High-performance multi-scale urban ventilation corridor construction-recognition-extraction method

Abstract

The invention relates to the field of urban planning and design, in particular to a high-performance multi-scale urban ventilation corridor construction-recognition-extraction method. Comprising the following steps: dividing building data through grids, sampling and constructing an urban average height grid model, and superposing a digital elevation model (DEM) and an azimuth gradient uplift model as a gallery foundation to obtain an azimuth uplift building (DSM); performing depression filling processing on the data to eliminate depression points; a D8 algorithm and a double-probabilistic roulette algorithm are used for distributing the flow direction according to the wind flow characteristics, the probabilistic roulette algorithm focuses on the climbing characteristics of the wind, the elevation absolute difference of the building is subjected to reverse weighting to calculate the probability, and the dominant direction of the wind influences the dominant weighting to calculate the probability; and the total number of each flow direction grid is counted and accumulated for multiple times in the whole grid, and an air volume map is obtained. And setting a threshold value to extract the minimum surface ventilation quantity, calculating an area of a grid value in a threshold value range, carrying out binaryzation, and extracting an air duct. And converting the grid air duct data into vector data for further analysis and display.

Description

Technical Field

[0001] The present invention relates to multiple fields such as urban planning and design, urban climatology, architecture, environmental science and ecology, and particularly relates to a method for constructing-identifying-extracting urban ventilation corridors with high performance and high precision for large ranges, multiple scales, and multiple directions. Background Art

[0003] Traditional research methods for urban ventilation corridors include wind tunnel experiment method, computer numerical simulation method, circuit data simulation, etc. Wind tunnel experiments are suitable for simulating wind environments of various types and scales. Although such simulations can objectively and realistically observe the detailed changes when air flows over the observed object, the use of this method is restricted by high experimental costs, limited simulation ranges, and strict requirements for the accuracy of physical models. CFD simulation has more advantages in high-resolution small-scale fluid simulations, and the simulation is very reliable. It is usually used for environmental prediction in urban planning and the diffusion of pollutants, but this method has a large amount of calculation and only adopts a single building scale or small community scale when simulating the urban wind environment; it is not suitable for simulating large-scale wind environments at the urban scale. Compared with CFD simulation, the circuit model has a faster calculation speed, is suitable for preliminary evaluation and scheme comparison, supports rapid modeling and optimization of various ventilation layouts, but lacks complex airflow interactions between buildings and it is difficult to express local effects.

[0004] In summary, due to the large amount of computing requirements and small spatial scales, high-precision numerical simulations are difficult to evaluate the ventilation conditions at the urban scale; the circuit models applicable to large ranges have certain limitations in the analysis of ventilation environments at the building scale due to the lack of strict aerodynamic analysis. Summary of the Invention

[0005] In order to solve the problems of large computing resources in the construction of urban ventilation corridors at the micro scale, inability to perform large-range and multi-scale calculations, and the lack of complex airflow interactions between buildings in the construction of urban ventilation corridors at the macro scale, while taking into account anisotropy, considering multi-directional duct heterogeneity, and performing identification and extraction. The present invention provides a method for constructing-identifying-extracting urban ventilation corridors with high performance and multiple scales, mainly including:

[0006] S1: Divide the building data into grids, extract the average height of buildings within the grids, superimpose the grid model of DEM and azimuth elevation, and perform noise processing on the grid model;

[0007] S2: Perform flow direction analysis on the grid model after filling depressions, use the D8 algorithm of wind flow characteristics and the double-probabilistic roulette algorithm to assign flow directions to obtain the flow direction results;

[0008] S3: Use full-grid multiple flow statistics on the flow direction results, cumulatively calculate the total number of each flow direction grid, and obtain the air volume grid;

[0009] S4: Extract the range of air volume grids with air volume greater than the set threshold according to the air volume grids to obtain a high air volume area, and the high air volume area is the ventilation corridor;

[0010] S5: Perform binarization processing on the high air volume area and the non-high air volume area, extract the grid air duct, and convert the grid data into vector data to complete the extraction of the air duct vector.

[0011] A computer device includes a memory, a processor, and a computer program stored on the memory. The processor executes the computer program to implement the steps of the above method.

[0012] A computer-readable storage medium stores a computer program. When the program is executed by a processor, the steps of the above method are implemented.

[0013] A computer program product includes a computer program or instruction. When the program or instruction is executed by a processor, the steps of the above method are implemented.

[0014] The beneficial effects brought by the technical solution provided by the present invention are:

[0015] The present invention improves the method of hydrological analysis in traditional GIS to study air ducts. Considering that the basic similarity between air flow and water flow lies in the continuity of fluid properties, which satisfies the continuity equation of fluid mechanics, and their movement laws both comply with the same fluid mechanics control equations and can form streamlines to reflect the movement trajectories of fluids. Their velocity fields, pressure fields, and temperature fields can be described by the same mathematical methods. Importantly, both air flow and water flow are affected by boundary conditions, and the solid shape will significantly affect their flow characteristics, such as flow separation around an object, pressure distribution, and resistance. In many cases, water flow and air flow can be simplified as incompressible fluids for research. By constructing a multi-scale fishing net to sample building data within the urban scale, overlaying the digital elevation model (DEM), then constructing a grid model with azimuthal elevation as the slope, and continuing to overlay to obtain an azimuthally elevated building DSM. After filling depressions, the elevated grid model ensures that the directionality of the flow conforms to the input direction. On this basis, the flow direction is calculated using a double-probabilistic roulette algorithm, with the probability calculated by reverse weighting of the slope difference and the probability calculated by large weighting of the dominant direction. These two probability parameters ensure the jump and streamline characteristics of the wind, and the azimuthally elevated grid model ensures the continuity of the fluid. The main flow channels are extracted by setting thresholds through multiple calculations. In this way, high-performance multi-scale urban ventilation corridor construction-identification-extraction is carried out. On the one hand, it solves the problems of a large amount of computing resources and small spatial scale in microscopic scale methods such as CFD simulation, and greatly improves the computing speed. On the other hand, it solves the problems of lack of complex air flow interaction between buildings and aerodynamic analysis in macroscopic scale methods such as circuit simulation. Therefore, the present invention is a high-performance, high-precision, multi-scale, multi-directional, and large-scale air duct identification method. Description of the Drawings

[0016] The present invention will be further described below in conjunction with the drawings and embodiments. In the drawings:

[0017] Figure 1 is a flowchart of a high-performance multi-scale urban ventilation corridor construction-identification-extraction method in an embodiment of the present invention;

[0018] Figure 2 is a display diagram of the DSM with northern azimuthal elevation in an embodiment of the present invention;

[0019] Figure 3 is a flow direction coding diagram of the high-performance multi-scale urban ventilation corridor construction-identification-extraction method in an embodiment of the present invention;

[0020] Figure 4 is a multi-scale and multi-directional extraction result diagram of air duct calculation in the scope of Wuhan City in an embodiment of the present invention;

[0021] Figure 5 is a result diagram of the comparative CFD simulation calculation in an embodiment of the present invention. Detailed Implementation Modes

[0022] For a clearer understanding of the technical features, objectives, and effects of the present invention, the detailed implementation modes of the present invention will now be described in detail with reference to the accompanying drawings.

[0023] Embodiment 1

[0024] Please refer to Figure 1 , Figure 1 which is a flowchart of a method for constructing - identifying - extracting a high - performance multi - scale urban ventilation corridor in an embodiment of the present invention, specifically including:

[0025] S1: Grid - divide the building data, extract the average height of buildings within the grid, superimpose the DEM and the grid model with azimuthal uplift, and perform noise - point processing on the grid model. The noise - point processing method is that if a pixel is lower than eight adjacent pixels, the lowest value of its adjacent pixels is assigned to this pixel, and the flow direction is defined as towards this pixel. If multiple adjacent pixels all have the lowest value, this value is still assigned to this pixel. However, the flow direction is defined as the pixel closest to the dominant wind direction towards it, for filtering out single - pixel sinks regarded as noise points.

[0026] In step S1, the building data is the buildings within the construction area, the fishing net is a grid with a custom resolution, and the data attributes include the outlines and heights of buildings. The DEM data is the terrain data within the construction area, and the terrain data covers a buffer zone of a certain range outside the circumscribed boundary of the buildings. The grid model with azimuthal uplift is an artificially created elevation model with slopes. Step S1 includes:

[0027] S11: Fishing - net sampling: Generate grids (e.g., 10m * 10m) at a certain resolution size to divide the study area into grids, extract the average height of buildings within the grids as the values of the grids, and obtain an urban building grid model that only includes buildings. The smaller the resolution, the finer the extracted air ducts;

[0028] S12: Superimpose the regional DEM: Superimpose the urban building grid model and the DEM to obtain a terrain grid model;

[0029] S13: Take the entire study area as the object, lift the height of the wind - source area (e.g., lift the northern area for the north wind), that is, artificially construct a slope;

[0030] S14: Construct the DSM of buildings with azimuthal uplift: Continue to superimpose the slope constructed in S13 and the terrain grid model obtained in S12 to obtain the DSM of buildings with azimuthal uplift. This DSM is a grid model containing building, terrain, and slope (wind - direction control) information.

[0031] S15: Depression filling: The purpose is to remove the depression points in the DSM to avoid errors in subsequent steps.

[0032] S2: Perform flow direction analysis on the grid model after filling depressions, and use the D8 algorithm of wind flow characteristics and the roulette algorithm of double probability to allocate the flow direction. The steps for calculating the flow direction of the building DSM with azimuth elevation are as follows:

[0033] S21: D8 algorithm based on wind flow characteristics: Traverse all raster values except the boundary rows. Take the current pixel as the central pixel and its elevation value. First, calculate the absolute height difference between the central pixel and each neighboring pixel. Then, calculate the horizontal distance between the central pixel and each neighboring pixel. Finally, divide the absolute height difference between the eight neighboring pixels around the central pixel by the relative distance between the pixels to obtain the slope, and the slope represents the blocking effect on the wind. The horizontal and numerical intervals use the size of the raster resolution. In this embodiment, the distances in the up, down, left, and right directions are set to 1, and the diagonal distance is √2. For the convenience of calculation, 1.4 is uniformly used. The calculation formula is as follows:

[0034]

[0035] Among them, neighbor_value is the elevation value of the 8 neighboring pixels around the current pixel, value is the elevation value of the current pixel, abs() means absolute value calculation, and distance is the horizontal distance between the current pixel and the neighboring pixel. Calculate the slopes of the 8 neighbors in this way.

[0036] S22: Roulette algorithm of double probability: After obtaining the calculation of the 8 neighboring slope changes, instead of taking the direction of the smallest slope as the flow direction, calculate the probability by reverse weighting of the slope difference. The smallest slope occupies the largest probability, and the largest slope occupies the smallest probability; calculate the probability by large weighting of the dominant direction, and the probability of the slope position consistent with the current wind direction is the largest. Integrate the reverse slope weight and the direction weight into the roulette algorithm to obtain the flow direction, and use the calculation result of the probability event as the flow direction raster of the current grid, mark the flow direction, and the output of the flow direction is an integer raster with a value range between 1 and 255.

[0037] S3: Use full raster multiple flow statistics on the flow direction results, accumulate and calculate the total number of each flow direction raster to obtain the air volume raster. Specifically:

[0038] S31: Flow statistics: Calculate the total number of each flow direction raster accumulated in the full raster area. Example: For a 100*100 grid, each row is used as a wind source to count its path, and each time the newly generated flow direction raster is used as the calculation basis to count all possible wind circulation paths, and the result is output as the air volume raster.

[0039] S32: Statistically calculate multiple times repeatedly (for example: 1000 times).

[0040] S4: For the air volume grid, extract the range where the air volume is greater than the specified value. Set a threshold to extract the air volume grid range. For the area with ventilation within a certain range (ventilation volume greater than the set threshold), obtain the high air volume area, which is the ventilation corridor. The high air volume area is the ventilation corridor under the grid resolution and in the azimuth elevation direction. The ventilation volume greater than the specified range, that is, greater than the set threshold, indicates that the air flow in this area can pass through and the air duct exists.

[0041] S5: Perform binary processing on the high air volume area and non-high air volume area in the air volume map, extract the grid air duct, convert the grid data into vector data, and complete the extraction of the air duct vector. The non-high air volume area means that most air flows cannot pass through and cannot be regarded as an air duct. The air volume map is obtained by calculating the cumulative value under multiple probabilities.

[0042] Specifically, construct a multi-scale fishing net for sampling, and the multi-resolution segmentation of the urban area can be completed. In this way, multi-scale urban air ducts can be extracted. This method is simpler and faster in calculation compared with CFD simulation calculation, saving a large amount of computing resources and is an efficient air duct extraction method. In addition, in order to evaluate the accuracy of the extracted air duct, perform a similarity analysis on the extracted air duct and the ventilation situation accurately simulated by CFD, compare whether the high wind speed area simulated by CFD coincides with the high air volume area extracted by this method, and use the similarity index for evaluation.

[0043] The present invention cuts the buildings in the grid area by constructing a fishing net of a certain scale, uses the average height of the buildings as the grid height, samples the building data within the urban scale range, superimposes the grid on the digital elevation model DEM, then constructs a raster model with azimuth elevation as the slope, and continues to superimpose to obtain an azimuth-elevated building DSM. After filling and depression treatment, the elevated grid model ensures that the directionality of the flow conforms to the input direction. On this basis, calculate the flow direction with the roulette algorithm of double probability, calculate the probability with the reverse weighting of the slope difference and calculate the probability with the large weighting of the dominant direction. These two probability parameters ensure the transition and streamline of the wind. The azimuth-elevated raster model ensures the continuity of the fluid, and extracts the main flow channel by calculating the threshold multiple times. The main flow channel is marked as the air duct. In this way, high-performance multi-scale urban ventilation corridor construction-identification-extraction is carried out. This method solves, on the one hand, the problems of a large amount of computing resources and small spatial scale in the micro scale such as the CFD simulation method, and greatly improves the computing speed; on the other hand, it solves the problems of the lack of complex air flow interaction and aerodynamic analysis between buildings in the macro scale such as circuit simulation.

[0044] For example: Taking the building data of Wuhan as an example, construct grids with resolutions of 5 meters, 10 meters, 15 meters, 20 meters, 25 meters, 30 meters, 35 meters, and 40 meters for fishing net segmentation sampling. Continue to superimpose the DEM raster and the azimuth-elevated raster, and the output raster is the azimuth-elevated building DSM, as shown inFigure 2 。

[0045] Perform flow direction calculation and flow volume statistics on the building DSM, use the double-probability roulette algorithm proposed in this application to perform flow direction calculation, obtain the flow direction grid, and the flow direction coding is shown as Figure 3 , and perform flow volume calculation after depression filling. Calculate the air flow result once, perform 1000 times of flow direction calculation and flow volume calculation, set the threshold to statistically analyze the high air volume area, and the air duct can be identified in the high air volume area. Vectorize the high air volume area to complete the air duct extraction. The extraction result is shown in Figure 4 shown. For the accuracy evaluation of the air duct result, perform CFD wind environment simulation with the same area and resolution. The simulation result is shown in Figure 5 shown. Compare the high wind speed area of the CFD simulation with the extracted air duct result. If the similarity is high, it indicates that this method is relatively reliable.

[0046] Table 1

[0047] Spatial resolution 5m 10m 15m 20m 25m 30m 35m 40m CFD-time 7:12:05 1:28:35 0:41:56 0:20:52 0:15:37 0:14:30 0:7:54 0:8:33 GIS-time 6.8m 2.4m 70s 33s 7s 1s 1s 1s Similarity 0.542 0.484 0.421 0.646 0.614 0.654 0.748 0.763

[0048] (CFD-time represents the time consumed by CFD simulation, and GIS-time represents the time consumed by the calculation of this method)

[0049] As can be seen from Table 1, under the same hardware conditions and system environment, the time required for the calculation of this method is much less than that of CFD simulation. From the similarity index of the two, after the resolution is greater than 20 meters, it can be well similar to the CFD simulation result. At a smaller resolution, the similarity is low because the length and width of most buildings are rarely 5 meters, 10 meters, etc. The selection of the grid scale in this method is optimal under the condition of conforming to the scale of most buildings. Through multi-scale grids, it can be applied to urban conditions under different development models, which helps relevant departments predict the specific ventilation effects of air ducts in the design of urban ventilation corridors, and then provides assistance for improving the urban climate.

[0050] Example 2

[0051] A computer device includes a memory, a processor, and a computer program stored on the memory. The processor executes the computer program to implement the steps of the above method.

[0052] Example 3

[0053] A computer-readable storage medium stores a computer program, and when the program is executed by a processor, it implements the steps of the above method.

[0054] Example 4

[0055] A computer program product includes a computer program or instruction, and when the program or instruction is executed by a processor, it implements the steps of the above method.

[0056] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A high-performance multi-scale urban ventilation corridor construction-identification-extraction method, characterized in that: include: S1: Grid the building data, extract the average height of the buildings in the grid, overlay the DEM and the azimuthally lifted grid model, and perform noise processing on the grid model; S2: Analyze the flow direction of the grid model after filling, use the D8 algorithm of wind flow characteristics and the double-probability roulette algorithm to assign the flow direction, and obtain the flow direction result; S3: Use multiple flow statistics of the entire grid for the flow direction results, and accumulate the total number of each flow direction grid to obtain the wind volume grid; S4: According to the air volume grid, extract the air volume grid range where the ventilation volume is greater than the set threshold value to obtain the high air volume area, which is the ventilation corridor; S5: Binarize the high air volume area and the non-high air volume area, extract the grid air duct, convert the raster data into vector data, and complete the air duct vector extraction.

2. A high-performance multi-scale urban ventilation corridor construction-identification-extraction method as claimed in claim 1, characterized in that: Step S1 includes: S11: Fishnet sampling: Generate a grid according to the set resolution size, divide the study area into grids, extract the average height of the buildings in the grid as the value of the grid, and obtain an urban building grid model that only includes buildings; S12: Overlay regional DEM: Overlay the urban building grid model with the DEM obtained by fishing net sampling to obtain a terrain grid model; S13: Taking the entire study area as the object, raise the height of the wind source area, that is, construct a slope; S14: constructing the DSM of the building with azimuth elevation: the slope constructed in S13 is further superimposed on the grid model obtained in S12 to obtain the DSM of the building with azimuth elevation; S15: Fill the depressions in the DSM of the buildings that are lifted and remove the depressions in the DSM.

3. A high-performance multi-scale urban ventilation corridor construction-identification-extraction method as claimed in claim 1, characterized in that: In step S2, flow direction calculation is performed for the DSM of the building with azimuth elevation, including the following steps: S21: D8 algorithm based on wind flow characteristics: traverse all grid values ​​except the boundary rows, take the current pixel as the center pixel, take the elevation value of the current pixel, first calculate the absolute height difference between the center pixel and each adjacent pixel, then calculate the horizontal distance between the center pixel and each adjacent pixel, and finally calculate the slope: Among them, slope is the slope of the 8 neighboring pixels, neighbor_value is the elevation value of the 8 neighboring pixels around the current pixel, value is the elevation value of the current pixel, abs() is the absolute value calculation, and distance is the horizontal distance between the current pixel and the neighboring pixel; S22: Double-probability roulette algorithm: After obtaining the 8 adjacent slope changes, the slope difference is reversely weighted to calculate the probability. The smallest slope has the highest probability, and the largest slope has the smallest probability. The probability is calculated by weighting the dominant direction, and the slope position consistent with the current wind direction has the highest probability. The reverse slope weight and direction weight are combined with the roulette algorithm to obtain the flow direction, and the calculation result of the probability event is used as the flow direction grid of the current grid. The flow direction is marked, and the output of the flow direction is an integer grid with a value range between 1 and 255.

4. A high-performance multi-scale urban ventilation corridor construction-identification-extraction method as claimed in claim 1, characterized in that: In step S3, the total number of each flow direction grid is calculated and accumulated in the entire grid area. For each row of the grid, its path is counted as a wind source. Each time, the newly generated flow direction grid is used as the basis for calculation, so as to count all possible wind flow paths, and the result is output as an air volume grid.

5. A high-performance multi-scale urban ventilation corridor construction-identification-extraction method as claimed in claim 1, characterized in that: In step S5, a multi-scale fishing net is constructed for sampling, so as to complete the multi-resolution segmentation of the urban area, thereby extracting multi-scale urban wind ducts.

6. A computer device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the high-performance multi-scale urban ventilation corridor construction-identification-extraction method described in any one of claims 1-5.

7. A computer-readable storage medium, characterized in that: A computer program is stored, and when the program is executed by a processor, the steps of the high-performance multi-scale urban ventilation corridor construction-identification-extraction method described in any one of claims 1 to 5 are implemented.

8. A computer program product, characterized in that The method comprises a computer program or an instruction, which, when executed by a processor, implements the steps of the high-performance multi-scale urban ventilation corridor construction-identification-extraction method described in any one of claims 1 to 5.

Images