Urban ventilation corridor planning method and system based on high-precision positioning of beidou

Abstract

The application relates to the field of Beidou high-precision positioning technology and provides a city ventilation corridor planning method and system based on Beidou high-precision positioning, which solves the problems of poor accuracy and effectiveness of city ventilation corridor planning. The application obtains the ground temperature distribution, building form parameters and three-dimensional point cloud data of the city ventilation corridor candidate area, filters the three-dimensional point cloud data and constructs a curved surface model; obtains the ground roughness distribution from the curved surface model, aligns the ground roughness distribution, the ground temperature distribution and the building form parameters through a unified spatial coordinate reference provided by the Beidou high-precision positioning technology; adopts a deep learning model to process the aligned temperature distribution and building form parameters, identifies ventilation weak areas and cold source inlets; and based on a path planning algorithm, generates a city ventilation corridor planning path that avoids the ventilation weak areas, with the minimum roughness and the maximum cold source connectivity as the targets, so that the accuracy and effectiveness of city ventilation corridor planning can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of Beidou high-precision positioning technology, and in particular to a city ventilation corridor planning method and system based on Beidou high-precision positioning. BACKGROUND

[0002] In the city ventilation corridor planning scenario, the Beidou high-precision positioning technology needs to meet the following technical requirements: first, the precise mapping of building-level spatial information needs to be realized, for example, through centimeter-level positioning data, the quantitative correlation of building form parameters such as window-wall ratio, roof shape and urban wind-heat environment is obtained; second, the dynamic coupling capability of multi-source data needs to be constructed, the wind speed and surface deformation data monitored by Beidou are spatiotemporally aligned with the surface temperature and vegetation index retrieved by remote sensing, forming a stereoscopic data system covering macroscopic wind field simulation and microscopic meteorological monitoring; finally, a real-time calibration mechanism needs to be supported, through Beidou short message communication, the dynamic correction of the wind field model in the area without public network is realized, and the adaptability of the planning scheme to the dynamic changes of the city such as newly built buildings and vegetation replacement is ensured.

[0003] At present, for the above-mentioned needs, the industry adopts a collaborative design framework of multi-source data fusion as the main technical scheme. This scheme integrates the three-dimensional structural data of building information model, the spatial analysis capability of geographic information system, the large-scale surface monitoring data of remote sensing, and the high-precision positioning and real-time communication functions of Beidou, to build a multi-scale dynamic coupling model. The specific implementation path includes: using Landsat9 thermal infrared band to retrieve surface temperature, combining Sentinel-2 multispectral data to extract vegetation index; simulating building energy consumption by EnergyPlus and correlating GIS spatial data to quantify the superposition effect of building layout on local wind-heat environment; finally, verifying the ENVI-met wind field model output by the wind speed data obtained by the Beidou monitoring station, forming a closed-loop process of “data collection to simulation analysis to real-time calibration”.

[0004] However, this scheme has three significant defects: first, multi-source data fusion relies on complex spatio-temporal alignment algorithms, for example, the coordinate conversion error between BIM models and remote sensing images may cause local wind field simulation deviation, especially in densely built-up areas, the lack of three-dimensional modeling accuracy will further amplify the error; second, the real-time performance of the dynamic coupling model is limited, although Beidou can provide millimeter-level positioning data, the existing framework still responds to urban dynamic changes such as temporary building construction and seasonal vegetation replacement in weeks, which cannot meet the emergency planning needs of sudden environmental events such as short-term strong winds and pollutant leaks; third, ENVI-met simulation has limitations in complex urban forms, for example, the wind speed attenuation law in the shadow area of high-rise buildings is difficult to accurately depict through existing algorithms, leading to performance deviation of the actual landing of the ventilation corridor layout scheme. In addition, the framework lacks the ability to integrate unstructured data such as social media public opinion and citizen sensory feedback, making it difficult to achieve public participation-driven planning optimization. SUMMARY

[0005] The purpose of the present application is to provide a city ventilation corridor planning method, system, electronic device and storage medium based on Beidou high-precision positioning, to solve the problem of poor accuracy and effectiveness of city ventilation corridor planning caused by data fusion error, insufficient real-time performance, and limitations of complex form simulation in the prior art.

[0006] To solve the above technical problems, in a first aspect, the present application provides a city ventilation corridor planning method based on Beidou high-precision positioning, comprising:

[0007] Obtaining the ground temperature distribution, building form parameters and three-dimensional point cloud data of the city ventilation corridor candidate area, filtering the three-dimensional point cloud data, and constructing a surface model based on the filtered three-dimensional point cloud data;

[0008] Obtaining the ground roughness distribution from the surface model, and aligning the spatial coordinates of the ground roughness distribution, the ground temperature distribution and the building form parameters based on the unified spatial coordinate reference provided by the Beidou high-precision positioning technology;

[0009] Processing the ground temperature distribution and the building form parameters after spatial coordinate alignment using a deep learning model to identify ventilation weak areas and cold source inlets;

[0010] Based on a path planning algorithm, the minimum roughness corresponding to the ground roughness distribution and the maximum cold source connectivity corresponding to the cold source inlets are taken as the target, and according to the target, a planning path of the city ventilation corridor avoiding the ventilation weak areas is generated.

[0011] Optionally, the deep learning model is used to process the surface temperature distribution and the building form parameters after the spatial coordinate alignment to identify ventilation weak areas and cold source inlets, including:

[0012] According to the preset different range of temperature intervals, the surface temperature distribution after the spatial coordinate alignment is divided into multiple temperature regions, and the temperature value and temperature change of each temperature region are obtained.

[0013] The building form parameters after the spatial coordinate alignment are divided into form regions corresponding to the temperature regions, and the building height, building quantity and building gap size in each form region are extracted.

[0014] The temperature value, temperature change, building height, building quantity and building gap size are combined to obtain a data set.

[0015] The deep learning model is used to analyze the data set, and the spatial range of the ventilation weak area and the spatial position of the cold source inlet are determined according to the analysis result.

[0016] Optionally, the deep learning model is used to analyze the data set, and the spatial range of the ventilation weak area and the spatial position of the cold source inlet are determined according to the analysis result, including:

[0017] The temperature value in the data set is associated with the building height to obtain a first association analysis result, and the temperature change, building quantity and building gap size are associated to obtain a second association analysis result.

[0018] According to the first association analysis result, the deep learning model is used to determine whether the corresponding temperature region and the corresponding form region meet the preset ventilation poor characteristics.

[0019] According to the second association analysis result, the deep learning model is used to determine whether the corresponding temperature region and the corresponding form region meet the preset cold source inlet characteristics.

[0020] All temperature regions and corresponding form regions that meet the preset ventilation poor characteristics are combined to generate the spatial range of the ventilation weak area.

[0021] All edge intersection points of temperature regions and corresponding form regions that meet the preset cold source inlet characteristics are marked as the spatial position of the cold source inlet.

[0022] Optionally, based on the path planning algorithm, the roughness minimization corresponding to the surface roughness distribution and the cold source connectivity maximization corresponding to the cold source inlet are taken as the target, and according to the target, a planning path of the urban ventilation corridor avoiding the ventilation weak area is generated, including:

[0023] selecting a starting point and a target point of the urban ventilation corridor, the starting point and the target point corresponding to different cold source inlets respectively;

[0024] determining a roughness value corresponding to each continuous region in the curved surface model according to the ground roughness distribution;

[0025] setting a space range of a ventilation weak area as an area that the path cannot pass through;

[0026] searching for a plurality of first candidate paths from the starting point to the target point in the curved surface model, and minimizing a sum of roughness values of areas passed through by each first candidate path and maximizing a connection number of each first candidate path and a cold source inlet as a target;

[0027] selecting, based on the target, a first candidate path avoiding the area that the path cannot pass through as a planned path of the urban ventilation corridor.

[0028] Optionally, the selecting, based on the target, a first candidate path avoiding the area that the path cannot pass through as a planned path of the urban ventilation corridor comprises:

[0029] selecting, from all the first candidate paths, a first candidate path avoiding the area that the path cannot pass through as a second candidate path;

[0030] dividing each second candidate path into a plurality of continuous path segments, calculating a roughness value corresponding to each path segment, accumulating all the roughness values constituting the second candidate path to obtain a total roughness value corresponding to each second candidate path;

[0031] counting a number of cold source inlets passed through by each second candidate path as a connection number corresponding to each second candidate path;

[0032] calculating, based on the total roughness value and the connection number corresponding to each second candidate path, a target value of each second candidate path according to a weight corresponding to each target, and selecting a second candidate path with a maximum target value as a planned path of the urban ventilation corridor.

[0033] Optionally, the filtering the three-dimensional point cloud data and constructing a curved surface model based on the filtered three-dimensional point cloud data comprises:

[0034] selecting, according to coordinates of each point cloud data in the three-dimensional point cloud data, a point in a preset height range in the candidate area of the urban ventilation corridor as a to-be-processed point;

[0035] filtering each of the to-be-processed points by calculating distances between each of the to-be-processed points and a preset number of adjacent points around the to-be-processed point, and removing points whose distance exceeds a preset value, to complete the filtering processing;

[0036] connecting the three-dimensional point cloud data after the filtering processing in sequence according to an arrangement order of corresponding coordinates to form a plurality of polygonal surfaces connected to each other, and constructing a curved surface model based on the plurality of polygonal surfaces.

[0037] Optionally, the obtaining of the ground temperature distribution, the building form parameters and the three-dimensional point cloud data of the urban ventilation corridor candidate area comprises:

[0038] obtaining satellite thermal images and unmanned aerial vehicle images of the urban ventilation corridor candidate area;

[0039] performing radiation calibration on the satellite thermal images and the unmanned aerial vehicle images respectively to obtain thermal radiation data and reflection characteristic data, and performing multispectral fusion on the thermal radiation data and the reflection characteristic data to extract the ground temperature distribution and the building form parameters;

[0040] scanning a three-dimensional terrain of the urban candidate corridor area by using an unmanned aerial vehicle laser radar to obtain three-dimensional point cloud data.

[0041] In a second aspect, the present application provides a city ventilation corridor planning system based on Beidou high-precision positioning, comprising:

[0042] an obtaining module configured to obtain ground temperature distribution, building form parameters and three-dimensional point cloud data of a city ventilation corridor candidate area, perform filtering processing on the three-dimensional point cloud data, and construct a curved surface model based on the three-dimensional point cloud data after the filtering processing;

[0043] an optimization module configured to obtain a ground roughness distribution from the curved surface model, and perform spatial coordinate alignment on the ground roughness distribution, the ground temperature distribution and the building form parameters by using a unified spatial coordinate reference provided by a Beidou high-precision positioning technology;

[0044] an identification module configured to process the ground temperature distribution and the building form parameters after the spatial coordinate alignment by using a deep learning model to identify ventilation weak areas and cold source inlets;

[0045] a generation module configured to generate a planning path of a city ventilation corridor avoiding the ventilation weak areas based on a path planning algorithm, with minimization of roughness corresponding to the ground roughness distribution and maximization of cold source connectivity corresponding to the cold source inlets as targets.

[0046] In a third aspect, the present application provides an electronic device, comprising:

[0047] a memory configured to store a computer program;

[0048] The processor is configured to execute the computer program to implement the steps of the method for planning an urban ventilation corridor based on high-precision Beidou positioning according to the first aspect.

[0049] In a fourth aspect, the present application provides a computer readable storage medium, wherein a computer program is stored in the computer readable storage medium, and the computer program is executable by a processor to implement the steps of the method for planning an urban ventilation corridor based on high-precision Beidou positioning according to the first aspect.

[0050] In the present application, a method for planning an urban ventilation corridor based on high-precision Beidou positioning is provided, which comprises the following steps: obtaining ground temperature distribution, building form parameters and three-dimensional point cloud data of a candidate area of the urban ventilation corridor, performing filtering processing on the three-dimensional point cloud data, and constructing a curved surface model based on the filtered three-dimensional point cloud data; obtaining ground roughness distribution from the curved surface model, and performing spatial coordinate alignment on the ground roughness distribution, the ground temperature distribution and the building form parameters by using a unified spatial coordinate reference provided by high-precision Beidou positioning technology; processing the ground temperature distribution and the building form parameters after spatial coordinate alignment by using a deep learning model to identify ventilation weak areas and cold source inlets; and generating a planning path of the urban ventilation corridor that avoids the ventilation weak areas based on a path planning algorithm, with the goal of minimizing roughness corresponding to the ground roughness distribution and maximizing cold source connectivity corresponding to the cold source inlets, and generating the planning path according to the goal.

[0051] The present application has the following beneficial effects:

[0052] The method for planning an urban ventilation corridor based on high-precision Beidou positioning provided by the present application can provide a basic model for subsequent acquisition of key data such as ground roughness by obtaining ground temperature distribution, building form parameters and three-dimensional point cloud data of a candidate area of the urban ventilation corridor, performing filtering processing on the three-dimensional point cloud data and constructing a curved surface model, and can improve the effectiveness of the point cloud data by filtering processing; the ground roughness distribution can be obtained from the curved surface model, and the ground roughness distribution, the ground temperature distribution and the building form parameters can be aligned by using a unified spatial coordinate reference of high-precision Beidou positioning technology, which can realize the unification of different types of data in the same spatial coordinate system and provide spatial consistency guarantee for subsequent joint analysis; the deep learning model can be used to process the ground temperature distribution and the building form parameters after spatial coordinate alignment to identify ventilation weak areas and cold source inlets, which can utilize the analysis capability of deep learning to accurately identify key areas that are crucial to ventilation and clearly define the core target areas for corridor planning; and the planning path that avoids the ventilation weak areas can be generated based on the path planning algorithm with the goal of minimizing roughness and maximizing cold source connectivity, which can generate a reasonable corridor path that takes into account ventilation efficiency, cold source utilization and avoidance of adverse areas.

[0053] Further, the surface temperature distribution after spatial coordinate alignment is divided into multiple temperature regions, and the temperature values and temperature change conditions of each region are obtained. The corresponding building form parameters are divided into form regions, and the building height, number and inter-building gap size are extracted to form a data set. Then, by correlating the temperature values and building height, temperature change conditions, etc. in the data set with the number of buildings and gap size, a deep learning model is used to determine whether the region meets the poor ventilation and cold source in-out characteristics, and then determine the spatial range of the ventilation weak area and the spatial position of the cold source inlet. The technical effect is that by subdividing the region and correlating the temperature and building form parameters, combined with the accurate judgment of the deep learning model, the specific and accurate identification of the spatial range of the ventilation weak area and the position of the cold source inlet is realized, and more detailed spatial basis is provided for urban ventilation corridor planning. BRIEF DESCRIPTION OF DRAWINGS

[0054] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0055] Figure 1 A flowchart of a city ventilation corridor planning method based on Beidou high-precision positioning provided by an embodiment of the present application;

[0056] Figure 2 A specific implementation schematic diagram of a city ventilation corridor planning method based on Beidou high-precision positioning provided by an embodiment of the present application;

[0057] Figure 3 A specific implementation schematic diagram of another city ventilation corridor planning method based on Beidou high-precision positioning provided by an embodiment of the present application;

[0058] Figure 4 A structure schematic diagram of a city ventilation corridor planning system based on Beidou high-precision positioning provided by an embodiment of the present application. DETAILED DESCRIPTION

[0059] In order to solve the problems of data fusion error, insufficient real-time performance and limitation of complex form simulation in the prior art, an urban ventilation corridor planning method based on Beidou high-precision positioning is provided, which adopts the following design concept: first, the ground surface temperature, building form and three-dimensional terrain point data of the candidate area are collected, and the terrain point data is filtered to establish a terrain surface model; the roughness of the ground surface is extracted from the model, and the data of the ground surface roughness, temperature and building form can be corresponded through a unified position standard; then, the corresponding data is processed by an intelligent analysis model to accurately find out the areas with poor ventilation and the positions suitable for cold source inlets; finally, when planning the path, not only the terrain resistance to ventilation is considered to be as small as possible, but also the path is considered to be connected to more cold sources while avoiding the areas with poor ventilation. In this way, not only the problems of non-uniform data and inaccurate identification of key areas are solved, but also the planned ventilation corridor is more in line with the actual demand and better plays the role of ventilation.

[0060] In order to enable the personnel in the technical field to better understand the present application, the present application will be further described in detail below in combination with the drawings and specific embodiments. Obviously, the described embodiments are only part of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by the person skilled in the art without making creative efforts fall within the scope of protection of the present application.

[0061] The core of the present application is to provide an urban ventilation corridor planning method based on Beidou high-precision positioning, and a flowchart of a specific embodiment thereof is shown in Figure 1 The method comprises the following steps.

[0062] S11, the ground surface temperature distribution, building form parameters and three-dimensional point cloud data of the urban ventilation corridor candidate area are obtained, the three-dimensional point cloud data is filtered, and a curved surface model is constructed based on the filtered three-dimensional point cloud data.

[0063] The ground surface temperature distribution is the temperature condition of different positions in the urban ventilation corridor candidate area, including the temperature value of each point and the distribution state of the temperature in the area; the building form parameters are data describing the characteristics of the buildings in the area, including the height, number and distance between buildings; the three-dimensional point cloud data is a data set composed of a large number of three-dimensional coordinate points, which can reflect the three-dimensional shape of the ground surface and buildings in the area; the filtering processing is to select useful points from the three-dimensional point cloud data and remove points that do not meet the requirements; the curved surface model is constructed based on the filtered three-dimensional point cloud data, which can reflect the surface model of the three-dimensional form of the ground surface and buildings in the area.

[0064] In the embodiments of the present application, firstly, the ground surface temperature distribution, building shape parameters and three-dimensional point cloud data of the candidate area of the urban ventilation corridor are obtained by the data acquisition device, for example, in the A area, the ground surface temperature distribution is obtained by collecting the temperature values of different streets and building periphery with a temperature sensor, the building shape parameters are obtained by recording the height, number and building spacing of each building in the area with a surveying tool, and the three-dimensional point cloud data is obtained by acquiring a large number of three-dimensional coordinate points including the ground and building surface with a laser scanning device. Secondly, the three-dimensional point cloud data obtained is filtered, the average distance of each point and a predetermined number of adjacent points around the point is calculated, the points whose distance exceeds a certain range of the average distance are determined as abnormal points and removed, and the points that can accurately reflect the ground surface and building shape are retained, for example, the distance of each point and 10 adjacent points around the point is calculated, the average distance is 2 meters, and the points whose distance exceeds 4 meters are removed. Finally, a curved surface model is constructed based on the three-dimensional point cloud data after filtering, the points in the filtered three-dimensional point cloud data are sequentially connected according to their spatial positions, and a continuous surface model reflecting the terrain and three-dimensional shape of the buildings in the A area is formed.

[0065] S12, obtaining the ground surface roughness distribution from the curved surface model, and performing spatial coordinate alignment on the ground surface roughness distribution, the ground surface temperature distribution and the building shape parameters through the unified spatial coordinate reference provided by the Beidou high-precision positioning technology.

[0066] Among them, the ground surface roughness distribution is data extracted from the curved surface model, reflecting the fluctuation degree of the ground surface and building surface in the area, and the greater the fluctuation, the higher the roughness; the Beidou high-precision positioning technology is a technology that can provide accurate spatial coordinates, and the unified spatial coordinate reference provided by the technology is a unified position reference standard, which can keep the position information of different data consistent; the spatial coordinate alignment is to match the ground surface roughness distribution, the ground surface temperature distribution and the building shape parameters according to the unified spatial coordinate reference, so that the three are one-to-one corresponding in position.

[0067] In the embodiments of the present application, as Figure 2As shown, first, the surface roughness distribution is obtained from the curved surface model constructed from S11, for example, in the curved surface model of area A, by analyzing the fluctuation changes of different positions on the model surface, it is determined that the roughness of the flat area is smaller, and the roughness of the building area with high and low staggered buildings is larger. Secondly, through the unified spatial coordinate reference provided by the Beidou high-precision positioning technology, the surface roughness distribution, the surface temperature distribution and the building form parameters are aligned in space coordinates, and a unified coordinate system is established with the bottom center point of a landmark building in area A as the coordinate origin. The roughness value of each position in the surface roughness distribution, the temperature value of each point in the surface temperature distribution and the position information of each building in the building form parameter are all corresponded to the specific coordinate points in the coordinate system, so as to ensure that the three types of data at the same coordinate point can be matched with each other, such as associating the roughness value, temperature value and building parameter corresponding to the coordinate (10, 20, 0).

[0068] S13, using a deep learning model to process the surface temperature distribution and building form parameters after spatial coordinate alignment to identify the ventilation weak area and the cold source inlet.

[0069] Among them, the deep learning model is an intelligent model that can analyze data and identify rules independently; the surface temperature distribution and building form parameters after spatial coordinate alignment are temperature data and building feature data matched with each other under the same coordinate system after processing by S12; the ventilation weak area is an area with poor ventilation effect in the region; the cold source inlet is a position where cold air flow can enter, such as a passageway around a large water body or a green land.

[0070] In the embodiments of the present application, first, the surface temperature distribution and building form parameters after spatial coordinate alignment in S12 are input into the deep learning model, for example, the temperature value, building height, building quantity and building spacing data under the same coordinate in area A are input into the model; secondly, the deep learning model processes these data, analyzes the relationship between temperature distribution and building form, and learns the rules that the area with high temperature and dense buildings and high height may have poor ventilation, the area with obvious temperature change and large gap between buildings may be a cold source inlet; finally, based on the analysis result, the ventilation weak area and the cold source inlet are identified, for example, in area A, the model identifies the eastern area with dense buildings and continuously high temperature as the ventilation weak area, and identifies the western green land around the area with large building spacing and obviously low temperature as the cold source inlet.

[0071] S14, based on a path planning algorithm, taking the minimization of the roughness corresponding to the surface roughness distribution and the maximization of the cold source connectivity corresponding to the cold source inlet as the target, generating a planning path of the urban ventilation corridor avoiding the ventilation weak area according to the target.

[0072] Among them, the path planning algorithm is a method used to plan a route from the starting point to the end point; roughness minimization means that the surface roughness of the area through which the planned path passes should be as small as possible, that is, the terrain and buildings should have less obstruction to ventilation; cold source connectivity maximization means that the planned path should connect to as many cold source inlets as possible to introduce more cold airflow; avoiding weak ventilation areas means that the planned path should not pass through areas with poor ventilation.

[0073] In the embodiments of this application, such as Figure 3 As shown, firstly, based on the path planning algorithm, the objectives are to minimize the surface roughness distribution obtained in S12 and maximize the cold source connectivity corresponding to the cold source inlets identified in S13. For example, in area A, the algorithm calculates the total roughness of the areas traversed by different routes, selects the route with the smaller total, and simultaneously counts the number of cold source inlets connected by the route, prioritizing routes with more connections. Secondly, based on the above objectives, the algorithm avoids the weak ventilation areas identified in S13 when planning the route. For example, the algorithm bypasses the weak ventilation area in the east and tries to make the route pass through the western cold source inlet and the flat area in the north. Finally, the algorithm generates planned routes for urban ventilation corridors that meet the conditions. For example, in area A, a route is planned starting from the western cold source inlet, passing through the flat square in the north, avoiding the weak ventilation area in the east, and finally reaching another cold source inlet.

[0074] This application provides the following specific example: In candidate areas for urban ventilation corridors, temperature data at various locations is first obtained using temperature detection equipment mounted on a drone to obtain the surface temperature distribution. Building height, quantity, and spacing are obtained through a building information registration system to obtain building morphology parameters. Three-dimensional point cloud data, including the ground and building exterior walls, is obtained using a ground laser scanner. When filtering the three-dimensional point cloud data, the distance between each point and its 10 neighboring points is calculated, yielding an average distance of 2 meters. Outliers exceeding 4 meters are removed. The remaining points are then stitched together according to their spatial location to construct a surface model reflecting the regional topography and building morphology. Surface roughness distribution is extracted from this surface model, determining that the northern flat plaza has low roughness, while the southern building complex area has high roughness. Using BeiDou high-precision positioning technology, a unified coordinate system is established with the regional central plaza as the origin (0,0,0), and the roughness, temperature values, and building parameters at each location are mapped to specific coordinates for alignment. The aligned data was input into a deep learning model. The model analysis revealed that in the eastern area, temperatures were mostly above 30°C, buildings were over 35 meters tall, there were more than three buildings, and the spacing between them was less than 5 meters, identifying this area as a poorly ventilated zone. In the western area, temperatures were below 25°C and the spacing between buildings was 15 meters, identifying this area as a cold source inlet. Finally, a path planning algorithm was used to calculate the route. Route 1, from the western cold source inlet to the southern cold source inlet, passed through the flat northern area, had a total roughness of 50, connected the two cold source inlets, and avoided the poorly ventilated eastern area, thus being determined as the final planned route.

[0075] By performing S11-S14, the embodiment of the application constructs an accurate curved surface model by acquiring multiple types of basic data and filtering processing, thereby providing a reliable basis for subsequent analysis; the spatial alignment of multiple types of data is realized by means of a unified coordinate reference, thereby avoiding errors caused by position mismatch; the ventilation weak area and cold source inlet are accurately identified by using a deep learning model, thereby providing a clear target area for planning; the path planning algorithm is combined to generate a path that takes into account small roughness, high cold source connectivity and avoidance of ventilation weak areas, so that the urban ventilation corridor can more effectively reduce ventilation obstacles, utilize cold sources and improve the overall ventilation efficiency of the region.

[0076] In a possible embodiment, S13, the spatial coordinate-aligned ground surface temperature distribution and building form parameters are processed by using a deep learning model to identify ventilation weak areas and cold source inlets, including:

[0077] Step 131, according to the preset different range of temperature intervals, the spatial coordinate-aligned ground surface temperature distribution is divided into multiple temperature regions, and the temperature value and temperature change of each temperature region are obtained.

[0078] Among them, the preset different range of temperature intervals is a plurality of continuous temperature ranges set in advance, used for dividing the region according to temperature; the spatial coordinate-aligned ground surface temperature distribution is the temperature distribution of the region in the unified coordinate system after processing; the temperature region is a region with similar temperature range divided according to the preset temperature interval; the temperature value is the temperature size in each temperature region; the temperature change is the temperature change state in each temperature region with time or space.

[0079] In the embodiment of the application, first, the preset different range of temperature intervals is set, for example, three continuous temperature ranges of 20-25℃, 25-30℃ and 30-35℃ are set, then according to the spatial coordinate-aligned ground surface temperature distribution, the region whose temperature meets each preset interval is divided into the corresponding temperature region, for example, the region whose temperature is 20-25℃ is classified as temperature region 1, the region whose temperature is 25-30℃ is classified as temperature region 2, and the region whose temperature is 30-35℃ is classified as temperature region 3, and finally the temperature value and temperature change of each temperature region are obtained, for example, the average temperature of temperature region 1 is 23℃ and changes by no more than 2℃ within a day, and the average temperature of temperature region 3 is 32℃ and is 5℃ higher in the afternoon than in the morning.

[0080] Step 132, the spatial coordinate-aligned building form parameters are divided into form regions corresponding to the temperature regions, and the building height, building quantity and building gap size in each form region are extracted.

[0081] The building form parameters after spatial coordinate alignment are building feature data in a unified coordinate system after processing; the form region is a region corresponding to the temperature region in the spatial position, that is, each form region corresponds to the spatial range of a temperature region; the building height is the height of the building in the form region; the building quantity is the total number of buildings in the form region; and the building gap size is the distance between adjacent buildings in the form region.

[0082] In the embodiment of the present application, first, according to the spatial range of the temperature region divided in step 131, the part of the building form parameters after spatial coordinate alignment that coincides with the spatial range of each temperature region is divided into the corresponding form region, for example, the temperature region 1 corresponds to the northwest region, and the building form parameters in the region are divided into the form region 1, then the building height, the building quantity and the building gap size in each form region are extracted, for example, in the form region 1, the height of each building is measured and the average value is calculated, the total number of buildings in the region is counted, and the distance between adjacent buildings is measured and the average value is calculated.

[0083] Step 133, combining the temperature value, the temperature change, the building height, the building quantity and the building gap size to obtain a data set.

[0084] The data set is a whole data group formed by combining the temperature value and the temperature change of each temperature region and the building height, the building quantity and the building gap size of the corresponding form region according to the spatial correspondence, which contains the correlation information of the temperature and the building in the region.

[0085] In the embodiment of the present application, first, the temperature value and the temperature change of each temperature region obtained in step 131 are associated with the building height, the building quantity and the building gap size of the corresponding form region extracted in step 132, for example, the temperature value of the temperature region 1 is 22℃ and the temperature change is 3℃, the average building height of the form region 1 is 17.7 meters, the building quantity is 3, and the average gap is 9 meters, then these associated data are integrated into a data group to form a data set, for example, the data set contains the content of “temperature region 1: temperature 22℃, temperature change 3℃, building height 17.7 meters, building quantity 3, and gap 9 meters”.

[0086] Step 134, using a deep learning model to analyze the data set, and determining the spatial range of the ventilation weak area and the spatial position of the cold source inlet according to the analysis result.

[0087] The deep learning model is an intelligent analysis tool that can learn the data rule autonomously; the data set is the data group containing the correlation information of the temperature and the building formed in step 133; the spatial range of the ventilation weak area is the specific spatial boundary of the region with poor ventilation effect; and the spatial position of the cold source inlet is the specific place where the cold air flow can enter the region.

[0088] In the embodiments of the present application, the data set formed in step 133 is first input into a deep learning model, for example, the data set of region A is input into the model, the model will automatically learn the correlation between temperature and building parameters in the data, for example, the area with high temperature, dense and tall buildings, and small gaps may have poor ventilation, and the area with low temperature, sparse buildings, and large gaps may be a cold source inlet, then the model analyzes the data set according to the learned rule to determine whether each region meets the characteristics of the ventilation weak area or the cold source inlet, and finally determines the spatial range of the ventilation weak area and the spatial position of the cold source inlet according to the analysis result, for example, the model determines that temperature region 3 and corresponding morphology region 3 are the ventilation weak area and delimits the spatial boundary, and determines that the edge of temperature region 1 and corresponding morphology region 1 is the cold source inlet and determines the specific position.

[0089] The present application provides the following specific examples: in the surface temperature distribution of the spatial coordinates of region A after alignment, three preset temperature intervals of 20-25℃, 25-30℃ and 30-35℃ are first set, the region of 20-25℃ in the northwest is divided into temperature region 1, the region of 25-30℃ in the middle is divided into temperature region 2, and the region of 30-35℃ in the southeast is divided into temperature region 3; when calculating the average temperature of temperature region 1, the temperature values of 5 points 21℃, 22℃, 23℃, 24℃ and 20℃ are selected, the total is 110℃, and 22℃ is obtained by dividing by 5, and the temperature change of the region within a day is recorded as 3℃, and the temperature information of other regions is obtained in the same way. Next, each temperature region is divided into a corresponding morphology region, morphology region 1 has 3 buildings with heights of 15m, 20m and 18m, an average height of (15+20+18) ÷ 3 = 17.7m, a building number of 3, and an average gap of (8+10) ÷ 2 = 9m between adjacent buildings. The building parameters of other morphology regions are extracted in the same way. Then, the temperature value 22℃ and the temperature change 3℃ of temperature region 1 are combined with the building parameters of morphology region 1, the temperature value 27℃ and the temperature change 4℃ of temperature region 2 are combined with the corresponding building parameters of morphology region 2, and the temperature value 33℃ and the temperature change 5℃ of temperature region 3 are combined with the corresponding building parameters of morphology region 3 to form a data set. Finally, the data set is input into a deep learning model, and the model analyzes that the temperature region 3 and the corresponding morphology region 3 in the southeast are the ventilation weak area, and the edge of the temperature region 1 and the corresponding morphology region 1 in the northwest is the cold source inlet.

[0090] By performing steps 131-134, the embodiment of the application provides a basis with clear temperature properties for subsequent analysis by dividing the area by the preset temperature interval and obtaining temperature information; realizes the spatial matching of temperature and building characteristics by dividing the corresponding form area and extracting building parameters; integrates scattered related information by combining data to form a data set; accurately identifies the ventilation weak area and the cold source inlet by analyzing the data set through the deep learning model. The whole process is progressive, making the correlation analysis of temperature and building characteristics more accurate, providing reliable key areas and location information for urban ventilation corridor planning, and improving the scientificity and effectiveness of planning.

[0091] In a possible embodiment, step 134, analyzing the data set by using a deep learning model, determining the spatial range of the ventilation weak area and the spatial position of the cold source inlet according to the analysis result, comprises:

[0092] a1, correlating and analyzing the temperature value in the data set with the building height to obtain a first correlation analysis result, and correlating and analyzing the temperature change, the number of buildings and the size of the gap between buildings to obtain a second correlation analysis result.

[0093] Wherein, the data set is a correlation data group containing the temperature value of the temperature area, the temperature change and the building height, the number of buildings and the size of the gap between buildings of the corresponding form area; the temperature value is the temperature size in the temperature area; the building height is the height of the building in the form area; the first correlation analysis result is the relationship conclusion obtained by correlating and analyzing the temperature value with the building height; the temperature change is the temperature variation state in the temperature area; the number of buildings is the total number of buildings in the form area; the size of the gap between buildings is the distance between adjacent buildings in the form area; the second correlation analysis result is the relationship conclusion obtained by correlating and analyzing the temperature change, the number of buildings and the size of the gap between buildings.

[0094] In the embodiment of the application, first, the temperature value of each temperature area and the building height of the corresponding form area are extracted from the data set, the relationship between the two is analyzed to obtain the first correlation analysis result, for example, in area A, by comparing the temperature values and building heights of different areas, it is found that the building height in the area with higher temperature is usually higher. Secondly, the temperature change of each temperature area, the number of buildings and the size of the gap between buildings of the corresponding form area are extracted from the data set, the relationship among the three is analyzed to obtain the second correlation analysis result, for example, it is found that in the area with obvious temperature change, the number of buildings is more and the gap between buildings is smaller.

[0095] a2, according to the first correlation analysis result, using a deep learning model to judge whether the corresponding temperature area and the corresponding form area meet the preset ventilation poor characteristics.

[0096] The first correlation analysis result is a relationship conclusion obtained after correlation analysis of a temperature value and a building height; the deep learning model is an intelligent analysis tool capable of autonomously learning data rules; the preset poor ventilation feature is a feature possessed by a region with poor ventilation effect, such as high temperature and tall building; the temperature region is a region divided according to a preset temperature interval; and the form region is a building feature region corresponding to the temperature region.

[0097] In the embodiment of the present application, the first correlation analysis result is first input into the deep learning model, for example, the correlation relationship between the temperature and the building height in the A region is input into the model. Secondly, the model judges whether each temperature region and the corresponding form region meet the preset poor ventilation feature one by one according to the preset poor ventilation feature. For example, if the preset feature is “high temperature and tall building”, the model will check whether the temperature value and the building height of each region meet this condition at the same time.

[0098] a3. According to the second correlation analysis result, the deep learning model is used to judge whether the corresponding temperature region and the corresponding form region meet the preset cold source in-out feature.

[0099] The second correlation analysis result is a relationship conclusion obtained after correlation analysis of temperature variation, building quantity, and building gap size; the deep learning model is an intelligent analysis tool capable of autonomously learning data rules; the preset cold source in-out feature is a feature possessed by a region suitable for cold air flow in and out, such as small temperature variation, few buildings, and large gap; the temperature region is a region divided according to a preset temperature interval; and the form region is a building feature region corresponding to the temperature region.

[0100] In the embodiment of the present application, the second correlation analysis result is first input into the deep learning model, for example, the correlation relationship between the temperature variation, the building quantity, and the gap size in the A region is input into the model. Secondly, the model judges whether each temperature region and the corresponding form region meet the preset cold source in-out feature one by one according to the preset cold source in-out feature. For example, if the preset feature is “small temperature variation, few buildings, and large gap”, the model will check whether the related parameters of each region meet this condition at the same time.

[0101] a4. All temperature regions and the corresponding form regions meeting the preset poor ventilation feature are combined to generate a spatial range of a weak ventilation area.

[0102] The preset poor ventilation feature is a feature possessed by a region with poor ventilation effect; the temperature region is a region divided according to a preset temperature interval; the form region is a building feature region corresponding to the temperature region; and the spatial range of the weak ventilation area is the specific spatial boundary of the region with poor ventilation effect.

[0103] In the embodiments of the present application, first, all temperature regions and corresponding form regions that meet the preset poor ventilation characteristics are collected, for example, in the A region, all regions that meet the "high temperature and high building" are collected. Secondly, the spatial ranges of these regions are combined to form a continuous boundary as the spatial range of the weak ventilation area.

[0104] a5, the edge intersection points of all temperature regions and corresponding form regions that meet the preset cold source inlet and outlet characteristics are marked as the spatial positions of the cold source inlets.

[0105] Wherein, the preset cold source inlet and outlet characteristics are characteristics of the region suitable for the inlet and outlet of cold air flow; the temperature region is a region divided according to a preset temperature interval; the form region is a building feature region corresponding to the temperature region; the edge intersection point is a point where the edge of the temperature region intersects with the corresponding form region; and the spatial position of the cold source inlet is a specific place where the cold air flow can enter the region.

[0106] In the embodiments of the present application, first, all temperature regions and corresponding form regions that meet the preset cold source inlet and outlet characteristics are collected, for example, in the A region, all regions that meet the "small temperature change, few buildings and large gap" are collected. Secondly, the points where the edges of these regions intersect are found, and these points are marked as the spatial positions of the cold source inlets.

[0107] The present application provides the following specific examples: in the data set of the A region, step a1 first extracts the temperature values of each temperature region and the building heights of the corresponding form regions, analyzes that the higher the temperature is, the higher the building height is, and obtains a first correlation analysis result; and then extracts the temperature change, the number of buildings and the gap between buildings of each region, analyzes that the greater the temperature change is, the more the number of buildings is and the smaller the gap is, and obtains a second correlation analysis result. Step a2 inputs the first correlation analysis result into the model, and determines that region 3 and the corresponding form region meet the preset poor ventilation characteristics according to the "temperature ≥ 30 ℃ and building height ≥ 28 m" characteristics. Step a3 inputs the second correlation analysis result into the model, and determines that region 1 and the corresponding form region meet the preset cold source inlet and outlet characteristics according to the "temperature change ≤ 3 ℃, building number ≤ 4 and gap ≥ 8 m" characteristics. Step a4 combines region 3 and the corresponding form region to delimit the southeast continuous region as the spatial range of the weak ventilation area. Step a5 finds the edge intersection points (such as the northeast and northwest coordinate points) of region 1 and the corresponding form region, and marks them as the spatial positions of the cold source inlets.

[0108] By performing a1~a5, the embodiment of the application analyzes the internal relationship between temperature and building parameters through correlation analysis, provides a basis for regional feature judgment, accurately identifies regions that meet the conditions of poor ventilation and cold source entry and exit by using a deep learning model combined with preset features, and then determines the spatial range of the ventilation weak area and the location of the cold source inlet by combining the regions and marking the intersection points. The whole process is progressive, making the identification of the key ventilation area more accurate and clear, providing a reliable spatial basis for urban ventilation corridor planning, and improving the scientificity and effectiveness of planning.

[0109] In a possible embodiment, S14, based on the path planning algorithm, aims to minimize the roughness corresponding to the surface roughness distribution and maximize the cold source connectivity corresponding to the cold source inlet, and generates a planning path of the urban ventilation corridor avoiding the ventilation weak area according to the target, including:

[0110] Step 141, selecting a starting point and a target point of the urban ventilation corridor, the starting point and the target point corresponding to different cold source inlets respectively.

[0111] Among them, the starting point of the urban ventilation corridor is the starting point of the ventilation corridor, and the target point is the terminal point of the ventilation corridor, both of which are end points of the planning path; the cold source inlet is the specific place where the cold air flow can enter the area, and the starting point and the target point correspond to different cold source inlets respectively, that is, the starting point and the terminal point are connected to a position where cold air flow enters respectively.

[0112] In the embodiment of the application, first, two different positions are selected from the identified cold source inlets as the starting point and the target point of the urban ventilation corridor, for example, in the A region, 4 cold source inlets are marked, the cold source inlet in the northwest is selected as the starting point, and the cold source inlet in the southeast is selected as the target point, to ensure that the two correspond to different cold source inlets and determine the starting point and the terminal point for subsequent path planning.

[0113] Step 142, determining the roughness value corresponding to each continuous region in the surface model according to the surface roughness distribution.

[0114] Among them, the surface roughness distribution is data reflecting the fluctuation degree of the ground surface and the building surface in the region; the surface model is a surface model reflecting the three-dimensional form of the region based on three-dimensional point cloud data; the continuous region is a part of the surface model in which the ground surface features are similar and connected; the roughness value is a value used to represent the fluctuation degree of each continuous region, and the larger the fluctuation, the larger the value.

[0115] In the embodiments of the present application, firstly, the surface roughness distribution of the ground is referenced to analyze the surface features of each part of the curved surface model, and the curved surface model is divided into multiple continuous regions, for example, in the curved surface model of region A, the flat square is divided into a continuous region, and the high and low staggered building group is divided into another continuous region. Secondly, a corresponding roughness value is assigned to each continuous region, which is set according to the undulating degree of the region, for example, the roughness value of the flat square is set to 2, and the roughness value of the building group region is set to 8, and the larger the value, the greater the hindrance of the region to ventilation.

[0116] Step 143, set the spatial range of the ventilation weak area as an area that the path cannot pass through.

[0117] Among them, the spatial range of the ventilation weak area is the specific spatial boundary of the area with poor ventilation effect, that is, the range of the area with poor ventilation; the area that the path cannot pass through is an area that the planned path cannot pass through, and setting the ventilation weak area as such an area means that the planned path needs to avoid places with poor ventilation effect.

[0118] In the embodiments of the present application, firstly, the spatial range of the determined ventilation weak area is determined, for example, in region A, the ventilation weak area is the building group region in the southeast, and the spatial range is a polygon region from the east boundary to the west boundary, and from the south boundary to the north boundary. Secondly, the range is set as an area that the path cannot pass through, that is, the ventilation corridor path planned subsequently cannot pass through the area, so as to ensure that the path avoids places with poor ventilation effect.

[0119] Step 144, search for multiple first candidate paths from the starting point to the target point in the curved surface model, and minimize the sum of roughness values of the regions passed by each first candidate path and maximize the number of connections of each first candidate path with the cold source inlet as the target.

[0120] Among them, the curved surface model is a surface model reflecting the three-dimensional form of the region; the first candidate path is a possible planned path from the starting point to the target point, and there are multiple; the sum of roughness values is the sum of the roughness values of all continuous regions passed by a path; the number of connections of the cold source inlet is the total number of cold source inlets passed by the path; the target refers to the conditions that the path planning needs to meet, that is, the minimization of the sum of roughness values and the maximization of the number of cold source connections.

[0121] In the embodiments of the present application, firstly, a plurality of possible paths from the starting point to the target point are searched in the curved surface model as first alternative paths, for example, in the A region, 5 different paths from the northwest starting point to the southeast target point are searched. Secondly, the sum of the roughness values of the regions through which each first alternative path passes is calculated, for example, the roughness values of the regions through which path 1 passes are 3, 5 and 4 respectively, and the sum is 3+5+4=12; the roughness values of the regions through which path 2 passes are 5, 9 and 6 respectively, and the sum is 5+9+6=20. At the same time, the number of cold source inlets through which each path passes is counted, for example, path 1 passes through 2 cold source inlets, and path 2 passes through 1 cold source inlet. Finally, minimizing the sum of the roughness values and maximizing the number of cold source connections are taken as the target of path selection, that is, paths with small sum and large number of connections are preferentially selected.

[0122] Step 145, based on the target, the first alternative path that avoids the region that the path cannot pass through is selected as the planning path of the urban ventilation corridor.

[0123] Wherein, the target refers to the conditions of “minimizing the sum of roughness values” and “maximizing the number of cold source connections” that the path planning needs to meet; the region that the path cannot pass through is the spatial range of the ventilation weak area, that is, the region that the planning path needs to avoid; the first alternative path is the possible path from the starting point to the target point; and the planning path of the urban ventilation corridor is the finally determined ventilation corridor route.

[0124] In the embodiments of the present application, firstly, according to the target set in step 144, paths that meet the conditions of “smaller sum of roughness values and more number of cold source connections” are selected from the plurality of first alternative paths, for example, among the 4 alternative paths in the A region, paths A and C are selected. Secondly, it is checked whether these selected paths pass through the region that the path cannot pass through (the ventilation weak area), for example, path C passes through the ventilation weak area in the southeast, and path A does not pass through this region. Finally, the path that meets the target and avoids the region that cannot be passed through is selected as the planning path, that is, path A is selected as the planning path of the urban ventilation corridor.

[0125] The application provides the following specific examples: in a region, step 141 selects location 1 in the northwest part of the marked cold source entrance as a starting point and location 3 in the southeast part as a target point, both of which correspond to different cold source entrances; step 142 divides the curved surface model into five continuous regions according to the ground roughness distribution and respectively assigns roughness values 3, 5, 9, 4 and 7; step 143 sets the southeast part of the weak ventilation area as an area that cannot be passed through; step 144 searches out four first alternative paths, and calculates that the sum of the roughness values of the areas passed through by path A is 3+4+5=12, path A connects two cold source entrances, the sum of path C is 3+5+6=14, path C connects two cold source entrances, and the sums of the other paths are higher or the connection numbers are less; and step 144 checks and finds that path C passes through the weak ventilation area and path A does not pass through, and therefore path A is selected as the planned path.

[0126] By performing steps 141-145, the embodiment of the application clearly defines the basic endpoints for path planning by selecting starting and ending points corresponding to different cold source entrances and ensures that cold air flows can be introduced at both ends; the influence of the ground on ventilation is quantified by setting roughness values, which provides a basis for path resistance evaluation; the weak ventilation area is set as a forbidden area to avoid the path passing through the low-efficiency area; and the low-resistance and high-cold-source-connection planning paths are ensured by multi-path search and target screening. The entire process is progressive, and the ventilation corridor planning path finally formed can effectively improve the regional ventilation efficiency and play a good ventilation role.

[0127] In a possible embodiment, step 145, based on the target, selects the first alternative path that avoids the area that the path cannot pass through as the planning path of the urban ventilation corridor, comprising:

[0128] b1, screening the first alternative path that avoids the area that the path cannot pass through from all the first alternative paths as the second alternative path.

[0129] Wherein, the first alternative path is a plurality of possible planning paths from the starting point to the target point, the area that the path cannot pass through is a weak ventilation area with poor ventilation effect, that is, an area that the planning path needs to avoid, and the second alternative path is a path selected from the first alternative path and not passing through the area that the path cannot pass through, which is a candidate path for further evaluation.

[0130] In the embodiment of the application, firstly, whether each first alternative path passes through the area that the path cannot pass through is checked one by one, for example, there are four first alternative paths in the A region, two of which pass through the weak ventilation area and the other two do not pass through, and secondly, the first alternative path that does not pass through the area that the path cannot pass through is screened out as the second alternative path, providing paths that meet the basic conditions for subsequent evaluation.

[0131] b2. Divide each second alternative path into multiple consecutive path segments, calculate the roughness value corresponding to each path segment, and sum up all the roughness values ​​that make up the second alternative path to obtain the total roughness value corresponding to each second alternative path.

[0132] Among them, the second alternative path is a candidate path that has not passed through the area that the path cannot pass through after screening. The path segment is a number of parts that divide the second alternative path into continuous spatial ranges. The roughness value is a value that represents the degree of undulation of the area where each path segment is located, and the larger the undulation, the larger the value. The total roughness value is the result of adding up the roughness values ​​of all path segments in a second alternative path, reflecting the degree of obstruction of the path to ventilation as a whole.

[0133] In this embodiment, each second alternative path is first divided into multiple continuous path segments. For example, in region A, the two second alternative paths are evenly divided into 3 and 4 segments according to their lengths, respectively. Then, the roughness value corresponding to each path segment is determined. For example, the first segment corresponds to a value of 2, the second segment corresponds to a value of 5, etc. Finally, the roughness values ​​of all path segments are summed to obtain the total roughness value. For example, the sum of 3 segments is 2+5+3=10, and the sum of 4 segments is 4+4+6+2=16.

[0134] b3. Count the number of cold source inlets that each second alternative path passes through, and use this as the number of connections corresponding to each second alternative path.

[0135] The second alternative path is a candidate path after screening. The cold source inlet is the specific location in the area that the cold air can enter. The number of connections refers to the total number of cold source inlets that a second alternative path passes through, reflecting the path's ability to introduce cold air.

[0136] In this embodiment of the application, the location of all cold source inlets in the area is first determined. For example, there are 4 cold source inlets in area A distributed in different directions. Then, the cold source inlets passed by each second alternative path are checked one by one and the number is counted. For example, one path passes through the cold source inlets in the northwest and northeast, and the number of connections is 2. Another path passes through the cold source inlets in the northwest and southwest, and the number of connections is also 2.

[0137] b4. Based on the total roughness value and number of connections corresponding to each second alternative path, calculate the target value of each second alternative path according to the weight of each objective, and select the second alternative path with the largest target value as the planning path of the urban ventilation corridor.

[0138] Wherein, the total roughness value is the overall ventilation resistance degree of the second alternative path, the connection number is the total number of cold source inlets passed by the path, the weight is a value reflecting the importance of the two objectives of total roughness minimization and connection number maximization, the target value is a comprehensive score calculated by combining the total roughness, the connection number and the corresponding weight for evaluating the overall advantages and disadvantages of the path, and the planning path of the urban ventilation corridor is the final selected optimal path.

[0139] In the embodiments of the present application, firstly, the weights of the two objectives are set, for example, the total roughness weight is 0.4 and the connection number weight is 0.6, secondly, the total roughness and the connection number are standardized, for example, the total roughness of 10 corresponds to 0.8 and the total roughness of 16 corresponds to 0.5, and the connection number of 2 corresponds to 0.9, and finally, the target value is calculated according to the formula: target value = total roughness standardized value × weight + connection number standardized value × weight, for example, the target values of the two paths are 0.86 and 0.74 respectively, and the path with the larger target value is selected as the planning path.

[0140] The present application provides the following specific examples: in the region, step b1 selects path B and path D which do not pass through the southeast ventilation weak area from the 4 first alternative paths as the second alternative paths; step b2 divides path B into 3 path segments with roughness values of 2, 5 and 3 respectively, and the total roughness is 2+5+3=10, and divides path D into 4 path segments with values of 4, 4, 6 and 2 respectively, and the total roughness is 4+4+6+2=16; step b3 statistics shows that both paths pass through 2 cold source inlets, and the connection number is 2; step b4 sets the total roughness weight to 0.4 and the connection number weight to 0.6, the total roughness standardized value of path B is 0.8, the connection number standardized value is 0.9, and the target value is 0.86, the total roughness standardized value of path D is 0.5, the connection number standardized value is 0.9, and the target value is 0.74, and finally path B is selected as the planning path.

[0141] By performing b1~b4, the embodiments of the present application ensure that the candidate paths meet the basic requirements by screening paths that avoid ventilation weak areas, and the ventilation resistance and cold source utilization capacity of the paths are quantified by calculating the total roughness and the connection number, and the advantages and disadvantages of the paths are comprehensively evaluated by calculating the target value with the weight. The whole process is progressive, so that the finally selected planning path can avoid ventilation inefficient areas and balance ventilation resistance and cold source utilization, thereby improving the rationality and effectiveness of urban ventilation corridor planning.

[0142] In a possible embodiment, S11, the three-dimensional point cloud data is filtered, and a curved surface model is constructed based on the filtered three-dimensional point cloud data, comprising:

[0143] Step 111, according to the coordinates of each point cloud data in the three-dimensional point cloud data, selecting points in a preset height range in the candidate region of the urban ventilation corridor as the to-be-processed points.

[0144] The three-dimensional point cloud data is a data set composed of a large number of three-dimensional coordinate points, which can reflect the three-dimensional shape of the ground surface and buildings in the region; the coordinates are the position information of each point in the three-dimensional space; the urban ventilation corridor candidate region is a region range in which a ventilation corridor can be planned; the preset height range is a height interval set in advance for filtering points; and the to-be-processed points are points selected from the three-dimensional point cloud data, located in the candidate region and in the preset height range, and are objects of subsequent processing.

[0145] In the embodiment of the present application, first, the coordinates of each point in the three-dimensional point cloud data are viewed to determine whether the point is located in the urban ventilation corridor candidate region, for example, in region A, the candidate region is the range of east longitude X1 to X2 and north latitude Y1 to Y2, and points with coordinates in the range are selected. Secondly, it is checked whether the height of the points is in the preset height range, for example, the preset height range is 0 to 50 meters, and points with a height between 0 and 50 meters are selected. The points that meet both the region and height conditions are taken as to-be-processed points.

[0146] Step 112, the distance between each to-be-processed point and a preset number of adjacent points around the to-be-processed point is calculated, and points with a distance exceeding a preset value are removed to complete the filtering process.

[0147] In the embodiment of the present application, first, the coordinates of each point in the three-dimensional point cloud data are viewed to determine whether the point is located in the urban ventilation corridor candidate region, for example, in region A, the candidate region is the range of east longitude X1 to X2 and north latitude Y1 to Y2, and points with coordinates in the range are selected. Secondly, it is checked whether the height of the points is in the preset height range, for example, the preset height range is 0 to 50 meters, and points with a height between 0 and 50 meters are selected. The points that meet both the region and height conditions are taken as to-be-processed points.

[0148] In the embodiment of the present application, first, a preset number of adjacent points around each to-be-processed point are selected, for example, 10 nearest adjacent points around each to-be-processed point in region A are selected. Secondly, the distance between the to-be-processed point and the 10 adjacent points is calculated, for example, the distances between a to-be-processed point and adjacent points are 2 meters, 3 meters, 5 meters, etc. Finally, points with a distance exceeding a preset value are removed, for example, the preset value is 4 meters, and points with a distance greater than 4 meters are removed to complete the filtering process.

[0149] Step 113, the three-dimensional point cloud data after filtering is sequentially connected according to the arrangement order of the corresponding coordinates to form a plurality of polygonal surfaces connected to each other, and a curved surface model is constructed based on the plurality of polygonal surfaces.

[0150] The three-dimensional point cloud data after filtering processing is an effective point reserved after step 112 processing; the arrangement order of the coordinates is an order of arranging the points according to the spatial coordinate positions; the polygonal face is a planar figure formed by connecting a plurality of points; the mutually connected polygonal faces are edges connected between the polygons, forming a continuous surface; and the curved surface model is a surface model based on a plurality of polygonal faces and capable of reflecting the three-dimensional form of the region.

[0151] In the embodiment of the present application, firstly, according to the coordinate arrangement order of the points after filtering processing, the adjacent points are sequentially connected to form a plurality of polygonal faces, for example, in the A region, four points with adjacent coordinates are connected to form a quadrilateral face, and a plurality of such faces are mutually connected. Secondly, the mutually connected polygonal faces are integrated to construct a continuous curved surface model, which can reflect the three-dimensional form of the ground surface and the building in the region.

[0152] The present application provides the following specific examples: in a region, first, the points with the coordinates of 100° to 101° east longitude and 30° to 31° north latitude (candidate region, the ranges of east longitude and north latitude in the present application are exemplary values, not real coordinates) and the height of 0 to 50 meters (preset height range) in the three-dimensional point cloud data are selected as the points to be processed; then, 10 adjacent points around each point to be processed are selected, the distances of the points to the adjacent points are calculated as 2 meters, 3 meters, 5 meters, etc., the preset value is 4 meters, the points with the distance of 5 meters are removed, and the filtering is completed; finally, the remaining points are connected to form mutually connected polygonal faces according to the coordinate order, and a curved surface model reflecting the three-dimensional form of the region is constructed.

[0153] By performing steps 111 to 113, the embodiment of the present application narrows the data range by screening the points to be processed, ensures that the processing object is related to the ventilation corridor planning, removes the abnormal points by filtering, improves the data accuracy, converts the discrete points into a continuous three-dimensional surface by constructing a curved surface model, and intuitively presents the region form. The entire process provides reliable basic data and model support for subsequent steps, and improves the pertinence and effectiveness of the early-stage data processing of planning.

[0154] In a possible embodiment, S11, the ground surface temperature distribution, the building form parameters and the three-dimensional point cloud data of the candidate region of the urban ventilation corridor are acquired, including:

[0155] c1, satellite thermal images and unmanned aerial vehicle images of the candidate region of the urban ventilation corridor are acquired.

[0156] The candidate region of the urban ventilation corridor is a region range in which the ventilation corridor is possibly planned; the satellite thermal images are images captured by a satellite and capable of reflecting the ground surface temperature distribution of the region; and the unmanned aerial vehicle images are images captured by an unmanned aerial vehicle and capable of presenting the details of the buildings and the terrain in the region.

[0157] In the embodiments of the present application, first, the range of the candidate urban ventilation corridor region is determined, for example, the range of region A is from longitude X1 to X2 and latitude Y1 to Y2. Second, satellite thermal images of the region are obtained by satellite remote sensing technology, which can show the temperature difference at different positions; at the same time, the unmanned aerial vehicle images are taken by controlling the unmanned aerial vehicle above the region to capture the details of the shape and distribution of buildings, for example, in region A, the satellite thermal images show that the southeast part has a higher temperature and the northwest part has a lower temperature, and the unmanned aerial vehicle images clearly show the height and arrangement of the buildings in the region.

[0158] c2, the satellite thermal images and the unmanned aerial vehicle images are respectively radiometrically calibrated to obtain thermal radiation data and reflection characteristic data, and the thermal radiation data and the reflection characteristic data are multispectral fused to extract the ground temperature distribution and the building morphology parameters.

[0159] Among them, the radiometric calibration is to correct the image data to eliminate the errors caused by instruments or environment, so that the data is more accurate; the thermal radiation data is the data obtained from the satellite thermal images after calibration, which reflects the intensity of the ground thermal radiation and can be used to calculate the temperature; the reflection characteristic data is the data obtained from the unmanned aerial vehicle images after calibration, which reflects the light reflection ability of the ground and buildings, and can be used to identify the building features; the multispectral fusion is to combine the thermal radiation data and the reflection characteristic data, and integrate the information of the two kinds of data; the ground temperature distribution is the temperature condition at different positions in the region; the building morphology parameters are data describing the building features, such as height, number, etc.

[0160] In the embodiments of the present application, first, the satellite thermal images are radiometrically calibrated, for example, by eliminating atmospheric interference and instrument errors, the gray value in the images is converted into actual thermal radiation data, so that the data can accurately reflect the ground temperature; at the same time, the unmanned aerial vehicle images are radiometrically calibrated to correct the influence of light changes during shooting, so as to obtain data that can truly reflect the reflection characteristics of buildings, for example, in region A, the calibrated thermal radiation data shows that the radiation value in the southeast part is 20 units higher than that in the northwest part, and the edges of the buildings in the calibrated unmanned aerial vehicle images are clearer. Second, the thermal radiation data and the reflection characteristic data are multispectral fused to combine the temperature information and the building features, and the ground temperature distribution and the building morphology parameters are extracted therefrom, for example, after fusion, it is clear that the high temperature area in the southeast of region A corresponds to dense high-rise buildings, and the low temperature area in the northwest corresponds to low buildings and green land.

[0161] c3, the three-dimensional terrain of the candidate urban corridor region is scanned by the unmanned aerial vehicle laser radar to obtain three-dimensional point cloud data.

[0162] The unmanned aerial vehicle laser radar is a device installed on the unmanned aerial vehicle, which measures the distance by emitting a laser beam and receiving the reflected signal to obtain the three-dimensional information of the terrain and objects; and the three-dimensional point cloud data is a data set composed of a large number of three-dimensional coordinate points, which are densely arranged and can reflect the three-dimensional shape of the ground surface and buildings in the region.

[0163] In the embodiment of the present application, first, the laser radar device is installed on the unmanned aerial vehicle, and the scanning parameters such as the scanning range and the laser emission frequency are set, for example, in the A area, the scanning range covers the entire candidate area, and the laser emission frequency is 1 million times per second. Secondly, the unmanned aerial vehicle is controlled to fly over the candidate area according to the preset flight route, the laser radar device continuously emits laser beams to the ground, receives the reflected signals and calculates the three-dimensional coordinates of each point, for example, the unmanned aerial vehicle takes off from the northwest of the A area, flies along the grid-shaped flight route, the laser beam covers every position on the ground, records the coordinate information of the ground, the top and side of the building, and finally generates three-dimensional point cloud data containing millions of points, which can clearly present the height of the building and the undulation of the terrain.

[0164] The present application provides the following specific examples: in the region, first, the candidate area of the urban ventilation corridor is defined as 100° to 101° east longitude and 30° to 31° north latitude, the thermal image of the region is obtained by satellite, and the image of the region is shot by unmanned aerial vehicle; then the satellite thermal image is radiometrically calibrated to convert the gray value into thermal radiation data, the unmanned aerial vehicle image is calibrated to correct the uneven brightness, and the two kinds of data are fused to extract the ground temperature distribution and building shape parameters; finally, the unmanned aerial vehicle carrying the laser radar is flown in the region at an interval of 50 meters, the laser emission frequency is set to 800,000 times per second, the laser beam is emitted and the reflected signal is received, the three-dimensional coordinates of the ground and the building are recorded, and three-dimensional point cloud data containing 5 million points are generated to present the three-dimensional shape of the region.

[0165] By performing c1-c3, the embodiment of the present application provides raw materials for analyzing the temperature and building characteristics of the region by obtaining the satellite thermal image and the unmanned aerial vehicle image; the data accuracy is improved and the accurate ground temperature distribution and building shape parameters are extracted by radiometric calibration and multispectral fusion; and detailed three-dimensional terrain data are obtained by scanning the unmanned aerial vehicle laser radar. The whole process collects and processes data from multiple dimensions to provide comprehensive and reliable basic information for the planning of the urban ventilation corridor, ensuring that the subsequent analysis and planning are more in line with the actual situation.

[0166] Figure 4 A structure diagram of a city ventilation corridor planning system based on Beidou high-precision positioning provided by the embodiment of the present application is shown in FIG. 1, which includes: Figure 4

[0167] ​The acquisition module 41 is configured to acquire the ground surface temperature distribution, building shape parameters and three-dimensional point cloud data of the urban ventilation corridor candidate area, filter the three-dimensional point cloud data, and construct a curved surface model based on the filtered three-dimensional point cloud data.

[0168] The optimization module 42 is configured to acquire the ground surface roughness distribution from the curved surface model, and align the ground surface roughness distribution, the ground surface temperature distribution and the building shape parameters in space coordinates based on a unified space coordinate reference provided by the Beidou high-precision positioning technology.

[0169] The recognition module 43 is configured to process the ground surface temperature distribution and the building shape parameters after the space coordinate alignment by using a deep learning model, so as to identify the ventilation weak area and the cold source inlet.

[0170] The generation module 44 is configured to generate a planning path of the urban ventilation corridor avoiding the ventilation weak area based on a path planning algorithm, and taking the minimization of the roughness corresponding to the ground surface roughness distribution and the maximization of the cold source connectivity corresponding to the cold source inlet as the target.

[0171] The urban ventilation corridor planning system based on the Beidou high-precision positioning provided in the embodiments of the present application is used to implement the urban ventilation corridor planning method based on the Beidou high-precision positioning described above, and therefore the specific implementation of the urban ventilation corridor planning system based on the Beidou high-precision positioning can be seen from the embodiment part of the urban ventilation corridor planning method based on the Beidou high-precision positioning described above, and the specific implementation can be referred to the description of the corresponding embodiment part, which will not be repeated here.

[0172] The present application also provides an electronic device, comprising: a memory for storing a computer program; a processor for executing the computer program to implement the steps of the urban ventilation corridor planning method based on the Beidou high-precision positioning described above.

[0173] The present application also provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the steps of the urban ventilation corridor planning method based on the Beidou high-precision positioning described above.

[0174] In an exemplary embodiment, the computer readable storage medium described above can include but is not limited to: a U disk, a read-only memory, a random access memory, a mobile hard disk, a magnetic disk or an optical disk, and various media that can store computer programs.

[0175] The embodiments of the present application also provide a computer program product, which comprises a computer program, and the computer program is executed by a processor to implement the steps in the urban ventilation corridor planning method based on the Beidou high-precision positioning described above.

[0176] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design goals of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0177] The foregoing has provided a detailed description of the urban ventilation corridor planning method, system, electronic equipment, and storage medium based on BeiDou high-precision positioning provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A method for planning urban ventilation corridors based on BeiDou high-precision positioning, characterized in that, include: The surface temperature distribution, building morphology parameters and three-dimensional point cloud data of the candidate area of ​​the urban ventilation corridor are obtained, the three-dimensional point cloud data is filtered, and a surface model is constructed based on the filtered three-dimensional point cloud data. The surface roughness distribution is obtained from the surface model, and the surface roughness distribution, the surface temperature distribution, and the building morphology parameters are aligned in spatial coordinates using the unified spatial coordinate reference provided by Beidou high-precision positioning technology. A deep learning model is used to process the surface temperature distribution and building morphology parameters after spatial coordinate alignment in order to identify weak ventilation areas and cold source inlets. Based on the path planning algorithm, with the objectives of minimizing the roughness corresponding to the surface roughness distribution and maximizing the cold source connectivity corresponding to the cold source inlet, a planned path for urban ventilation corridors that avoids the weak ventilation zone is generated.

2. The urban ventilation corridor planning method based on BeiDou high-precision positioning according to claim 1, characterized in that, The process of using a deep learning model to process the spatially aligned surface temperature distribution and building morphology parameters to identify weak ventilation zones and cold source inlets includes: According to preset temperature ranges, the surface temperature distribution after spatial coordinate alignment is divided into multiple temperature regions, and the temperature value and temperature change of each temperature region are obtained. The building morphology parameters after spatial coordinate alignment are divided into morphology regions corresponding to the temperature region, and the building height, number of buildings and size of the gap between buildings are extracted in each morphology region. The temperature value, the temperature change, the building height, the number of buildings, and the size of the gap between buildings are combined to obtain a data set; The dataset is analyzed using a deep learning model, and the spatial range of the weak ventilation zone and the spatial location of the cold source inlet are determined based on the analysis results.

3. The urban ventilation corridor planning method based on BeiDou high-precision positioning according to claim 2, characterized in that, The process of analyzing the dataset using a deep learning model and determining the spatial range of the weak ventilation zone and the spatial location of the cold source inlet based on the analysis results includes: The temperature values ​​in the dataset are correlated with the building height to obtain a first correlation analysis result. The temperature changes, the number of buildings, and the size of the gaps between buildings are correlated to obtain a second correlation analysis result. Based on the first correlation analysis results, a deep learning model is used to determine whether the corresponding temperature region and the corresponding morphological region meet the preset poor ventilation characteristics. Based on the results of the second correlation analysis, a deep learning model is used to determine whether the corresponding temperature region and the corresponding morphological region conform to the preset cold source entry and exit characteristics. Combine all temperature zones and corresponding morphological zones that meet the preset characteristics of poor ventilation to generate the spatial range of weak ventilation zones. Mark the intersection of the edges of all temperature regions that meet the preset cold source entry and exit characteristics with the corresponding morphological regions as the spatial location of the cold source inlet.

4. The urban ventilation corridor planning method based on BeiDou high-precision positioning according to claim 1, characterized in that, The path planning algorithm aims to minimize the surface roughness corresponding to the surface roughness distribution and maximize the cold source connectivity corresponding to the cold source inlet. Based on the objectives, it generates planned paths for urban ventilation corridors that avoid the weak ventilation zone, including: The starting point and the target point of the urban ventilation corridor are selected, and the starting point and the target point correspond to different cold source inlets; Based on the surface roughness distribution, determine the roughness value corresponding to each continuous region in the surface model; Define the spatial range of the poorly ventilated area as an area that cannot be traversed by any path; In the surface model, multiple first alternative paths from the starting point to the target point are searched, and the goal is to minimize the sum of the roughness values ​​of the regions traversed by each first alternative path and maximize the number of connections between each first alternative path and the cold source inlet. Based on the stated objectives, a first alternative route that avoids areas that the stated path cannot traverse is selected as the planned route for the urban ventilation corridor.

5. The urban ventilation corridor planning method based on BeiDou high-precision positioning according to claim 4, characterized in that, The step of selecting a first alternative route that avoids areas inaccessible by the original route as the planning route for the urban ventilation corridor, based on the stated objective, includes: Select the first alternative path from all the first alternative paths that avoids the area that the path cannot pass through, and use it as the second alternative path; Each second alternative path is divided into multiple consecutive path segments, and the roughness value corresponding to each path segment is calculated. All the roughness values ​​that make up the second alternative path are summed to obtain the total roughness value corresponding to each second alternative path. The number of cold source inlets passed through by each second alternative path is counted, and this number is taken as the number of connections corresponding to each second alternative path. Based on the total roughness value and number of connections corresponding to each second alternative path, the target value of each second alternative path is calculated according to the weight of each objective, and the second alternative path with the largest target value is selected as the planning path of the urban ventilation corridor.

6. The urban ventilation corridor planning method based on BeiDou high-precision positioning according to claim 1, characterized in that, The step of filtering the 3D point cloud data and constructing a surface model based on the filtered 3D point cloud data includes: Based on the coordinates of each point cloud data in the three-dimensional point cloud data, points within a preset height range in the candidate area of ​​the urban ventilation corridor are selected as points to be processed. Calculate the distance between each point to be processed and a preset number of neighboring points, and remove points whose distance exceeds the preset value to complete the filtering process; The filtered 3D point cloud data are connected sequentially according to the order of their corresponding coordinates to form multiple interconnected polygonal surfaces. A surface model is then constructed based on these polygonal surfaces.

7. The urban ventilation corridor planning method based on BeiDou high-precision positioning according to claim 1, characterized in that, The acquisition of surface temperature distribution, building morphology parameters, and 3D point cloud data of candidate areas for urban ventilation corridors includes: Acquire satellite thermal and drone images of candidate areas for urban ventilation corridors; Radiometric calibration is performed on the satellite thermal image and the UAV image respectively to obtain thermal radiation data and reflectance data. The thermal radiation data and the reflectance data are then fused in a multispectral manner to extract the surface temperature distribution and building morphology parameters. The candidate urban corridor area is scanned in three dimensions using a drone-based lidar system to obtain three-dimensional point cloud data.

8. A city ventilation corridor planning system based on BeiDou high-precision positioning, characterized in that, include: The acquisition module is used to acquire the surface temperature distribution, building morphology parameters and three-dimensional point cloud data of the candidate area of ​​urban ventilation corridor, filter the three-dimensional point cloud data, and construct a surface model based on the filtered three-dimensional point cloud data. The optimization module is used to obtain the surface roughness distribution from the surface model and align the surface roughness distribution, the surface temperature distribution, and the building morphology parameters with spatial coordinates using a unified spatial coordinate reference provided by BeiDou high-precision positioning technology. The identification module is used to process the surface temperature distribution and building morphology parameters after spatial coordinate alignment using a deep learning model in order to identify weak ventilation areas and cold source inlets. The generation module is used to generate a planned path for urban ventilation corridors that avoid the weak ventilation zone, based on a path planning algorithm, with the goal of minimizing the roughness corresponding to the surface roughness distribution and maximizing the cold source connectivity corresponding to the cold source inlet.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the steps of the urban ventilation corridor planning method based on BeiDou high-precision positioning as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, enables the urban ventilation corridor planning method based on BeiDou high-precision positioning as described in any one of claims 1 to 7.

Images