Urban community surface ponding simulation method considering roof structure

Abstract

The invention relates to the technical field of urban hydrological simulation. The invention discloses an urban community surface ponding simulation method considering a roof structure, and the method comprises the steps: obtaining a regional multi-angle picture through an oblique photography unmanned plane, generating an oblique photography model through image processing, and reconstructing a digital surface model according to the oblique photography model; ground data are separated from the DSM through a PMF method, roof data are extracted in combination with building features, mask data are output, other ground features are removed, and building roof elevation grid data are obtained; calculating the gradient of the roof, dividing the gradient into flat and sloping roofs, building drainage structures into roof drainage models, integrating the roof drainage models with the urban community hydrological model considering the roofs, and running to realize surface ponding simulation. The method comprises the following steps: building a high-precision model through an oblique photography unmanned aerial vehicle and a GIS, separating ground data through a PMF, and accurately dividing a flat / slope roof through median filtering and an SVM; special drainage structures are arranged for the two types of roofs; and integrating the roof drainage model and a two-dimensional coupling hydrological model, and outputting an accumulated water thematic map.

Description

Technical Field

[0001] This invention relates to the field of urban hydrological simulation technology, and in particular to a method for simulating ground water accumulation in urban communities that takes into account roof structures. Background Technology

[0002] With the accelerating pace of urbanization and the continuous expansion of urban areas, vast amounts of land have been converted into artificial surfaces, significantly increasing the number of impermeable areas within cities. The most significant impact of impermeable surfaces on urban hydrology is the disruption of the original natural hydrological cycle, leading to a sharp increase in stormwater runoff and placing unprecedented pressure on urban drainage systems. Especially against the backdrop of frequent extreme weather events, the resulting urban flooding problem has become an increasingly serious environmental and social issue. Researchers are employing urban hydrological models based on urban hydrophysical processes to scientifically and effectively address urban flooding. With technological advancements, the simulation of urban hydrological processes has expanded from one-dimensional models to one- or two-dimensional coupled models. One- or two-dimensional coupled model technology can simulate urban flooding in complex urban geographical environments. However, most existing urban flooding models use data with low precision, thus failing to meet the requirements for preventing and controlling urban flooding problems.

[0003] Most existing urban flooding models indeed tend to overlook the crucial role of building roofs in the water accumulation process, especially the unique functions of different roof types. Based on structural characteristics, urban buildings can be broadly categorized into flat roofs and pitched roofs. Roof design is critical for controlling flooding, serving both as temporary rainwater storage to slow surface runoff and reducing the risk of flooding through rapid drainage. Flat roofs act as temporary rainwater storage and slow surface runoff, while pitched roofs reduce the risk of flooding through rapid drainage. Furthermore, timely drainage prevents structural damage from long-term load-bearing.

[0004] Most existing methods for simulating waterlogging have limitations in terms of data accuracy and model complexity. These methods typically rely on low-precision meteorological, topographic, and land-use data, which limits the accuracy of model predictions. Furthermore, many models neglect the impact of building roofs on waterlogging formation, especially the differences in stormwater management and runoff control between different types of roofs, such as flat roofs and pitched roofs. Such simplification prevents the models from fully reflecting reality, thus affecting the reliability and practicality of the model results. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention discloses a method for simulating urban community ground water accumulation that considers the differences in rainwater management and runoff control between different types of roofs, such as flat roofs and pitched roofs, and constructs a high-precision one- or two-dimensional coupled hydrological model of the city that takes roofs into account, thereby achieving the accuracy requirements of high-precision water accumulation simulation data.

[0006] This invention discloses a method for simulating ground water accumulation in urban communities considering roof structures, which includes the following steps:

[0007] Oblique photography drones are used to acquire multi-angle photos of a region, which are then stitched together using image processing software to generate an oblique photography model.

[0008] Reconstructing the DSM digital surface model based on oblique photogrammetry;

[0009] Ground data was separated from the DSM using the PMF progressive morphological filtering method. The remaining ground feature data after the ground data was separated from the DSM included building, tree and municipal facility data, and building data included roof data.

[0010] Roof data is extracted by combining visual and building feature features, and roof range mask data is output. Based on this mask data, other features are removed in DSM to obtain building roof elevation raster data.

[0011] Roof slope is calculated based on elevation raster data, and building roofs are classified into flat roofs and pitched roofs according to their slope.

[0012] Drainage structures are constructed based on the two types of roof characteristics, and a roof drainage model considering roof type is established.

[0013] By integrating the roof drainage model with the urban community hydrological model that takes into account the building roof, the model can be run to simulate ground water accumulation.

[0014] Furthermore, the steps for extracting roof data using visual information combined with building feature characteristics include:

[0015] Identify the dark areas in the image layers corresponding to the DSM and oblique photogrammetry model, then filter out features with regular geometric shapes, and finally verify whether the surface has regular repeating textures.

[0016] The areas that meet the three characteristics are manually selected and marked to generate initial roof range data;

[0017] Remove non-roof features from the initial data and output the final roof area mask data;

[0018] By cropping the DSM based on the mask data and preserving the roof elevation information, the building roof elevation raster data is generated.

[0019] Furthermore, the steps for classifying building roofs into flat roofs and pitched roofs based on their slope include:

[0020] Based on roof elevation raster data, the slope value is calculated pixel by pixel to generate a slope raster layer;

[0021] Median filtering is applied to the slope raster layer to eliminate noise interference from isolated pixels;

[0022] Based on the threshold of the field survey, roofs with an average slope greater than the threshold are classified as pitched roofs, and those with an average slope less than the threshold are classified as flat roofs.

[0023] Known types of roofs with uniform distribution are selected as training samples, and a classifier is trained using SVM with slope as the feature.

[0024] The trained classifier is applied to the automatic classification of the slope layer of the entire region, and the error is manually corrected by combining the oblique photogrammetry model to output the classification results.

[0025] Furthermore, the construction of a flat roof drainage structure includes:

[0026] Measure the plan dimensions of the flat roof, determine the coordinates of the geometric center, and set the drainage node at those coordinates;

[0027] Starting from the drainage node, the drainage pipes are laid along a gentle direction, and the end of the pipes is connected to the building's outdoor drainage interface.

[0028] Configure pipe parameters: Select cast iron pipe or plastic pipe, set the diameter according to the roof catchment area, set the roughness according to the characteristics of the pipe material, and set the slope according to the drainage flow direction.

[0029] Furthermore, the construction of the drainage structure for the pitched roof includes:

[0030] Extract the ridge line and eaves edge line of the pitched roof, and evenly set drainage nodes along the eaves line;

[0031] A downpipe is vertically installed at each drainage node, and the bottom of the downpipe is connected to the ground drainage network.

[0032] Configuration parameters: Select cast iron pipe or plastic pipe as the pipe material; set the downpipe diameter based on the single slope area; set the roughness based on the pipe material characteristics; and set the slope based on the eaves height.

[0033] Furthermore, the construction of the roof drainage model includes:

[0034] Based on the two types of roof drainage structure designs, draw a topology diagram of the drainage network and mark the pipe routes, nodes and downpipes;

[0035] Determine the core parameters of the pipeline network: pipe material, diameter, roughness, slope, and node width;

[0036] Based on the topology diagram and core parameters of the pipe network, the integrity of the pipe network topology is checked, and a roof drainage model is generated.

[0037] Furthermore, the determination of drainage network parameters includes:

[0038] Piping materials: Select the pipe type for flat roofs and pitched roofs from cast iron pipes and plastic pipes, based on the building's usage environment;

[0039] Diameter: The diameter of the pipe for a flat roof is determined by the area of ​​the flat roof, and the diameter of the downpipe for a pitched roof is determined by the area of ​​the single slope.

[0040] Roughness: Cast iron pipes are set with a fixed coefficient, while plastic pipes are set with a coefficient lower than that of cast iron pipes;

[0041] Slope: For flat roofs, pipes should be designed with a gentle slope; for pitched roofs, downpipes should be designed with a vertical or near-vertical slope.

[0042] Node width: Calculated using a formula based on the maximum overflow volume, overflow outlet width coefficient, and maximum design water accumulation height.

[0043] Furthermore, the steps for integrating the roof drainage model with the urban community hydrological model include:

[0044] Roof catchment areas are divided as follows: a single flat roof is an independent catchment area; flat roofs with elevation differences are divided into catchment areas according to elevation; pitched roofs are divided into single-slope catchment areas along the ridge line.

[0045] Match drainage nodes to each catchment area and record the relationships between them;

[0046] Calculate the hydrological parameters of the catchment area: impermeable surface ratio is 100%, slope is taken as average, Manning coefficient is set according to roof material, and width is calculated based on runoff coefficient and area.

[0047] The runoff data and catchment area parameters of the roof drainage model are imported into the urban hydrological model and integrated with surface runoff and soil permeability data. Pipe roughness and node type are set to complete the integration.

[0048] Furthermore, the construction of urban community hydrological models that take into account building rooftops includes:

[0049] Construct a one-dimensional model: Import data on catchment areas, land use, and drainage networks; set pipe roughness and node types; simulate water flow in the network; and calculate node overflow volume.

[0050] Constructing a two-dimensional model: Import high-precision terrain data (DEM), divide it into grid cells, assign surface roughness, slope, runoff coefficient, and infiltration rate to the grid cells, and set initial and boundary conditions;

[0051] Input the overflow volume from the one-dimensional model into the two-dimensional model, calculate the water accumulation height of the grid according to the shallow water equation, set the simulation time step and duration, and complete the model construction.

[0052] Furthermore, the simulated water accumulation output is water depth data in the form of a raster, with each raster corresponding to a depth value.

[0053] The beneficial effects of this invention are:

[0054] This invention effectively solves the problems of low data accuracy and insufficient consideration of the impact of roofs in existing urban flooding simulation models. Existing models often rely on low-precision meteorological and topographical data and frequently ignore the role of building roofs in flooding formation, resulting in poor simulation reliability. This invention, however, uses oblique photogrammetry drones to acquire multi-angle photographs of a region, generating a high-precision oblique photogrammetry model. This model is then reconstructed using GIS software, and ground data is separated using the PMF method. Roof data is accurately extracted by combining building feature characteristics. Furthermore, median filtering and an SVM classifier are used to accurately classify flat and pitched roofs, providing high-resolution and high-accuracy basic data support for subsequent flooding simulations and improving the accuracy of the simulation data.

[0055] This invention designs customized drainage schemes for different roof structures, overcoming the shortcomings of existing models that simplify the handling of roof drainage differences. For flat roofs with a gentle slope and slow water flow, this invention places drainage nodes at the geometric center, with pipes laid out from these nodes. For pitched roofs with a steep slope and rapid water flow, drainage nodes and downpipes are installed along the eaves edge. Parameters such as pipe material, diameter, and roughness are specified for both types of roofs. This differentiated design accurately reflects the actual characteristics of flat roofs—water storage and slowing runoff—and pitched roofs—rapid drainage, avoiding distortion in roof drainage simulation that could affect overall water accumulation prediction results and making the drainage process simulation more realistic.

[0056] This invention enhances the practicality and decision-making value of flood simulation by integrating a roof drainage model with a coupled one- and two-dimensional urban community hydrological model that considers building roofs. The one-dimensional model simulates the dynamics of pipe network water flow and calculates the overflow volume at nodes, while the two-dimensional model, based on a high-precision DEM grid, calculates the water level height by combining surface attribute parameters. The combination of the two models comprehensively reflects the process of pipe network drainage and surface water diffusion. The thematic maps output by the simulation, including water depth distribution, duration, and impact range, can directly provide accurate data for urban flood control and drainage planning, drainage system optimization, etc., helping to scientifically assess flood risk and improve the city's ability to cope with rainstorm flooding. Attached Figure Description

[0057] Figure 1 This is a design drawing of the drainage network structure in a method for simulating ground water accumulation in an urban community that considers the roof structure, as described in this application.

[0058] Figure 2 This is a roof drainage model diagram in an urban community ground water accumulation simulation method considering roof structure, as described in the embodiments of this application.

[0059] Figure 3 This is a flowchart illustrating a method for simulating ground water accumulation in urban communities that considers roof structure, categorized into flat roofs and pitched roofs according to roof type, as described in this application.

[0060] Figure 4 In the embodiments of this application, a method for simulating ground water accumulation in urban communities that considers roof structure is used to construct drainage structures for flat roofs and pitched roofs based on the characteristics of flat roofs and pitched roofs, thereby establishing a roof drainage model process that considers roof type.

[0061] Figure 5 In one embodiment of this application, a method for simulating ground water accumulation in urban communities that considers roof structure is implemented by integrating a roof drainage model with a hydrological model of urban communities that takes roof structure into account. The process of simulating ground water accumulation in urban communities is shown in the figure.

[0062] Figure 6 This is a flowchart illustrating a method for simulating ground water accumulation in urban communities that considers roof structures, as described in this application. Detailed Implementation

[0063] To enable those skilled in the art to better understand the present invention, the technical solutions in the specific embodiments of the present invention will be clearly and completely described below.

[0064] 1. The first objective of this invention is to extract building roofs from oblique photogrammetry model data and digital surface data, and to classify them into flat roofs and pitched roofs according to roof type.

[0065] The solution to achieve the first objective of this invention includes the following steps:

[0066] 1) Detailed 3D Surface Data Acquisition: The oblique photogrammetry data used in this invention is obtained by taking photos of the area from different angles using an oblique photogrammetry drone (a drone equipped with multiple cameras at different angles). The oblique photogrammetry drone must be equipped with no fewer than four lenses, and the camera resolution must be no less than 20 million pixels. The flight altitude is set to 100 to 150 meters according to the modeling accuracy requirements. Then, oblique photogrammetry image processing software is used to stitch together the multi-angle photos of the covered area to generate an oblique photogrammetry model. A major feature of this model is that it can obtain three-dimensional and detailed 3D surface data of the area. The digital surface data is obtained by reconstructing the digital surface model (which records the elevation information of the surface and features, including all natural terrain features and man-made structures) using GIS (Geographic Information System) software through the oblique photogrammetry model. This model contains the elevation information of the ground and features on the ground.

[0067] 2) Separating Ground Data: First, it's necessary to distinguish ground data from non-building data. This method uses the Progressive Morphological Filter (PMF) method to extract ground data. The PMF method iterates through the digital surface model by adjusting the window, distinguishing between ground and non-ground data based on a set threshold. Typically, the window size is 5×5 (pixels), and the threshold is 10 meters. The extracted ground data is then examined, and GIS software is used to separate the ground data from the digital surface model.

[0068] The PMF (Progressive Morphological Filtering) parameters need to be dynamically adjusted according to scene characteristics. For flat communities with dense buildings and few trees, a 5×5 pixel window size and a 10-meter height threshold are used. For communities with significant terrain undulations (slope > 15°), the window size is adjusted to 7×7 pixels to cover a larger terrain unit, and the height threshold is reduced to 8 meters to reduce the interference of terrain undulations on ground determination. For densely wooded areas (tree coverage > 30%), some tree point clouds are first removed through preprocessing using vegetation indices (such as NDVI), and then a 6×6 pixel window and a 9-meter height threshold are used for filtering. After filtering, random sampling is required for inspection. Five sampling points are randomly selected within a certain range to ensure that the ground data extraction accuracy is not less than 95%. If the standard is not met, the parameters are readjusted and the filtering process is repeated.

[0069] 3) Roof Data Extraction: After separating the digital surface model, the remaining data mainly consists of ground features (buildings, trees, municipal facilities). Therefore, a method combining visual features with ground feature characteristics (color, shape, texture) is used to extract these features in GIS software. The main ground feature characteristics of buildings are as follows: most buildings are dark in color; most shapes are regular geometric shapes; and textures are regular with repeating patterns. The manually extracted roof data is output as roof extent mask data in the GIS software. Then, using the roof extent mask data, other ground features are removed from the digital surface model to obtain high-precision building roof elevation raster data.

[0070] 4) Calculating Roof Slope: This invention uses high-precision building roof elevation raster data as input to calculate the roof slope in GIS software. The "Slope" tool in the software is used to calculate the slope at each point. The slope calculation method is based on the elevation difference between the center pixel and its eight surrounding pixels, as shown in the following formula:

[0071]

[0072] In the formula, Z i，j x represents the elevation value (m) of the cell in the i-th row and j-th column; c dx represents the pixel size (m); s represents the slope; dx represents the rate of elevation change along the x-axis in the elevation raster data; dy represents the rate of elevation change along the y-axis in the elevation raster data.

[0073] The formula group is used to iterate through the building roof elevation raster data cells, and the calculated slope raster layer is saved.

[0074] 5) Roof Type Classification: This invention first reduces the noise impact of other interferences on the slope layer through median filtering. Based on the field survey results, roofs with an average slope greater than 5° are defined as pitched roofs, and those with an average slope less than 5° are defined as flat roofs. In snowy northern regions, this threshold is adjusted to 7°, while in rainy southern regions, it remains unchanged at 5°. During implementation, at least thirty roof slope survey samples from the field are collected within the region. The threshold is then calibrated using the sample data to ensure classification accuracy. At least 12 known roof types are selected as training samples. These samples need to be evenly distributed within the region to ensure the model's generalization ability. Simultaneously, the number of samples for flat and pitched roofs should be relatively balanced to avoid class imbalance. Using slope as the classification feature, an SVM (Support Vector Machine) classifier is selected for training. The training samples and their corresponding roof types are input, and after parameter optimization, the entire region is classified to complete the roof type classification. Based on the oblique photogrammetry model, the classification results are further corrected in GIS software by manual inspection and editing tools according to the actual regional images. The classification results are then converted into vector files, and the building roof data of flat roofs and pitched roofs are extracted according to the range of the vector files.

[0075] When the roof type is a special roof, the drainage structure needs to be designed specifically according to the structural characteristics. For folded plate roofs, the roof is divided into sections (each section length ≤ 15 meters), and each section is considered an independent slope. The drainage method is determined according to the slope (refer to the classification standard for flat / sloping roofs). When the slope is < 5°, a drainage node is set at the geometric center of each section and pipes are laid along the gentle direction. When the slope is > 5°, a drainage node is set along the eaves of each section and downpipes are laid vertically. For dome roofs, the lowest point of the dome is the core drainage node. 3-4 radial drainage channels (10cm wide, 5cm deep) are set along the surface of the dome. Downpipes are connected to the ends of the channels. The diameter of the downpipes is determined according to the projected area of ​​the dome (refer to the standard for the single slope area of ​​a sloping roof). For curved roofs, drainage nodes are set at points where the difference in elevation between the contour lines is > 0.5 meters. Drainage pipes are laid along the tangent direction of the contour lines. The slope of the pipes is set at 1.2 times the slope of the contour lines to enhance drainage power. Special roof drainage structures need to be verified through hydraulic calculations (ensuring a pipe flow velocity of 1.0-2.0 m / s). If the calculations fail, the node positions or pipe parameters should be adjusted.

[0076] 2. The second objective of this invention is to construct drainage structures for flat roofs and pitched roofs based on their characteristics, thereby establishing a roof drainage model that considers roof type. The specific steps are as follows:

[0077] 1) Drainage Pipeline Structure Design: For flat roofs, due to their gentle slope and slow water flow, water typically converges under gravity. This invention places the drainage node at the geometric center of the flat roof, using this node as the starting point for the roof drainage pipes. For pitched roofs, with their steeper slope and faster water flow, unlike flat roofs, pitched roofs usually have downpipes at the edges. Water flows into these downpipes under gravity and then into the main drainage system. Therefore, the drainage nodes and pipes for pitched roofs are located at the roof edges. See the detailed drainage pipeline structure design below. Figure 1 .

[0078] 2) Determining Drainage Network Parameters: The parameters of drainage nodes and pipes are set as needed. The main parameters to be set are as follows: pipe material, pipe diameter, pipe roughness, pipe slope, and node width. Drainage network parameters for flat roofs and pitched roofs are set separately according to the roof type. Specific pipe parameters are shown in Table 1.

[0079] Table 1. Specific Pipeline Parameter Settings

[0080]

[0081] Table 1

[0082] The diameter of the drainage pipes must be matched according to the standard quantitative specifications based on the roof's catchment area. Flat roof catchment area <150m² 2 Use 75mm diameter pipes, 150-300m 2 Use a 100mm diameter pipe for lengths >300m 2 Two pipes (each 100mm in diameter) are installed in parallel; the single-slope area of ​​the pitched roof is less than 100m². 2 Use a 100mm diameter downpipe, 100-250m long. 2 Use a 150mm diameter downpipe for lengths >250m 2 An additional 150mm diameter downpipe will be installed every 20 meters along the eaves line. The catchment area calculation must include the roof projection area and the roof extension, and the calculation result should be rounded up to the nearest integer to ensure drainage capacity redundancy.

[0083] In the case of heavy rainfall during a typhoon, the diameter of the drainage pipe is increased by 20% based on the conventional calculation, and the overflow outlet width coefficient N is taken as 0.9.

[0084] Node width is an important parameter used when calculating the node overflow width. The formula for calculating the node overflow width is as follows:

[0085] ;

[0086] In the formula, b is the width of the node overflow outlet (m); Q qThe maximum overflow flow rate is (L / s); N is the overflow outlet width coefficient; g is the gravitational acceleration, taken as 9.81 m / s²; h max Maximum design water accumulation height on the roof (m); h b The distance (m) is from the bottom of the node to the roof.

[0087] 3) Roof Drainage Model Construction: Based on the confirmed drainage network structure design and parameter determination results, the designed drainage network structure and determined parameter values ​​are input using GIS software to check the topological relationships and construct the roof drainage model. This step ensures the accuracy and rationality of all input data, thereby obtaining reliable simulation results.

[0088] 3. The third objective of this invention is to integrate the roof drainage model with the urban community hydrological model that takes into account building roofs, thereby simulating surface water accumulation in urban communities. The specific steps are as follows:

[0089] 1) Roof Drainage Model Integration: Due to the relatively small slope of a single flat roof, the entire flat roof is considered a single catchment area. If a single flat roof has elevation differences or multiple drainage points, the roof is divided into smaller areas based on elevation and drainage path, with each area corresponding to a main drainage point. For pitched roofs, the catchment area is divided along the ridge line. Using the ridge line as the boundary, each section forms an independent catchment area, defined by the ridge line and the roof edge. After completing the catchment area division, GIS software is needed to calculate the parameters related to the catchment area and assign them to it. The impermeable surface ratio is 100% (infiltration rate 0), the slope is taken as the average slope within the catchment area, and the Manning roughness coefficient is taken as 0.012 (the Manning coefficients of common roof materials such as asphalt and ceramic tiles are close, and an empirical value can be used uniformly). The catchment area width is calculated using the following formula:

[0090] ;

[0091] Where W is the width of the catchment area; K is the runoff coefficient; and S is the area of ​​the catchment area (m²). 2 ).

[0092] Finally, in the GIS software, based on the geographical location of the drainage nodes matching the geographical extent of the catchment area, one drainage node needs to be matched and marked for each catchment area. This completes the labeling work for the catchment area division. The rooftop drainage nodes are then connected to the nodes in the urban drainage network system to ensure that rooftop drainage data can accurately flow into the urban drainage hydrological model. Then, the runoff data generated by the rooftop drainage model is used as input and combined with other input data (surface runoff, soil permeability) from the urban community hydrological model. By setting appropriate model parameters (pipe roughness coefficient, node type), the rooftop drainage model integration is completed. See the specific rooftop drainage model for details. Figure 2 .

[0093] Surface runoff data were obtained through a combination of field observations and remote sensing inversion. Field observations were conducted on 3-5 representative plots within the community, such as asphalt pavements, grasslands, and bare land, where runoff monitoring instruments were deployed. Runoff coefficients were obtained by continuously monitoring three rainstorms. Soil permeability data were obtained through field sampling using a dual-ring infiltration meter, with a sampling density of 1000 m³ / m². 2 A single sampling point with a sampling depth of 0-50 cm was used to measure the saturated hydraulic conductivity. Then, soil permeability raster data with a resolution of 2 meters was generated by Kriging interpolation.

[0094] 2) Construction of urban community hydrological models considering building rooftops: Input the labeled catchment area and land use data into a professional hydrological model, and set the various parameters required by the model, such as the roughness coefficient of pipes. The software establishes a one-dimensional urban community hydrological model, specifying the node types. This one-dimensional model primarily simulates the water flow dynamics of urban drainage systems, and its core equations are the Manning equation and the Saint-Venant equation, as follows:

[0095] ;

[0096] ;

[0097] In the formula, v is the water flow velocity (m / s); n is the roughness coefficient; R is the hydraulic radius (m), which is equal to the cross-sectional area A divided by the wetted perimeter P; S is the slope; Q is the pipe flow rate (m³ / s); t is the simulation time (s); x is the length along the pipe direction (m); A is the cross-sectional area (m²); g is the acceleration due to gravity, which is taken as 9.81 m / s² here; z is the ground elevation (m); S f This refers to the friction slope.

[0098] To obtain v and Q using the above formulas, the overflow water volume Q at each node needs to be calculated. o The calculation is performed using the following formula:

[0099] ;

[0100] In the formula, Q o Q represents the overflow volume of the node (m3); i Let Q represent the inflow volume (m³) at each node. Roof drainage is included in the node inflow volume Q of the one-dimensional model. i .

[0101] When constructing a two-dimensional urban community hydrological model, a high-precision digital elevation model (DEM) and land use data are first used to divide the area into regular or irregular grid cells using GIS software. The physical properties of each grid cell, such as surface roughness and slope, are defined in the software, and initial and boundary conditions are set. Based on the land cover type and soil properties, parameters such as runoff coefficient and infiltration rate are set for each grid cell. After completing the above settings in the GIS software, the grid cells and the output of the one-dimensional urban hydrological model are input into the two-dimensional urban community hydrological model. The physical properties of each grid cell, surface roughness and slope, need to be defined in the model. In addition, the runoff system and infiltration rate of each grid cell also need to be set according to the land cover type and soil properties. The core equation of the two-dimensional model is the shallow water equation, which describes the horizontal flow of water, as follows:

[0102] ;

[0103] ;

[0104] ;

[0105] In the formula, h is the surface water depth (m); t is the simulation time (s); x and y are the distances along the pipe and perpendicular to the pipe on the horizontal plane (m); u and v are the water flow velocities in the x and y directions (m / s); and g is the acceleration due to gravity, which is taken as 9.81 m / s² here. 2 z represents ground elevation (m); S fx S fy The friction slopes are in the x and y directions; A cell The area of ​​the grid cell is (m2).

[0106] The node overflow water volume data Q output by the one-dimensional model o As input to the two-dimensional model, the water accumulation height h of the grid cells is calculated. By setting appropriate simulation time steps Δt and simulation duration t, the urban community hydrological model considering building rooftops is completed.

[0107] 3) Simulation of ground water accumulation in urban communities: By integrating the above-mentioned roof drainage model and constructing an urban community hydrological model that takes into account building roofs, the urban community hydrological model can be run to simulate ground water accumulation in urban communities. The simulation results are a raster layer based on water depth.

[0108] In this invention, building roofs are extracted using oblique photogrammetry model data and digital surface data, and classified into flat roofs and pitched roofs according to roof type. (Flowchart shown below) Figure 3As shown, based on the characteristics of flat roofs and pitched roofs, drainage structures for flat roofs and pitched roofs are constructed respectively, thus establishing a roof drainage model considering roof type. The process is as follows: Figure 4 As shown, the roof drainage model is integrated with a roof-based urban community hydrological model to realize the simulation process of surface water accumulation in urban communities. Figure 5 The overall flowchart of this invention is as follows: Figure 6 As shown,

[0109] Methods for extracting and classifying building rooftops

[0110] The extraction and classification of building roofs is a fundamental method for water accumulation simulation. Building roofs are extracted using oblique photogrammetry model data and digital surface data, and classified into flat roofs and pitched roofs according to their type.

[0111] Drainage structure design based on roof type

[0112] For flat roofs, due to their gentle slope, water flow converges slowly, typically relying on gravity. This invention places the drainage node at the geometric center of the flat roof, using this node as the starting point for the roof drainage pipes. For pitched roofs, with their steeper slope, water flow converges quickly. Unlike flat roofs, pitched roofs usually have downpipes at the edges. Water flows into these downpipes by gravity and then into the main drainage system. Therefore, the drainage nodes and pipes for pitched roofs are located at the roof edges.

[0113] Construction of roof drainage model

[0114] Based on the characteristics of flat roofs and pitched roofs, drainage structures for flat roofs and pitched roofs are constructed respectively, thereby establishing a roof drainage model that takes into account the roof type.

[0115] Integration of roof drainage models with urban community hydrological models that take into account building rooftops

[0116] By integrating roof drainage models with urban community hydrological models that take into account building roofs, surface water accumulation simulation in urban communities can be achieved.

[0117] This invention, which considers building rooftops, constructs a hydrological model for urban communities to more accurately assess the risk of waterlogging and drainage capacity. The model comprehensively considers the drainage characteristics of different roof types, including the water storage and drainage capacity of flat and pitched roofs, and integrates multiple factors such as meteorology, topography, and land use to establish a multi-factor coupled hydrological model. By utilizing high-resolution remote sensing and topographic data, the spatial resolution and accuracy of the model are improved. Furthermore, through real-time monitoring and dynamic updates, the model enables real-time simulation and prediction of urban hydrological processes.

[0118] It should be understood that those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A method for simulating ground water accumulation in urban communities considering roof structures, characterized in that, Includes the following steps: Oblique photography drones are used to acquire multi-angle photos of a region, which are then stitched together using image processing software to generate an oblique photography model. Reconstructing the DSM digital surface model based on oblique photogrammetry; Ground data was separated from the DSM using the PMF progressive morphological filtering method. The remaining ground feature data after the ground data was separated from the DSM included building, tree and municipal facility data, and building data included roof data. Roof data is extracted by combining visual and building feature features, and roof range mask data is output. Based on this mask data, other features are removed in DSM to obtain building roof elevation raster data. Roof slope is calculated based on elevation raster data, and building roofs are classified into flat roofs and pitched roofs according to their slope. Drainage structures are constructed based on the two types of roof characteristics, and a roof drainage model considering roof type is established. By integrating the roof drainage model with the urban community hydrological model that takes into account the building roof, the model can be run to simulate ground water accumulation.

2. The method for simulating ground water accumulation in urban communities considering roof structures according to claim 1, characterized in that, The steps for extracting roof data using visual and architectural features include: Identify the dark areas in the image layers corresponding to the DSM and oblique photogrammetry model, then filter out features with regular geometric shapes, and finally verify whether the surface has regular repeating textures. The areas that meet the three characteristics are manually selected and marked to generate initial roof range data; Remove non-roof features from the initial data and output the final roof area mask data; By cropping the DSM based on the mask data and preserving the roof elevation information, the building roof elevation raster data is generated.

3. The method for simulating ground water accumulation in urban communities considering roof structures according to claim 1, characterized in that, The steps for classifying building roofs into flat roofs and pitched roofs based on their slope include: Based on roof elevation raster data, the slope value is calculated pixel by pixel to generate a slope raster layer; Median filtering is applied to the slope raster layer to eliminate noise interference from isolated pixels; Based on the threshold of the field survey, roofs with an average slope greater than the threshold are classified as pitched roofs, and those with an average slope less than the threshold are classified as flat roofs. Known types of roofs with uniform distribution are selected as training samples, and a classifier is trained using SVM with slope as the feature. The trained classifier is applied to the automatic classification of the slope layer of the entire region, and the error is manually corrected by combining the oblique photogrammetry model to output the classification results.

4. The method for simulating ground water accumulation in urban communities considering roof structures according to claim 1, characterized in that, The construction of a flat roof drainage structure includes: Measure the plan dimensions of the flat roof, determine the coordinates of the geometric center, and set the drainage node at those coordinates; Starting from the drainage node, the drainage pipes are laid along a gentle direction, and the end of the pipes is connected to the building's outdoor drainage interface. Configure pipe parameters: Select cast iron pipe or plastic pipe, set the diameter according to the roof catchment area, set the roughness according to the characteristics of the pipe material, and set the slope according to the drainage flow direction.

5. The method for simulating ground water accumulation in urban communities considering roof structures according to claim 1, characterized in that, The construction of a pitched roof drainage structure includes: Extract the ridge line and eaves edge line of the pitched roof, and evenly set drainage nodes along the eaves line; A downpipe is vertically installed at each drainage node, and the bottom of the downpipe is connected to the ground drainage network. Configuration parameters: Select cast iron pipe or plastic pipe as the pipe material; set the downpipe diameter based on the single slope area; set the roughness based on the pipe material characteristics; and set the slope based on the eaves height.

6. The method for simulating ground water accumulation in urban communities considering roof structures according to claim 1, characterized in that, The construction of the roof drainage model includes: Based on the two types of roof drainage structure designs, draw a topology diagram of the drainage network and mark the pipe routes, nodes and downpipes; Determine the core parameters of the pipeline network: pipe material, diameter, roughness, slope, and node width; Based on the topology diagram and core parameters of the pipe network, the integrity of the pipe network topology is checked, and a roof drainage model is generated.

7. The method for simulating ground water accumulation in urban communities considering roof structures according to claim 1, characterized in that, The determination of drainage network parameters includes: Piping materials: Select the pipe type for flat roofs and pitched roofs from cast iron pipes and plastic pipes, based on the building's usage environment; Diameter: The diameter of the pipe for a flat roof is determined by the area of ​​the flat roof, and the diameter of the downpipe for a pitched roof is determined by the area of ​​the single slope. Roughness: Cast iron pipes are set with a fixed coefficient, while plastic pipes are set with a coefficient lower than that of cast iron pipes; Slope: For flat roofs, pipes should be designed with a gentle slope; for pitched roofs, downpipes should be designed with a vertical or near-vertical slope. Node width: Calculated using a formula based on the maximum overflow volume, overflow outlet width coefficient, and maximum design water accumulation height.

8. The method for simulating ground water accumulation in urban communities considering roof structures according to claim 1, characterized in that, The steps involved in integrating roof drainage models with urban community hydrological models include: Roof catchment areas are divided as follows: a single flat roof is an independent catchment area; flat roofs with elevation differences are divided into catchment areas according to elevation; pitched roofs are divided into single-slope catchment areas along the ridge line. Match drainage nodes to each catchment area and record the relationships between them; Calculate the hydrological parameters of the catchment area: impermeable surface ratio is 100%, slope is taken as average, Manning coefficient is set according to roof material, and width is calculated based on runoff coefficient and area. The runoff data and catchment area parameters of the roof drainage model are imported into the urban hydrological model and integrated with surface runoff and soil permeability data. Pipe roughness and node type are set to complete the integration.

9. A method for simulating ground water accumulation in urban communities considering roof structures, as described in claim 1, is characterized in that... The construction of urban community hydrological models that take into account building rooftops includes: Construct a one-dimensional model: Import data on catchment areas, land use, and drainage networks; set pipe roughness and node types; simulate water flow in the network; and calculate node overflow volume. Constructing a two-dimensional model: Import high-precision terrain data (DEM), divide it into grid cells, assign surface roughness, slope, runoff coefficient, and infiltration rate to the grid cells, and set initial and boundary conditions; Input the overflow volume from the one-dimensional model into the two-dimensional model, calculate the water accumulation height of the grid according to the shallow water equation, set the simulation time step and duration, and complete the model construction.

10. A method for simulating ground water accumulation in urban communities considering roof structures according to claim 1, characterized in that: The output of the surface water accumulation simulation is water depth data in the form of a raster, with each raster corresponding to a depth value.

Images