:

[0093] 1. Pixel Accuracy (PA), that is, the proportion of the number of correctly classified pixels to the total number of pixels, and its expression is:

[0094]

[0095] 2. Mean Pixel Accuracy (MPA), that is, the average value of the accuracy of each type of land use, and its expression is:

[0096]

[0097] 3. Intersection over Union (IoU), which reflects the degree of coincidence between the image segmentation prediction result and the original image true value, and its expression is:

[0098]

[0099] ​Wherein, the true positive (TP) represents the pixel of the actual land use type correctly predicted by the model; the false positive (FP) represents the pixel of the actual land use type incorrectly predicted by the model; the false negative (FN) represents the pixel of the actual non-land use type incorrectly predicted by the model; the true negative (TN) represents the pixel of the actual non-land use type correctly predicted by the model; A P represents the pixel area of the land use type predicted by the model; A r represents the pixel area of the actual land use type.

[0100] In some embodiments, the green measure type includes green roof, rainwater garden and permeable pavement; the green measure area to be arranged includes one or any multiple of commercial building, office building, hospital, school, hotel, transportation building, sports building and residential community; the vertical coordination optimization algorithm of the green measure is a multi-objective Bayesian optimization algorithm; and the optimal green measure arrangement scheme represents the green measure type of the green measure area to be arranged in each catchment area and the corresponding optimal green measure arrangement area obtained by solving the Pareto frontier of the total outflow of the green measure and the total cost of the green measure arrangement in the to-be-drainage-managed city.

[0101] In step S20, a vertical water dynamic coupling simulation model integrating the gray measure and the green measure is called, and a preset rainfall is input into the vertical water dynamic coupling simulation model to determine the waterlogging accumulation and the green measure accumulation of the to-be-drainage-managed city corresponding to the to-be-drainage-managed city.

[0102] After the land identification segmentation of the remote sensing image of the to-be-drainage-managed city is performed by using the preset semantic segmentation model and the vertical arrangement feasibility of the green measure of the remote sensing image of the to-be-drainage-managed city is evaluated by using the preset vertical feasibility evaluation module to determine the green measure area to be arranged in each catchment area, a vertical water dynamic coupling simulation model integrating the gray measure and the green measure is called, and a preset rainfall is input into the vertical water dynamic coupling simulation model to determine the waterlogging accumulation and the green measure accumulation of the to-be-drainage-managed city corresponding to the to-be-drainage-managed city.

[0103] In some embodiments, the step of calling the vertical water dynamic coupling simulation model integrating the gray measure and the green measure, inputting the preset rainfall into the vertical water dynamic coupling simulation model to determine the waterlogging accumulation and the green measure accumulation of the to-be-drainage-managed city corresponding to the to-be-drainage-managed city, includes:

[0104] In step S201, the rainfall intensity corresponding to the to-be-drainage-managed city is determined according to a preset rainfall intensity calculation formula, and the rainfall corresponding to the to-be-drainage-managed city is determined according to a fourth product between the rainfall intensity corresponding to the to-be-drainage-managed city and a preset rainfall duration.

[0105] Step S202, input the rainfall into the preset vertical water dynamic coupling simulation model to determine the corresponding waterlogging water accumulation and green measure detention of the water management city, wherein the waterlogging water accumulation includes green measure detention and residual flood flow, and the green measure detention includes aboveground detention and underground detention.

[0106] Specifically, the vertical water dynamic coupling simulation model integrating gray measures and green measures represents a vertical water dynamic coupling simulation model integrating gray infrastructure (such as drainage pipe network) and green measures, and a preset rainfall is input into the vertical water dynamic coupling simulation model to determine the corresponding waterlogging water accumulation and green measure detention of the water management city. Please refer to Figure 5 and Figure 6 wherein, Figure 5 is a flowchart for simulating runoff transfer and interaction calculation between multi-layer green measures and underground pipe network, Figure 6 is a vertical water dynamic coupling simulation model integrating gray infrastructure (such as drainage pipe network) and green measures; the vertical water dynamic coupling simulation model integrating gray infrastructure (such as drainage pipe network) and green measures is constructed to simulate runoff transfer and interaction calculation between multi-layer green measures and underground pipe network, and the general formula is as follows:

[0107] P = SI + D + SS + US + RFV,

[0108] wherein, P represents rainfall, which is the input of the entire vertical water dynamic coupling simulation model; SI represents soil infiltration, D represents pipe network discharge, SS represents aboveground detention, US represents underground detention, and RFV represents residual flood flow. The rainfall intensity can be simulated to change with time at different return periods according to the storm intensity calculation formula of the water management city, wherein the expression of the storm intensity calculation formula corresponding to the water management city is as follows:

[0109]

[0110] wherein, q represents rainfall flow per hectare per second, with the unit of (L / s) / hm 2 ); P represents the return period of storm occurrence, indicating the average number of years of a specific intensity of storm event in a certain period of time, and the value of P can be 50, with the unit of year; t represents the duration of rainfall event, with the unit of minute;

[0111] Rainwater first undergoes natural infiltration process of underlying surface soil, and the soil infiltration SI of the region is calculated according to the Horton infiltration formula as follows:

[0112]

[0113] wherein f min is the minimum infiltration rate, f max is the maximum infiltration rate, f min and f max can be obtained by the empirical value of the soil composition combined with the infiltration coefficient, K d is the attenuation coefficient, which can be obtained by empirical method, and i is the rainfall rate.

[0114] Further, the rainwater exceeding the soil infiltration capacity flows into the pipe network system, and the pipe network discharge D is calculated by the difference between the node inflow NIF and the waterlogging accumulation TFV (i.e. the total overflow of the system), and the formula is as follows:

[0115] D = NIF - TFV,

[0116] wherein NIF is the node inflow, and TFV is the waterlogging accumulation, which can be obtained by the hydrodynamic simulation model.

[0117] Further, referring to Figure 7 , the waterlogging accumulation is introduced into the aboveground detention facility, the outflow of the green measures of the upper building is introduced into the green measures of the lower building, the vertical runoff connection and conversion are performed through the green measures of different building floors, so that the runoff at different heights in the same plot can experience iterative calculation of multiple green measures, and the expression of the aboveground detention volume SS includes:

[0118]

[0119] wherein Q SS1(j) represents the inflow rate of the jth floor building, q SS1(j) represents the overflow rate of the jth floor building, e SS1(j) represents the evaporation rate of the jth floor building, A j represents the green measure area of the jth floor building, and SS represents the aboveground detention volume, represents the area ratio of different floors, which is used for normalizing the flow; q SS1(j) , e SS1(j) can adopt the calibrated value in the SWMM model.

[0120] Further, referring to Figure 8 , the water exceeding the aboveground detention capacity will overflow into the plot where the underground infiltration facility is arranged with green measures, i.e. the final outflow q SS1(1) of the aboveground detention facility is introduced into the plot where the underground infiltration facility is arranged with green measures, and the water quantity is further transferred and converted through the structure layer of different types of measures. The successfully intercepted water quantity is calculated according to the LID control unit, and the expression of the underground detention volume US is as follows:

[0121]

[0122] wherein q SS1(1) represents the final outflow of the above-ground storage area, A ss represents the green measure area of the above-ground outflow layer, A US represents the green measure area of the underground infiltration storage facility, q1 represents the surface layer overflow rate of the underground infiltration storage facility, q4 represents the water storage layer outflow rate of the underground infiltration storage facility, e1 represents the surface layer evaporation rate of the underground infiltration storage facility, e2 represents the paving layer evaporation rate of the underground infiltration storage facility, e3 represents the soil layer evaporation rate of the underground infiltration storage facility, and e4 represents the water storage layer evaporation rate of the underground infiltration storage facility.

[0123] In some embodiments, the SWMM model of the water drainage management city can include 136 catchment areas, 117 pipe network nodes, 117 pipes and 17 drainage outlets. Other input information includes: catchment area and slope percentage, calculated by ArcGIS tool on fine land classification image and DEM data; impermeable percentage, which is the proportion of impervious area in the fine land classification result, obtained by weighted average according to different land use types and their corresponding parameters, combined with area; the input information of the pipe mainly includes pipe diameter, pipe length and node height.

[0124] After running the SWMM model, the output catchment area outflow is compared with the vertical layout control optimization method of the city drainage green measure of the application. In the simulation scene of a single measure, the results of the model of the method and the SWMM model are highly consistent, and the fitting degrees of the model of the method under the rainwater garden, green roof and permeable pavement three measures are 0.773, 0.999 and 0.971 respectively; the synergistic effect of multiple measures in the vertical simulation scene, the fitting degree index of the model of the method reaches 0.924, the simulation performance is good, and the vertical layout of multiple measures not only achieves a substantial reduction in outflow, but also significantly delays the outflow time.

[0125] Step S30, using GIS geographic analysis technology to determine a plurality of layout evaluation indexes according to the region to be laid green measure, the waterlogging water storage amount, the green measure storage amount and the population information of the water drainage management city, and based on the fuzzy comprehensive evaluation method, according to the plurality of layout evaluation indexes and the weight corresponding to each layout evaluation index, to determine the layout priority coefficient of each green measure type in each catchment area;

[0126] The vertical water dynamic coupling simulation model of the fusion of the gray measures and the green measures is called, preset rainfall is input into the vertical water dynamic coupling simulation model, after the waterlogging accumulated water quantity and the green measure lag storage quantity corresponding to the to-be-drained managed city are determined, GIS geographic analysis technology is used to determine a plurality of layout evaluation indexes according to the to-be-laid green measure region, the waterlogging accumulated water quantity, the green measure lag storage quantity and population information of the to-be-drained managed city, and the layout priority coefficient of each green measure type in each catchment area is determined based on the fuzzy comprehensive evaluation method according to the plurality of layout evaluation indexes and the weight corresponding to each layout evaluation index; wherein the green measure type includes a green roof, a rainwater garden, permeable pavement and the like.

[0127] In some embodiments, the step of determining the layout priority coefficient of each green measure type in each catchment area based on the fuzzy comprehensive evaluation method according to the plurality of layout evaluation indexes and the weight corresponding to each layout evaluation index comprises:

[0128] In step S301, a plurality of layout evaluation indexes and the weight corresponding to each layout evaluation index are obtained, wherein the layout evaluation index includes one or any multiple of land use nature, vertical layout suitability, building density, population density and waterlogging risk.

[0129] In step S302, based on the fuzzy comprehensive evaluation method, the difference between the maximum value and the minimum value of the layout evaluation index is used to determine a plurality of evaluation grades of the layout evaluation index in each catchment area, and a preset membership function is used to normalize each layout evaluation index into a membership value in the interval [0, 1] to determine the normalized membership value of the layout evaluation index in each catchment area.

[0130] In step S303, according to the normalized membership value of the layout evaluation index in each catchment area and the plurality of evaluation grades of the layout evaluation index in each catchment area, a membership matrix corresponding to each catchment area is constructed.

[0131] In step S304, the importance weight of each layout evaluation index is determined by using the analytic hierarchy process, and the third product between the membership matrix and the importance weight of each layout evaluation index is used to determine the layout priority coefficient of each green measure type in each catchment area.

[0132] Specifically, the GIS geographic analysis technology is adopted to construct an evaluation set including five types of layout evaluation indexes, i.e., land use property, vertical layout suitability, waterlogging risk, building density and population density, according to the region to be laid out with green measures, the waterlogging accumulation, the green measure accumulation and the population information of the city to be drained, and each layout evaluation index is divided into five levels, i.e., I to V, and a higher priority means a better layout condition and should be considered preferentially, wherein the land use property (LUP) means that the layout potential of each type of measure in different regions is obtained by assigning corresponding weights to each type of land according to the applicability of the green measure in different land types and then performing weighted calculation in combination with the land area; the vertical layout suitability (VLS) means that the feasibility of vertical layout is determined by performing detailed analysis on specific measures based on the area of land around the building; the building density (BD) means that the urgency of laying out measures is quantitatively evaluated by calculating the proportion of buildings in the area of catchment; the population density (PD) means that the area per capita of each catchment is calculated in combination with the population database; and the waterlogging risk (WR) means that the node overflow flow of the regional pipe network system is calculated according to a water dynamic model, and the total overflow flow of the system is counted to reflect the degree of waterlogging risk.

[0133] Further, please refer to Figure 9 , based on the fuzzy comprehensive evaluation method, each layout evaluation index is normalized to calculate the layout priority coefficient of each green measure type in each catchment. The specific process is as follows:

[0134] Step S210, each layout evaluation index is normalized to a membership value in the interval [0, 1] by using a preset membership function, and five intervals are divided according to the difference between the maximum value and the minimum value of the layout evaluation index to determine the evaluation levels of the layout evaluation index in each catchment. On this basis, a membership matrix corresponding to each catchment is constructed according to the normalized membership value of the layout evaluation index in each catchment and the evaluation levels of the layout evaluation index in each catchment, wherein the membership matrix is an evaluation fuzzy matrix R, and the expression of the evaluation fuzzy matrix R is:

[0135]

[0136] wherein r ij is an element in the evaluation fuzzy matrix R, which represents the evaluation level membership value of each catchment under different layout evaluation indexes; m represents the number of evaluation indexes; and n represents the number of evaluation levels.

[0137] Step S220, the weights of each layout evaluation index are determined by using the analytic hierarchy process (AHP) through literature review and expert consultation to construct a weight matrix W (w1, w2, …, w i), wherein the weights of land use, vertical layout suitability, waterlogging risk, building density and population density are 0.195, 0.462, 0.195, 0.074 and 0.074 respectively.

[0138] Step S230, finally, the above steps are fused to construct a priority multi-index evaluation system for vertical layout of green measures, and a layout priority coefficient P of the green measures is calculated to quantify the layout priority of each green measure type in each catchment area, wherein the number of layout evaluation indexes and the number of evaluation grades of the three green measure types are the same, and the expression of the layout priority coefficient P is:

[0139] P = W x R = (P1, P2, …, P3),

[0140] wherein W represents a weight matrix containing the weight values of each layout priority index, and P represents the layout priority coefficient of the green measures, which represents the layout priority coefficient of each type of green measures in each catchment area.

[0141] Through the above steps, the layout priority coefficients of green roofs, rainwater gardens and permeable pavements in each catchment area can be calculated and determined, and the comprehensive layout priority coefficients and priority levels in spatial layout of the three types of green measures are as shown in Figure 10 , wherein the lower the layout priority coefficient of each type of green measures in each catchment area, the more priority the type of green measures has.

[0142] Step S40, the green measure parameters corresponding to the to-be-laid green measure area of the to-be-drained management city are obtained, and a preset vertical coordination optimization algorithm of green measures is adopted to obtain the Pareto optimal solution set of the total outflow of green measures and the total cost of green measure layout of all catchment areas in the to-be-drained management city according to the layout priority coefficient and the to-be-laid green measure parameters, and determine the optimal green measure layout scheme of the to-be-laid green measure area in each catchment area.

[0143] The GIS geographic analysis technology is adopted to determine a plurality of layout evaluation indexes according to the green measure region to be laid, the waterlogging water volume, the green measure storage volume and the population information of the drainage management city, the fuzzy comprehensive evaluation method is adopted to determine the layout priority coefficient of each green measure type in each catchment area based on the plurality of layout evaluation indexes and the weight corresponding to each layout evaluation index, the green measure parameter corresponding to the green measure region to be laid of the drainage management city is obtained, the preset green measure vertical collaborative optimization algorithm is adopted to obtain the Pareto optimal solution set of the total outflow of the green measure and the total cost of the green measure layout in all catchment areas of the drainage management city according to the layout priority coefficient and the green measure parameter to be laid, and the optimal green measure layout scheme of the green measure region to be laid in each catchment area is determined, wherein the green measure vertical collaborative optimization algorithm is a multi-objective Bayesian optimization algorithm; the optimal green measure layout scheme represents that the green measure type of the green measure region to be laid in each catchment area and the corresponding optimal green measure layout area are obtained by solving the Pareto frontier of the total outflow of the green measure and the total cost of the green measure layout in the drainage management city.

[0144] In some embodiments, the step of determining the total outflow of the green measure comprises:

[0145] Step S401: obtaining the number of catchment areas of the drainage management city, the number of green measure types, the unit time outflow of the green measure of each catchment area and the green measure area of each catchment area;

[0146] Step S402: calculating and determining the first product between the unit time outflow of each green measure type of each catchment area and the outflow green measure area of each green measure type of each catchment area to determine the green measure outflow water volume of each green measure type of each catchment area.

[0147] Step S403: determining the total outflow of the green measure of all catchment areas of the drainage management city according to the green measure outflow water volume of each green measure type of each catchment area, the number of catchment areas of the drainage management city and the number of green measure types.

[0148] Specifically, the expression of the total outflow of the green measure is:

[0149]

[0150] wherein Q represents the green measure outflow water volume of all catchment areas of the drainage management city, the unit of which is m 3 , n represents the number of catchment areas of the drainage management city, the value of n can be 136 or the like; m represents the number of green measure types, the value of m can be 3 or the like; q represents the unit time outflow of the green measure of each catchment area, the value of q can be 0.0001 or the like; and A represents the green measure area of each catchment area, the value of A can be 0.0001 or the like. 3 , n represents the number of catchment areas of the drainage management city, the value of n can be 136 or the like; m represents the number of green measure types, the value of m can be 3 or the like; qlid(ij) represents the outflow of the i-th catchment area j-th green measure per unit time, with the unit of mm / h; A lid(ij) represents the outflow green measure area of the i-th catchment area j-th green measure, with the unit of m 2 .

[0151] In some embodiments, the step of determining the total cost of green measure arrangement includes:

[0152] Step S4001, obtaining the number of catchment areas of the to-be-drained city, the number of green measure types, the arrangement priority coefficient of each type of green measure in each catchment area, the unit area cost of each type of green measure, and the arrangement area of each type of green measure in each catchment area;

[0153] Step S4002, calculating a second product between the arrangement priority coefficient of each type of green measure in each catchment area, the unit area cost of each type of green measure, and the arrangement area of each type of green measure in each catchment area;

[0154] Step S4003, determining the total cost of green measure arrangement of all catchment areas of the to-be-drained city according to the number of catchment areas of the to-be-drained city, the number of green measure types, and the second product.

[0155] Specifically, the expression of the total cost of green measure arrangement is:

[0156]

[0157] Wherein, C represents the total cost of arranging all green measures in all catchment areas of the to-be-drained city, n represents the number of catchment areas of the to-be-drained city, m represents the number of green measure types, and the value of m can be 3 or the like; P ij represents the arrangement priority coefficient of the i-th catchment area j-th green measure, C j represents the unit area cost of the j-th green measure, with the unit of yuan / m 2 ; A ij represents the arrangement area of the i-th catchment area j-th green measure, with the unit of m 2 .

[0158] In some embodiments, the step of determining the optimal green measure arrangement scheme of the to-be-arranged green measure area in each catchment area according to the arrangement priority coefficient and the to-be-arranged green measure parameters includes:

[0159] Step S40001, a preset green measure vertical collaborative optimization algorithm is called, and a total outflow of green measures of all catchment areas and a total cost of green measure arrangement are taken as double optimization objectives according to the arrangement priority coefficient and the to-be-arranged green measure parameter, a constraint condition is taken as the arrangeable area of different green measure types in each catchment area, a Gaussian regression model is used to perform probability modeling according to a nonlinear update between each green measure parameter to be arranged in each catchment area and a response value of the objective function, so as to predict an expected mean value and a variance of an objective function value of a potential sample point corresponding to the to-be-arranged green measure area, wherein the to-be-arranged green measure parameter represents a green measure sample point corresponding to each green measure type to be arranged in each catchment area;

[0160] Step S40002, an expected hyper volume improvement algorithm is used to calculate a collection score of each potential sample point according to the probability distribution of the Gaussian regression model, and a sample point with the highest score is selected as a target of next function evaluation;

[0161] Step S40003, the above steps are repeated, a Pareto optimal solution set satisfying the constraint condition is iteratively generated by quantifying an expected gain of a new sample on a hyper volume measure of an existing Pareto frontier, so as to determine an optimal green measure arrangement scheme of the to-be-arranged green measure area in each catchment area.

[0162] Specifically, first, different green measure types are determined according to each catchment area in the to-be-drained management city, for example, a rainwater garden, a green roof, permeable pavement and the like, and an arrangeable area of each green measure type in different catchment areas is defined, wherein the arrangeable area of each catchment area is affected by geographical location, land use and the like. For the to-be-arranged green measure parameter of each catchment area, such as the area of a rainwater garden, permeable pavement and green roof, a green measure area corresponding to each green measure type to be arranged in each catchment area is predicted by a Gaussian process regression model. The Gaussian process regression can calculate an expected mean value and a variance of an objective function value of a potential to-be-arranged green measure sample point corresponding to the to-be-arranged green measure area by correlating the measured data and the predicted data. Through the Gaussian process regression model, the expected mean value and the variance of the objective function value of the to-be-arranged green measure sample point corresponding to the to-be-arranged green measure area are obtained as inputs of a multi-objective Bayesian optimization algorithm, wherein the green measure arrangement collaborative optimization algorithm is a multi-objective Bayesian optimization algorithm.

[0163] The preset green measure layout collaborative optimization algorithm is called, different green measure types in each catchment area are taken as constraint conditions, a Gaussian process regression model is used to perform probability modeling according to nonlinear updating between each green measure parameter to be laid out in each catchment area and a response value of a target function, to predict expected mean and variance of a target function value of a potential green measure sample point corresponding to the green measure area to be laid out, and then an expected hyper volume improvement algorithm is used to calculate a collection score of each potential sample point according to the probability distribution of the Gaussian regression model, and the sample point with the highest score is selected as a target for next function evaluation; after the expected mean and variance of the green measure parameter corresponding to the green measure area to be laid out are obtained through the Gaussian process regression model, the expected hyper volume improvement algorithm is used to calculate a collection score of each green measure area to be laid out according to the expected mean and variance of the green measure parameter corresponding to the green measure area to be laid out, and the green measure area to be laid out with the highest score is selected as a target for next function evaluation.

[0164] More specifically, the expected hyper volume improvement (EHVI) algorithm is used to evaluate the “collection score” of each potential green measure area to be laid out, in combination with the predicted expected mean and variance of the target function value of the potential green measure sample point corresponding to the green measure area to be laid out. The expected hyper volume improvement (EHVI) algorithm selects the area with the maximum improvement (hyper volume) to perform next layout evaluation. According to the calculation result, the sample point with the highest collection score is selected as a target for next function evaluation. The green measure layout scheme of the area involved in the sample point is considered as the area that is most likely to optimize the flow and cost at present.

[0165] According to the foregoing steps S40001 and S40003, the multi-objective Bayesian optimization algorithm is repeatedly executed, in each iteration, a new green measure layout scheme is adjusted based on the current flow and cost prediction result, to minimize the total flow and layout cost, and through multiple iterations, the green measure layout scheme is gradually adjusted, until the Pareto optimal solution set of the total flow and layout total cost of all catchment areas is solved. In each iteration, the Bayesian optimization algorithm is used to further adjust the parameters according to the feedback of the two target functions, i.e., the total flow function of the green measure and the total cost function of the green measure layout, to ensure the convergence of the global optimal solution, so as to determine the optimal green measure layout scheme of the green measure area to be laid out in each catchment area. The optimal green measure layout scheme represents the optimal green measure type and the corresponding optimal green measure area of the green measure area to be laid out in each catchment area selected from the Pareto optimal solution set in which the total flow and layout total cost of the green measure in the drainage management city meet the constraint conditions.

[0166] From the above embodiments, compared with the prior art, the present application is aimed at the problem that the prior art focuses on two-dimensional plane optimization and cannot fully utilize urban vertical space resources, resulting in that the flood control potential is not fully tapped, and the present application includes but is not limited to the following beneficial effects:

[0167] Firstly, the present application can significantly improve the flood control efficiency of the urban drainage system. By constructing a vertical runoff transmission model linked with multiple layers of buildings on the ground and underground, the urban rainwater flow process can be more comprehensively simulated, breaking through the limitations of traditional two-dimensional plane models, thereby significantly improving the storage capacity. This multi-layer linked modeling method can accurately predict the runoff situation after rainfall, helping to develop more effective drainage schemes and reduce urban waterlogging and flood risks.

[0168] Secondly, the present application can optimize resource allocation and measure implementation efficiency. By comprehensively considering multiple indicators such as land use nature, vertical layout suitability, waterlogging risk, building density, and population density, and quantifying priorities through fuzzy evaluation method, the areas that need to be prioritized can be accurately identified, which not only helps to improve the scientificity of resource allocation, but also improves the efficiency and effectiveness of measure implementation.

[0169] Thirdly, the present application can balance flood control and economic cost. By using a multi-objective Bayesian optimization algorithm to generate a Pareto optimal solution, the balance between flood control efficiency (such as minimizing total outflow) and economic cost (such as minimizing layout cost) can be achieved. This method can provide multiple optimization schemes for decision-makers to flexibly select the best scheme according to actual budget and demand, thereby ensuring the best flood control effect under limited resources.

[0170] Fourthly, the present application predicts the effect of green measures to be laid out (such as total outflow and layout cost) through a Gaussian process regression model, and selects the optimal area for green measure layout through an expected hyper volume improvement algorithm. This dynamic optimization method can real-time adjust and update the green measure layout scheme, thereby realizing the flexibility and adaptability of urban drainage management.

[0171] Fifthly, the present application can reduce the operation cost of the drainage system and environmental impact. By optimizing the layout of green measures, the outflow of the drainage system can be effectively reduced, the probability of waterlogging and flooding can be reduced, and the construction and maintenance cost of the drainage system can be reduced by reducing excessive infrastructure construction (such as expensive underground drainage pipe network). At the same time, the vertical layout optimization of green measures such as rainwater gardens and permeable pavement can help to improve the urban ecological environment, increase urban green space, and improve the sustainability of the city.

[0172] Further, through the multi-level and multi-objective optimization model, various resources of the urban drainage system can be effectively integrated, green drainage measures can be scientifically and reasonably laid out, flood control efficiency can be significantly improved, economic cost can be reduced, urban environmental quality can be improved, and strong support can be provided for sustainable development of the city.

[0173] Please refer to Figure 11 , which is provided for one of the purposes of the present application, a vertical layout control optimization device for urban drainage green measures, comprising a to-be-laid-out area determination module 1100, a feasibility data determination module 1200, a layout priority determination module 1300, and an optimal scheme determination module 1400. Wherein, the to-be-laid-out area determination module 1100 is set to respond to the instruction of optimizing the layout control of the green measures of the to-be-drainage-managed city, adopts a preset semantic segmentation model to perform land identification segmentation on the remote sensing image of the to-be-drainage-managed city, and calls a preset vertical feasibility evaluation module to perform vertical layout feasibility evaluation on the remote sensing image of the to-be-drainage-managed city, to determine the to-be-laid-out green measure area in each catchment area, wherein the to-be-drainage-managed city comprises a plurality of catchment areas, and the catchment area comprises a plurality of to-be-laid-out green measure areas; the feasibility data determination module 1200 is set to call a vertical water dynamic coupling simulation model that fuses gray measures and green measures, input a preset rainfall into the vertical water dynamic coupling simulation model, to determine the corresponding waterlogging accumulation and green measure detention of the to-be-drainage-managed city; the layout priority determination module 1300 is set to adopt GIS geographic analysis technology to determine a plurality of layout evaluation indexes according to the to-be-laid-out green measure area, the waterlogging accumulation, the green measure detention, and the population information of the to-be-drainage-managed city, and determine the layout priority coefficient of each green measure type in each catchment area based on the fuzzy comprehensive evaluation method according to the plurality of layout evaluation indexes and the weight corresponding to each layout evaluation index; the optimal scheme determination module 1400 is set to obtain the to-be-laid-out green measure parameters corresponding to the to-be-laid-out green measure area of the to-be-drainage-managed city, adopt a preset green measure vertical collaborative optimization algorithm, and obtain the Pareto optimal solution set of the total outflow of the green measures and the total cost of the green measure layout of all catchment areas in the to-be-drainage-managed city according to the layout priority coefficient and the to-be-laid-out green measure parameters, to determine the optimal green measure layout scheme of the to-be-laid-out green measure area in each catchment area.

[0174] On the basis of any embodiment of the present application, please refer to Figure 12 Another embodiment of the present application further provides an electronic device, which can be realized by a computer device, such as Figure 12As shown, the internal structure diagram of the computer device is shown. The computer device includes a processor, a computer readable storage medium, a memory and a network interface connected by a system bus. Among them, the computer readable storage medium of the computer device stores an operating system, a database and computer readable instructions, the database can store control information sequence, the computer readable instructions are executed by the processor, and the processor can realize a kind of urban drainage green measure vertical layout control optimization method. The processor of the computer device is used to provide computing and control capability, support the operation of entire computer device. The memory of the computer device can store computer readable instructions, the computer readable instructions are executed by the processor, and the processor can execute the urban drainage green measure vertical layout control optimization method of the application. The network interface of the computer device is used to connect with terminal communication. Those skilled in the art can understand, Figure 12 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the application, and does not constitute a limitation on the computer device to which the scheme of the application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0175] The processor in the embodiment is used to execute the specific functions of each module in Figure 11 The memory stores the program codes and various data required for executing the above-mentioned modules. The network interface is used for data transmission between user terminals or servers. The memory in the embodiment stores the program codes and data required for executing all modules in the urban drainage green measure vertical layout control optimization device of the application, and the server can call the program codes and data of the server to execute the functions of all modules.

[0176] The application also provides a storage medium storing computer readable instructions, which are executed by one or more processors to make the one or more processors execute the steps of the urban drainage green measure vertical layout control optimization method described in any embodiment of the application.

[0177] The application also provides a computer program product, including computer program / instructions, which are executed by one or more processors to realize the steps of the urban drainage green measure vertical layout control optimization method described in any embodiment of the application.

[0178] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments of the application can be completed by a computer program instructing relevant hardware, and the computer program can be stored in a computer readable storage medium. When the program is executed, the processes of the above-mentioned embodiments of the methods can be included. The storage medium can be a computer readable storage medium such as a magnetic disc, an optical disc, a read-only memory (ROM), or a random access memory (RAM).

[0179] The above only describes some embodiments of the application. It should be pointed out that those skilled in the art can make some improvements and refinements without departing from the principles of the application, and these improvements and refinements should also be considered as the protection scope of the application.

Claims

1. A method for optimizing the vertical layout of urban drainage greening measures, characterized in that, include: In response to the instruction to optimize the deployment of green measures in cities under drainage management, a preset semantic segmentation model is used to perform land identification and segmentation on the remote sensing images of the cities under drainage management, and a preset vertical feasibility assessment module is invoked to perform a vertical deployment feasibility assessment of the remote sensing images of the cities under drainage management, so as to determine the areas to be deployed with green measures in each catchment area. The cities under drainage management include multiple catchment areas, and each catchment area includes multiple areas to be deployed with green measures. The vertical hydrodynamic coupling simulation model that integrates gray and green measures is invoked, and the preset rainfall is input into the vertical hydrodynamic coupling simulation model to determine the corresponding waterlogging volume and green measure retention volume of the city to be managed for drainage. Using GIS geographic analysis technology, multiple deployment evaluation indicators are determined based on the area to be deployed green measures, the amount of waterlogging, the storage capacity of green measures, and the population information of the city to be managed for drainage. Based on the fuzzy comprehensive evaluation method, the deployment priority coefficient of each type of green measure in each catchment area is determined according to the multiple deployment evaluation indicators and the corresponding weights of each deployment evaluation indicator. The parameters of the green measures to be deployed in the areas of the city to be managed for drainage are obtained. A preset green measure vertical collaborative optimization algorithm is used to obtain the Pareto optimal solution set of the total outflow of green measures and the total cost of green measure deployment in all catchment areas of the city to be managed for drainage, based on the deployment priority coefficient and the parameters of the green measures to be deployed. The optimal green measure deployment scheme for the areas of the city to be managed for drainage is then determined.

2. The method for vertical layout control and optimization of urban drainage greening measures according to claim 1, characterized in that, The steps of calling a vertical hydrodynamic coupling simulation model that integrates gray and green measures, and inputting a preset rainfall amount into the vertical hydrodynamic coupling simulation model to determine the corresponding urban flooding volume and green measure retention capacity of the city to be managed include: The rainfall intensity corresponding to the city to be managed for drainage is calculated and determined according to the preset rainstorm intensity calculation formula. The rainfall amount corresponding to the city to be managed for drainage is determined according to the fourth product between the rainfall intensity corresponding to the city to be managed for drainage and the preset rainfall duration. The rainfall is input into a preset vertical hydrodynamic coupling simulation model to determine the corresponding urban flooding volume and green measure retention volume of the city to be managed for drainage. The urban flooding volume includes the green measure retention volume and the remaining flood flow. The green measure retention volume includes the above-ground retention volume and the underground retention volume.

3. The method for vertical layout control and optimization of urban drainage greening measures according to claim 1, characterized in that, The steps for determining the deployment priority coefficient of each type of green measure in each catchment area based on the fuzzy comprehensive evaluation method, according to the multiple deployment evaluation indicators and their corresponding weights, include: Multiple deployment evaluation indicators and their corresponding weights are obtained. The deployment evaluation indicators include one or more of the following: land use, vertical deployment suitability, building density, population density, and waterlogging risk. Based on the fuzzy comprehensive evaluation method, the difference between the maximum and minimum values ​​of the deployment evaluation index is used to determine the multiple evaluation levels of the deployment evaluation index in each catchment area. A preset membership function is used to normalize each deployment evaluation index to a membership value in the interval [0,1] to determine the normalized membership value of the deployment evaluation index in each catchment area. Based on the normalized membership values ​​of the deployment evaluation indicators in each catchment area and the multiple evaluation levels of the deployment evaluation indicators in each catchment area, a membership matrix corresponding to each catchment area is constructed. The importance weights of each deployment evaluation index are determined using the analytic hierarchy process (AHP). The deployment priority coefficient of each type of green measure in each catchment area is determined based on the third product between the membership matrix and the importance weights of each deployment evaluation index.

4. The method for vertical layout control and optimization of urban drainage greening measures according to claim 1, characterized in that, The steps to determine the total outflow from green initiatives include: Obtain the number of catchment areas, the number of green measure types, the outflow rate of green measures per unit time for each catchment area, and the area of ​​green measures for each catchment area in the city to be managed for drainage; Calculate the first product between the unit time outflow of each green measure type in each catchment area and the outflow green measure area of ​​each green measure type in each catchment area to determine the green measure outflow volume of each green measure type in each catchment area; The total outflow of green measures in all catchments of the city to be managed is determined based on the outflow volume of green measures for each type of green measure in each catchment area, the number of catchments in the city to be managed, and the number of green measure types.

5. The method for vertical layout control and optimization of urban drainage greening measures according to claim 1, characterized in that, The steps to determine the total cost of implementing green initiatives include: Obtain the number of catchment areas, the number of green measure types, the deployment priority coefficient of each type of green measure in each catchment area, the unit area cost of each type of green measure, and the deployment area of ​​each type of green measure in each catchment area of ​​the city to be managed for drainage. Calculate and determine the deployment priority coefficient of each type of greening measure in each catchment area, the unit area cost of each type of greening measure, and the second product between the deployment area of ​​each type of greening measure in each catchment area; The total cost of deploying green measures in all watersheds of the city to be managed is determined based on the number of watersheds in the city to be managed, the number of green measure types, and the second product.

6. The method for vertical layout control and optimization of urban drainage greening measures according to claim 1, characterized in that, The steps of using a pre-defined vertical collaborative optimization algorithm for green measures, based on the deployment priority coefficient and the parameters of the green measures to be deployed, to obtain the Pareto optimal solution set of the total outflow of green measures and the total deployment cost of green measures in all catchment areas of the city to be managed for drainage, and to determine the optimal green measure deployment scheme for the areas to be deployed in each catchment area, include: The preset vertical collaborative optimization algorithm for green measures is invoked. Based on the deployment priority coefficient and the parameters of the green measures to be deployed, the algorithm takes the total outflow of green measures in all catchment areas and the total deployment cost of green measures as dual optimization objectives, and the deployable area of ​​different types of green measures in each catchment area as constraints. A Gaussian regression model is used to perform probabilistic modeling based on the nonlinear update between the parameters of each green measure to be deployed in each catchment area and the response value of the objective function, so as to predict the expected mean and variance of the objective function values ​​of the potential sample points corresponding to the area to be deployed green measures. The parameters of the green measures to be deployed represent the green measure sample points corresponding to each type of green measure to be deployed in each catchment area. The preset expected hypervolume improvement algorithm is used to calculate the collection score of each potential sample point according to the probability distribution of the Gaussian regression model, and the sample point with the highest score is selected as the target of the next function evaluation. Repeat the above steps, and iteratively generate a Pareto optimal solution set that satisfies the constraints by quantifying the expected gain of the new samples on the existing Pareto front hypervolume measure, so as to determine the optimal greening measure deployment scheme for the areas to be greened in each catchment area.

7. The method for vertical layout control and optimization of urban drainage greening measures according to any one of claims 1 to 6, characterized in that, The types of green initiatives include green roofs, rain gardens, and permeable paving. The areas where greening measures are to be implemented include one or more of the following: commercial buildings, office buildings, hospitals, schools, hotels, transportation buildings, sports buildings, and residential communities. The vertical collaborative optimization algorithm for the green measures is a multi-objective Bayesian optimization algorithm; The optimal green measure deployment scheme is characterized by solving the Pareto front of the total outflow of green measures and the total cost of green measure deployment in the city to be managed for drainage, thereby obtaining the type of green measures to be deployed in each catchment area and the corresponding optimal deployment area of ​​green measures.

8. A vertical layout control and optimization device for urban drainage greening measures, characterized in that, include: The module for determining the area to be deployed is configured to respond to instructions for optimizing the deployment of green measures in cities under drainage management. It uses a preset semantic segmentation model to perform land identification and segmentation on the remote sensing images of the cities under drainage management and calls a preset vertical feasibility assessment module to perform a vertical deployment feasibility assessment of the remote sensing images of the cities under drainage management, so as to determine the areas to be deployed in each catchment area. The cities under drainage management include multiple catchment areas, and each catchment area includes multiple areas to be deployed in green measures. The feasibility data determination module is set to call a vertical hydrodynamic coupling simulation model that integrates gray measures and green measures, and input the preset rainfall into the vertical hydrodynamic coupling simulation model to determine the corresponding waterlogging volume and green measure retention volume of the city to be managed for drainage. The deployment priority determination module is configured to use GIS geographic analysis technology to determine multiple deployment evaluation indicators based on the area to be deployed green measures, the amount of waterlogging, the amount of water stored by green measures, and the population information of the city to be managed for drainage. Based on the fuzzy comprehensive evaluation method, the deployment priority coefficient of each type of green measure in each catchment area is determined according to the multiple deployment evaluation indicators and the corresponding weights of each deployment evaluation indicator. The optimal solution determination module is configured to obtain the parameters of the green measures to be deployed corresponding to the areas of the cities to be managed for drainage, and use a preset green measure vertical collaborative optimization algorithm to obtain the Pareto optimal solution set of the total outflow of green measures and the total cost of green measure deployment in all catchment areas of the cities to be managed for drainage, based on the deployment priority coefficient and the parameters of the green measures to be deployed, and determine the optimal green measure deployment scheme for the areas of the cities to be managed for drainage in each catchment area.

9. An electronic device comprising a central processing unit and a memory, characterized in that, The central processing unit is used to invoke and run a computer program stored in the memory to perform the steps of the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores, in the form of computer-readable instructions, a computer program implemented according to any one of claims 1 to 7, which, when invoked by a computer, executes the steps included in the corresponding method.

Images