An AI-based method for automatically generating grid-like urban road networks

Abstract

This invention discloses an artificial intelligence-based automatic generation method for grid-type urban road networks, belonging to the field of urban planning; it includes: S1 collecting relevant base data to obtain three layer sets: target base boundary, main road centerline, and urban land use category; S2 merging the target base boundary and main road centerline to generate a plot layer, overlaying urban land use category, deleting land use with non-road network coverage attributes from the plot layer, and then obtaining a street area database based on distance settings and the advancing wavefront method; S3 randomly placing external control points at the plot boundaries, randomly placing uniform grids inside the plots and adapting to site rotation, offsetting the external and internal control points to the nearest intersection of the street area, obtaining a control point sequence set; S4 sorting the external and internal control point sequence sets in ascending order, searching for the street area line segment with the smallest bend angle between each pair of control points to obtain the main road network; S5 three-dimensional interactive projection of the road scheme to obtain engineering report drawings.

Description

Technical Field

[0001] This invention belongs to the field of urban planning, and specifically relates to an artificial intelligence-based method for automatically generating grid-type urban road networks. Background Technology

[0002] In urban road planning and design, grid-style urban roads are one of the most common road network types, often used in cities with flat terrain. Traditional grid-style urban road planning mainly relies on manpower combined with existing standards for planning and design. It is difficult to fully comply with national standards and specifications for urban roads during the planning and design process. It is relatively dependent on the work experience of the planners and designers, and the results are not scientific enough. Moreover, it is even more inadequate when facing complex natural terrain, making the work quite challenging.

[0003] Existing methods for generating grid-based urban road networks fall into two categories. One involves computer graphics modeling, which automatically generates road network models based on predefined specifications. While this method meets the criteria, the algorithm is mechanical and crude. Evolutionary algorithms are then used to optimize this process, but the roads generated by these algorithms lack internal logic, resulting in chaotic and impractical solutions. The other method is based on image learning. This involves training a generative adversarial network (GAN) model by learning patterns from massive amounts of road network samples. This model generates road networks within strictly defined plot sizes, but it is a time-consuming pure image learning method. This approach suffers from slow model training, a lack of automated verification, and high costs associated with manually selecting feasible road networks. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to provide an artificial intelligence-based automatic generation method for grid-type urban road networks.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A method for automatically generating a grid-like urban road network using artificial intelligence includes the following steps:

[0007] S1. Collect relevant base data and input it into the geographic information system platform to obtain three layer sets: target base boundary, main road centerline, and urban land use category.

[0008] S2. Merge the target base boundary and the center line of the main road to generate a land parcel layer, overlay the urban land use category to assign land parcel attributes, delete land use with non-road network coverage attributes in the land parcel layer, and then obtain the street area database based on distance settings and the advancing wavefront method.

[0009] S3. Randomly place external control points at the boundary of the plot, and randomly place a uniform grid inside the plot and adapt it to the site rotation. The external control points and internal control points are offset to the intersection of the nearest block field to obtain the control point sequence set.

[0010] S4. Sort the external control point and internal control point sequence sets in ascending order, and search for the block field line segments with the smallest bend angle between each pair of control points to obtain the trunk road network.

[0011] S5, 3D interactive projection road scheme, and laser printing of the selected scheme to finally obtain the engineering report drawings of the road scheme.

[0012] Furthermore, the relevant base data includes: target base boundary data, main road data within the target base boundary, and urban land use data.

[0013] Furthermore, the target base boundary data and the main road data within the target base boundary are collected using drone equipment; the urban land use data within the target base boundary are collected using an urban open-source data platform.

[0014] Furthermore, the major urban land use categories include: Category A public management and public service land, Category B commercial service facility land, Category R residential land, Category M industrial land, Category G green space and square land, and Category E non-construction land.

[0015] Furthermore, the land use with non-road network coverage attributes includes: Class A public management and service facilities land, Class G green space and square land, and Class E non-construction land.

[0016] Furthermore, the advancing wavefront method involves advancing the boundary of the site layer that can generate road network from the edge of the site towards the center of the site at a set distance, randomly and automatically generating a grid to obtain the street area.

[0017] Furthermore, the step of obtaining the control point sequence set in S3 is as follows:

[0018] S31, randomly place external control points with a certain spacing between any two endpoints of the boundary of the road network site layer that can be generated, take the largest bounding rectangle of plots N1, N2, N3...Nn and calculate the spatial deflection angles θ1, θ2, θ3...θn;

[0019] S32, generate a standard uniform grid within the block plot according to the recommended block scale values ​​for different functional areas in the city, rotate the grid by an angle θn and filter out the block boundary grid to obtain internal control points;

[0020] S33, offset the external control point and the internal control point to the nearest intersection of the street field obtained in S2;

[0021] S34. Based on the latitude and longitude coordinates, obtain the sequence of external control points and internal control points in sequence, and obtain the control point sequence set (X1, Y1), (X2, Y2), (X3, Y3)...(Xn, Yn).

[0022] Furthermore, the step of obtaining the trunk road network in S4 is as follows:

[0023] S41, based on the control point sequence set described in S3, the block field line segments are searched according to the principle of minimizing the angle between control point (Xn-1, Yn-1) and control point (Xn, Yn). The principle of minimizing the angle means that there are N ways to connect the block field between two control points. The sum of the angles between the line segments of the block field in each connection method is calculated, and the block field connection method with the smallest sum of angles is selected.

[0024] S42, following the ascending order of the X coordinates and Y coordinates of the control points, generate the transverse and longitudinal arterial roads within the plot to obtain the arterial road network.

[0025] The beneficial effects of this invention are:

[0026] 1. This method combines urban road regulations with artificial intelligence technology, which can generate road network plans simultaneously in a short time, reducing labor costs and improving design efficiency;

[0027] 2. This method is based on the latest urban road planning and design specifications. By setting the block area distance for each type of land use attribute, the standardization of the generated road network can be guaranteed.

[0028] 3. This method uses the wavefront advancement method and the uniform grid method, combined with artificial intelligence algorithms, to generate urban grid-like road networks, which improves the standardization of road networks at all levels and ensures the scientific nature of automatically generated scheme sets;

[0029] 4. The results generated by this method are displayed on a 3D platform, allowing for interactive manual addition or subtraction of roads or changes in alignment via joysticks. This facilitates real-time adjustments to the plan and ultimately outputs road planning and design engineering reports and drawings. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 This is a flowchart of the generation method of the present invention;

[0032] Figure 2This is a schematic diagram of the initial data entry process for this invention;

[0033] Figure 3 This is a schematic diagram illustrating the generation of the street-area dataset in this invention;

[0034] Figure 4 This is a schematic diagram of the trunk road network generation process of the present invention;

[0035] Figure 5 These are road traffic planning engineering drawings generated by this invention. Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] like Figure 1 As shown, an artificial intelligence-based method for automatically generating a grid-like urban road network includes the following steps:

[0038] S1: Data acquisition and input;

[0039] like Figure 2 As shown, relevant base data is collected and input into the geographic information platform. The format of the relevant base data is uniformly converted into Shapefile using the format conversion tool in the geographic information platform. The coordinates of the relevant base data are uniformly converted into WorldGeodetic System-1984 using the projection tool in the geographic information platform, thus obtaining three layer sets: target base boundary, main road centerline, and urban land use category.

[0040] The relevant base data includes target base boundary data, main road data within the target base boundary, and urban land use data. Among them, the target base boundary data and the main road data within the target base boundary are collected using a P580 multi-rotor UAV, equipped with a camera with a resolution of 20 megapixels or higher, a 1 / 2.3-inch CMOS sensor, a focal length of 14.66mm, and a storage capacity of 256GB or more. The urban land use data within the target base boundary is collected through an urban open-source data platform. The urban land use categories include: Category A public management and public service land, Category B commercial service facilities land, Category R residential land, Category M industrial land, Category G green space and square land, and Category E non-construction land.

[0041] S2: Street area generation;

[0042] Merge the target site boundary and main road centerline layers obtained from S1 to generate a plot layer. Overlay the urban land use category to assign land use attributes to the plot layer, and obtain a plot layer with urban land use category attributes. According to the actual intention requirements of urban planning and design, delete the land use that does not have road network coverage attributes in the plot layer to obtain a site layer in which road network can be generated within the target site boundary. Number the plots in the layer as N1, N2, N3...Nn.

[0043] According to the "GBT 51328-2018 Urban Comprehensive Transportation System Planning Standard", the street area distance for each type of land use is set as follows: the street area distance for Class B commercial service facilities land is 100-200m; the street area distance for Class R residential land is no more than 300m; and the street area distance for Class M industrial land is no more than 600m.

[0044] like Figure 3 As shown, the block field is generated based on the set block field distance for each type of land use attribute and the advancing wavefront method. The block field intersection is generated by the feature inflection point conversion tool in the geographic information platform to obtain the block field dataset.

[0045] The aforementioned advancing wavefront method involves advancing the boundary of the site layer that can generate road network from the edge of the site towards the center of the site at a set distance, randomly and automatically generating a grid to obtain the street area.

[0046] S3: Road control point placement and optimization; specific steps are as follows:

[0047] S31, randomly place external control points at a certain interval between any two endpoints of the boundary of the road network site layer that can be generated, take the largest bounding rectangle of the plots N1, N2, N3...Nn and calculate the spatial deflection angles θ1, θ2, θ3...θn, the interval range is a specified setting of 400-600 meters;

[0048] S32, within the block plot, according to the recommended values ​​of block scale for different functional areas in the city in the "GBT51328-2018 Urban Comprehensive Transportation System Planning Standard", a standard uniform grid is generated, the grid is rotated by an angle θn and the plot boundary grid is selected to obtain internal control points;

[0049] S33, offset the external control point and the internal control point to the nearest intersection of the street field obtained in S2;

[0050] S34. Based on the latitude and longitude coordinates of the World Geodetic System-1984, obtain the sequence of external control points and internal control points in sequence, and obtain the control point sequence set (X1, Y1), (X2, Y2), (X3, Y3)...(Xn, Yn);

[0051] S4: Road scheme generation; the specific steps are as follows:

[0052] S41, based on the control point sequence set described in S3, the block field line segments are searched according to the principle of minimizing the angle between control point (Xn-1, Yn-1) and control point (Xn, Yn). The principle of minimizing the angle means that there are N ways to connect the block field between two control points. The sum of the angles between the line segments of the block field in each connection method is calculated, and the block field connection method with the smallest sum of angles is selected.

[0053] S42, as Figure 4 As shown, by sequentially ascending the X coordinates and Y coordinates of the control points, the transverse and longitudinal arterial roads within the plot are generated, thus obtaining the arterial road network.

[0054] According to the requirements for the road red line width of secondary arterial roads without special requirements in the "GBT51328-2018 Urban Comprehensive Transportation System Planning Standard", the secondary arterial roads should be widened to form a road red line of 20-35m, and a road red line chamfer should be generated at the intersection. The chamfer distance at the intersection of secondary arterial roads should be 15-20m.

[0055] S5: Road plan output;

[0056] The plan is loaded onto the 3D display platform, and roads can be manually added or removed or the alignment can be changed using a joystick, with the indicators changing in real time.

[0057] like Figure 5 As shown, the output generates trunk road schemes and index evaluation parameters, and forms engineering report drawings, which are then printed into text format using a C3222L / 3226 wireless A3 multifunction printer and color laser printer.

[0058] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0059] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention.

Claims

1. A method for automatically generating a grid-like urban road network using artificial intelligence, characterized in that, Includes the following steps: S1. Collect relevant base data and input it into the geographic information system platform to obtain three layer sets: target base boundary, main road centerline, and urban land use category. S2. Merge the target base boundary and the center line of the main road to generate a land parcel layer, overlay the urban land use category to assign land parcel attributes, delete land use with non-road network coverage attributes in the land parcel layer, and then obtain the street area database based on distance settings and the advancing wavefront method. S3. Randomly place external control points at the boundary of the plot, and randomly place a uniform grid inside the plot and adapt it to the site rotation. The external control points and internal control points are offset to the intersection of the nearest block field to obtain the control point sequence set. S4. Sort the external control point and internal control point sequence sets in ascending order, and search for the block field line segments with the smallest bend angle between each pair of control points to obtain the trunk road network. S5, 3D interactive projection road scheme, and laser printing of the selected scheme to finally obtain the engineering report drawings of the road scheme; The steps for obtaining the control point sequence set in S3 are as follows: S31, randomly place external control points with a certain spacing between any two endpoints of the boundary of the site layer that can generate road network, take the largest bounding rectangle of plots N1, N2, N3...Nn and calculate the spatial deflection angles θ1, θ2, θ3...θn; S32, generate a standard uniform grid within the block plot according to the recommended block scale values ​​for different functional areas in the city, rotate the grid by an angle θn and filter out the block boundary grid to obtain internal control points; S33, offset the external control point and the internal control point to the nearest intersection of the street field obtained in S2; S34, obtain the sequence of external control points and internal control points according to latitude and longitude coordinates, and obtain the control point sequence set (X1, Y1), (X2, Y2), (X3, Y3)...(Xn, Yn); The steps for obtaining the trunk road network in S4 are as follows: S41, based on the control point sequence set described in S3, the block field line segments are searched according to the principle of minimizing the angle between control point (Xn-1, Yn-1) and control point (Xn, Yn). The principle of minimizing the angle means that there are N ways to connect the block field between two control points. The sum of the angles between the line segments of the block field in each connection method is calculated, and the block field connection method with the smallest sum of angles is selected. S42, following the ascending order of the X coordinates and Y coordinates of the control points, generate the transverse and longitudinal arterial roads within the plot to obtain the arterial road network.

2. The method for automatically generating a grid-like urban road network using artificial intelligence according to claim 1, characterized in that, The relevant base data includes: target base boundary data, main road data within the target base boundary, and urban land use data.

3. The method for automatically generating a grid-like urban road network using artificial intelligence according to claim 2, characterized in that, The target base boundary data and the main road data within the target base boundary are collected by drone equipment; the urban land use data within the target base boundary are collected by an urban open source data platform.

4. The method for automatically generating a grid-like urban road network using artificial intelligence according to claim 2, characterized in that, The major land use categories in the city include: Category A public management and public service land, Category B commercial service facilities land, Category R residential land, Category M industrial land, Category G green space and square land, and Category E non-construction land.

5. The method for automatically generating a grid-like urban road network using artificial intelligence according to claim 1, characterized in that, The land use that is not covered by the road network includes: Class A public management and service facilities land, Class G green space and square land, and Class E non-construction land.

6. The method for automatically generating a grid-like urban road network using artificial intelligence according to claim 1, characterized in that, The aforementioned advancing wavefront method involves advancing the boundary of the site layer that can generate road network from the edge of the plot towards the center of the plot at a set distance, randomly and automatically generating a grid to obtain the street area.

Images