A Smart Generation Method for Urban Road Networks Based on Parameter Rules and Iterative Step-by-Step Implementation

By using a step-by-step iterative method based on parameter rules to generate urban road networks, the problem of insufficient factor balance in traditional methods is solved, and a more scientific, flexible and interactive road network generation is achieved.

CN117668982BActive Publication Date: 2026-11-17SOUTHEAST UNIV
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202311653689.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-12-05
Publication Date
2026-11-17
Estimated Expiration
2043-12-05

AI Technical Summary

Technical Problem

Traditional methods for generating urban road networks struggle to balance the influence of various factors, such as the natural geographical environment, planning intentions, and site conditions. They lack scientific rationality and cannot be dynamically and interactively adjusted and improved.

Method used

A step-by-step iterative approach based on parameter rules is adopted. By collecting vector data of the design site, a digital sand table is constructed, which is then classified into a set of dominant forces—points, lines, and surfaces. A parameter rule library is built, and the main road network is generated using the tensor field algorithm. The secondary road network is generated by combining topology extraction and spline fitting models, and interactive optimization is performed.

Benefits of technology

It improves the accuracy, scientific rigor, flexibility, and completeness of urban road network generation, enhances the interactivity and presentation of generated solutions, and ensures data standardization and accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117668982B_ABST
    Figure CN117668982B_ABST
Patent Text Reader

Abstract

This invention provides an intelligent method for generating urban road networks based on iterative parameter rules. The method includes: inputting site condition data; classifying the data into dominant force-point sets, dominant force-line sets, and dominant force-surface sets as different constraint inputs for tensor field generation, and pre-setting parameter rules according to relevant specifications; obtaining feature vectors of the tensor field and any point within the site using a tensor field algorithm based on the parameter rules and constraint inputs; generating arterial roads using a topology extraction algorithm based on the feature vectors, resulting in a multi-mode arterial road network; identifying and generating three types of control points based on the arterial road network; connecting the control points using a spline fitting model to form a connected secondary arterial road network; generating a branch road network based on a finite element mesh; and outputting the urban road network scheme for interactive optimization. This invention enables the optimization of urban road network structure through iterative parameter rule-based urban road network layout.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of urban planning technology, specifically to an intelligent method for generating urban road networks based on step-by-step iteration of parameter rules. Background Technology

[0002] Urban road network layout is a crucial component of urban planning and design, serving as a vital means of organizing spatial form and shaping the city's image, profoundly impacting urban transportation and development. A well-designed urban road network can significantly enhance urban accessibility, optimize urban functions and spatial organization, improve traffic efficiency and reduce congestion, and shape the city's image. Traditional methods for generating urban road networks struggle to balance the influence of various factors such as the natural geographical environment, planning intentions, and site conditions. Furthermore, they fail to meet the design needs of different levels of road networks, exhibiting excessive subjectivity and insufficient scientific rationality, and are unable to dynamically, intuitively, and effectively adjust and improve urban road network generation schemes. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides an intelligent urban road network generation method based on parameter rules and iterative progression. This method solves the problems of traditional urban road network generation, which struggles to balance the influence of multiple factors such as natural geographical environment, planning intent, and site conditions, lacks scientific rationality, and cannot dynamically and interactively adjust and improve the urban road network generation scheme.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A method for intelligently generating urban road networks based on step-by-step iteration of parameter rules, comprising the following steps:

[0006] Collect relevant vector data of design site conditions and input them into the geographic information platform for format standardization processing to construct a digital sand table of the design site and form a design site dataset.

[0007] The design site condition dataset is classified into dominant force-point set, dominant force-line set, and dominant force-surface set, which are used as inputs for different constraint conditions in the generation of tensor field. A parameter rule library is constructed based on the preset parameter rules in relevant specifications.

[0008] Based on the parameter rule base, the different input constraints are transformed into multiple tensor fields and fused using the tensor field algorithm to obtain a tensor field that changes continuously and smoothly between different constraints, and to generate feature vectors for any point within the design site; based on the feature vectors, the topology extraction algorithm is used to extract the main roads and obtain a multi-mode main road network.

[0009] The main roads generated based on tensor fields divide the site into enclosed areas, identify and generate three types of control points: connection points, external control points, and internal control points; select internal control points to connect to the external control points on the edge lines, and connect the control points based on spline fitting models to form a connected secondary road network.

[0010] Input the urban land use function of the blocks divided by the generated main and secondary roads, discretize the block boundaries, select the best generation node of the branch road grid and generate the control grid, and then gradually generate the global grid from the block boundary inward. After completion, the grid is adjusted.

[0011] Output a progressively generated urban road network, make local adjustments to improve the overall form, and interactively optimize the generated overall road network in conjunction with the entire urban design process.

[0012] Preferably, the process of collecting relevant design site vector data and inputting it into a geographic information platform for format standardization to construct a digital sand table of the design site and form a design site condition dataset specifically includes:

[0013] Relevant design site vector data is collected through the OpenStreetMap open data platform. The collected design site vector data is then input into the geographic information platform. The format of the above design site vector data is uniformly converted into Shapefile through a format conversion tool to obtain design site boundary data, design site slope data, and design site natural boundary data. The relevant design site vector data includes the geographic coordinate information, plot height information, plot shape information, and plot area information of the design site.

[0014] Based on the planning intent of the design site, point vector data containing geographic coordinate information is obtained through annotation and drawing on a geospatial information platform. This data is used to represent the nodes of the design site; line vector data is used to represent the urban spatial axes of the design site; and area vector data is used to represent the different urban functional zones of the design site, including commercial areas, residential areas, and industrial areas.

[0015] The above vector data is unified in terms of dimensions and coordinates and integrated into the Ulrich holographic sand table. A digital sand table of the design site is constructed through spatial alignment to obtain the design site condition dataset.

[0016] Preferably, the step of classifying the design site condition dataset into dominant force-point set, dominant force-line set, and dominant force-surface set as inputs for different constraint conditions in tensor field generation, and constructing a parameter rule library according to relevant standard preset parameter rules, specifically includes:

[0017] The design site condition dataset is classified and processed according to the different geometric morphology types of its elements to obtain the following sets: dominant force-point set, used to control the formation of a radial road network; dominant force-line set, used to guide and control the orientation of the road shape; and dominant force-surface set, used to assist in the control of the road shape so that the road shape fits the natural boundary of the design site. These are used as different constraint inputs for generating the tensor field of the main road network.

[0018] Based on the "Urban Road Traffic Planning and Design Code" and the "Urban Road Intersection Planning Code", a parameter rule base for generating road networks at all levels has been compiled and constructed.

[0019] Table 1: Road Network Generation Rule Parameter Control

[0020]

[0021]

[0022] Preferably, based on the parameter rule base, the different input constraints are transformed into multiple tensor fields and fused using a tensor field algorithm to obtain a tensor field that continuously and smoothly varies between different constraints, and a feature vector of any point within the design site is generated; based on the feature vector, a topology extraction algorithm is used to extract and generate main roads, resulting in a multi-mode main road network, specifically including:

[0023] Based on the different constraint parameter sets and parameter rule bases of the dominant force-point set, dominant force-line set, and dominant force-surface set, different modes of basic tensor fields defined on the field where the design site is located are generated respectively. The influence range and magnitude of the force of each basic tensor field are controlled according to the parameter rules. The generated basic tensor fields are fused by the decaying radial basis function to obtain the tensor field T(p) that changes continuously and smoothly among multiple constraints.

[0024] The formula for calculating the attenuation radial basis function is as follows:

[0025]

[0026] Where T(p) is the fused tensor field, T i (p) represents the fundamental tensor field corresponding to the constraint conditions, d is the decay constant, and p is a point on the entire field where the design site is located. i The closer to the constraint, the higher the weight. The larger the value, the greater the influence of the constraints; conversely, the greater the distance, the weaker the influence.

[0027] By dividing the entire field space of the design site into a two-dimensional network model, based on the generated tensor field, using the tensors of the four grid points of any grid in the two-dimensional network model, the tensor at any position inside the grid is calculated by bilinear interpolation, thereby obtaining the mutually perpendicular primary and secondary eigenvectors of the tensor at any position in the entire field of the design site.

[0028] By visualizing tensor fields and using a topological approach, topology is extracted along the primary and secondary eigenvectors of the tensor at any position within the field for the basic pattern of the tensor field. This yields two different sets of path lines, which are then further combined to form different patterns of the basic tensor field, including a multi-mode backbone network of grid, radial, and organic types.

[0029] Preferably, the different constraint parameter sets based on the input dominant force-point set, dominant force-line set, and dominant force-surface set respectively generate basic tensor fields of different modes defined on the entire field of the design site, specifically including:

[0030] A tensor is generated based on the given point element P0 = (x0, y0) in the input dominant force-point set. Its main eigenvectors are circular, and the secondary eigenvectors emanate from the center, thus obtaining the basic tensor field of the node pattern, which has the following definition;

[0031]

[0032] Where x = x p -x0, x = x p -x0, P=(x,y) is any point in the field;

[0033] The direction (u) is defined by the given line element in the input dominant force-line set. x ,u y The tensor is generated by generating a tensor whose primary and secondary eigenvectors are perpendicular to each other, thus obtaining the basic tensor field of the grid pattern, which has the following definition;

[0034]

[0035] in,

[0036] Based on the input dominant force-surface set, extract the boundary line from its surface features. For any line segment AB within the boundary line, assign a regular element at point A, making its principal feature vector... This generates a smooth tensor field that adapts to the boundary curve; and then obtains the basic tensor field domains of different modes generated under the constraints of the set of dominant forces.

[0037] Preferably, the step of dividing the site into enclosed areas and identifying and generating three types of control points—connection points, external control points, and internal control points—specifically includes:

[0038] The main road generated based on the tensor field divides the site into enclosed areas Bn;

[0039] Identify the connection points Pi(x,y) on the main road of the enclosed area, excluding the area's vertices, that connect to other roads. The latitude and longitude of these points are determined by the projected coordinate system in which they are located, where x represents longitude and y represents latitude.

[0040] Between two intersections on the main road of the enclosed area, points are randomly generated according to the secondary road spacing index (350-500m) in a clockwise generation order from the upper left area to the lower right area, and identified as external control points Pj(x,y). The latitude and longitude are determined by the projection coordinate system in which they are located, where x represents longitude and y represents latitude.

[0041] Identify the area of ​​the enclosed area Bn, calculate the number of intersections to be set according to the intersection density parameters, and randomly generate a rectangular dot matrix based on the calculated number. The dot matrix spacing is 350-500m. The generated points are identified as internal control Pn(x,y), and their latitude and longitude are determined by the projection coordinate system in which they are located, where x represents longitude and y represents latitude.

[0042] Rotate the generated rectangular dot matrix to the same direction based on the horizontal direction, and delete the internal control points located outside the area or on the edge of the area; delete the internal control points with a vertical distance of less than 350m from the edge of the area; add, adjust and delete internal control points as appropriate according to the shape control factors.

[0043] Preferably, internal control points are selected and connected to external control points along the edge lines. These control points are then connected based on a spline fitting model to form a connected secondary trunk road network. Specifically, this includes:

[0044] For each internal control point, select an external control point that meets the following three requirements: (1) It should be as close as possible to the target internal control point; (2) If the two points form a transverse road, select an external control point with the smallest possible latitude difference; if the two points form a longitudinal road, select an external control point with the smallest possible longitude difference; (3) The road should be as perpendicular as possible to the edge of the plot after the two points are connected.

[0045] The objective function for selecting external control points using goal programming is as follows:

[0046]

[0047] (z row and z colLet w represent the objective functions for selecting external points for the lateral and longitudinal paths, respectively, with the variables being the coordinates of the external control points; w represents the weights of each variable, and d represents the weights of the external control points. IO d represents the distance between the outermost internal control point and the target's external control point. y and d x These represent the latitude and longitude differences between the internal control point and the external control point of the target, respectively, and θ is the angle between the internal control point and the external control point of the target relative to the edge of the land parcel.

[0048] The selected internal and external control points are connected using a spline fitting model to generate a secondary trunk road network.

[0049] Preferably, the urban land use function of the blocks divided by the generated primary and secondary roads is input, the block boundaries are discretized, the optimal generation node of the branch road grid is selected and a control grid is generated, and then the global grid is gradually generated from the block boundary inward. After completion, the grid is adjusted to obtain the branch road network, specifically including:

[0050] Divide the blocks into blocks based on the existing main and secondary road network, and input urban land use functions for each block;

[0051] The block boundaries are discretized, and the discretization length depends on the functional type of the block, as shown in the table below:

[0052] Table 2: Boundary Discrete Length and Generated Mesh Segment Length Control

[0053] Street Function Discrete length / Length of generated grid segment (meters) Business District 150-200 Residential area 200-300 Industrial Zone 300-600

[0054] Measure the interior angles of the block boundary points to determine the optimal generation point P1 with the smallest included angle among all discrete line segments. Use discrete line segment P1P2 as the base edge. If the included angle θ is less than 110 degrees, take the existing boundary point P on the adjacent edge and take boundary point G so that PG1 is parallel to P1P2. The length of PG1 must meet the control of the generated grid segment length (see Table 2). If the included angle θ is greater than or equal to 110 degrees, take the boundary point G1 on the angle bisector of the included angle θ. The length of P1G1 must meet the control of the generated grid segment length. Then select boundary point G2 again. G1G2 needs to be parallel to P1P2.

[0055] The first control grid is generated based on the interior angle of the optimal generation point and the coordinates of the corresponding boundary points.

[0056] Based on the generated control grid, generation points Pn are inserted sequentially, and the global grid is generated iteratively from the block boundary inwards, and the grid is adjusted.

[0057] Preferably, the output urban road plan is interactively optimized, specifically including:

[0058] After evaluating and comparing different schemes and improving the overall shape of the road network, the generated road network is input into the smart city design platform to generate a visualized 3D model, which is then displayed on a holographic sand table in combination with the intelligently generated 3D model of the land parcels.

[0059] VR / AR technology is used throughout the entire urban design process to adjust the road network layout in real time, perform interactive optimization, and simultaneously display the adjusted road network structure and information of each road segment.

[0060] Beneficial effects:

[0061] (1) The present invention is a method for intelligent generation of urban road networks based on step-by-step iteration of parameter rules. By collecting relevant design site vector data and inputting it into a geographic information platform for format unification processing, a digital sand table of the design site is constructed to form a dataset of design site conditions, which ensures the standardization and accuracy of the data and greatly improves the accuracy and efficiency of urban road network generation.

[0062] (2) The present invention is a method for intelligent generation of urban road networks based on step-by-step iteration of parameter rules. By classifying the design site condition dataset into dominant force-point set, dominant force-line set, and dominant force-surface set as inputs for different constraints in the generation of tensor fields, and constructing a parameter rule library according to the preset parameter rules in relevant specifications, it can comprehensively consider the influence of various factors such as natural geographical environment, planning intention, site conditions and relevant specifications, and can incorporate road-related parameters, making the overall urban road network generation more scientific and reasonable.

[0063] (3) The present invention provides an intelligent urban road network generation method based on step-by-step iteration of parameter rules. According to the parameter rule library, the different input constraints are converted into multiple tensor fields and fused through tensor field algorithm to obtain a tensor field that changes continuously and smoothly between different constraints. The main road is generated by extracting from the tensor field. It can integrate multiple influencing factors to obtain a multi-mode main road network, which can fully improve the flexibility and accuracy of urban road network generation.

[0064] (4) The present invention is a method for intelligent generation of urban road networks based on step-by-step iteration of parameter rules. By inputting the urban land use function of the blocks divided by the generated main and secondary roads, the block boundaries are discretized, the best generation node of the branch road grid is selected and the control grid is generated, and then the whole-domain grid is gradually generated from the block boundary to the inside. After completion, the grid is adjusted. It can generate a branch road network with appropriate orientation and organic growth through multiple parameters, which greatly improves the integrity and organic nature of urban road network generation.

[0065] (5) The present invention is a method for intelligent generation of urban road networks based on step-by-step iteration of parameter rules. By inputting the generated road network into the intelligent city design platform, a visualized three-dimensional model is generated. Combined with the intelligently generated three-dimensional model of the land parcel, it is displayed on a holographic sand table. In the entire process of urban design, VR / AR technology is used to assist in adjusting the road network layout in real time and performing interactive optimization. The adjusted road network structure and information of each road segment are displayed synchronously, which greatly improves the interactivity and display of the urban road network generation scheme. Attached Figure Description

[0066] Figure 1 This is a flowchart of the method of the present invention;

[0067] Figure 2 This is the selection logic for the optimal generation node of the branch grid in the embodiments of the present invention. Detailed Implementation

[0068] The technical solutions in 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.

[0069] Example

[0070] like Figure 1 and Figure 2 As shown, the technical solution of the present invention will be described in detail below using urban design of a certain area of ​​a certain city as an example.

[0071] (1) Collect relevant design site vector data and input it into the geographic information platform for format standardization processing to construct a digital sand table of the design site and form a design site condition dataset, specifically including:

[0072] (1.1) Collect relevant design site vector data through the OpenStreetMap open data platform, including the geographic coordinate information, plot height information, plot shape information, and plot area information of the design site;

[0073] (1.2) Input the collected design site vector data into the geographic information platform, and use the format conversion tool to convert the above design site vector data into a unified Shapefile to obtain the design site boundary data, design site slope data and design site natural boundary data.

[0074] (1.3) Based on the planning intent of the design site, through the geospatial information platform, by annotation and drawing, point element vector data (used to represent the nodes of the design site), line element vector data (used to represent the urban spatial axis of the design site), and area element vector data (used to represent the different urban functional zones of the design site, including commercial area, residential area, and industrial area) containing geographic coordinate information are obtained respectively.

[0075] (1.4) The above vector data is unified in terms of dimensions and coordinates and integrated into the Ulrich holographic sand table. The design site digital sand table is constructed through spatial alignment to obtain the design site condition dataset.

[0076] (2) The design site condition dataset obtained in step one is classified into dominant force-point set, dominant force-line set, and dominant force-surface set, which are used as inputs for different constraint conditions in the generation of the tensor field. A parameter rule library is constructed based on the preset parameter rules in relevant specifications, specifically including:

[0077] (2.1) The design site condition dataset is classified and processed according to the different geometric morphology types of its elements, and used as different constraint inputs for generating the tensor field of the main road network. They are: dominant force-point set, used to control the formation of radial road network; dominant force-line set, used to guide and control the orientation of road shape; dominant force-surface set, used to assist in road shape control so that the road shape fits the natural boundary of the design site.

[0078] (2.2) Based on the "Urban Road Traffic Planning and Design Code" and the "Urban Road Intersection Planning Code", a parameter rule base for generating road networks at all levels was compiled and constructed:

[0079] Table 1: Road Network Generation Rule Parameter Control

[0080]

[0081]

[0082] (3) Based on the parameter rule base of step two, using the tensor field algorithm, and based on the different constraint parameter sets of the dominant force-point set, dominant force-line set, and dominant force-surface set input in step three, basic tensor fields of different modes defined on the entire field of the design site are generated respectively, and fused to obtain a tensor field that changes continuously and smoothly between different constraints, and feature vectors of any point in the design site are generated; using the topology extraction algorithm, based on the feature vectors, the main roads are generated by topology extraction to obtain a multi-mode main road network, specifically including:

[0083] (3.1) Based on the given point element P0 = (x0, y0) in the input dominant force-point set, a tensor is generated. Its main eigenvectors are circular, and the secondary eigenvectors emanate from the center, thus obtaining the basic tensor field of the node pattern, which has the following definition:

[0084]

[0085] Where x = x p -x0, x = x p -x0, P=(x,y) is any point in the field;

[0086] (3.2) The direction (u) defined by the given line element in the input dominant force-line set is given. x ,u y The generation of tensors, whose principal and secondary eigenvectors are mutually perpendicular, yields the fundamental tensor field of the grid pattern, defined as follows:

[0087]

[0088] in,

[0089] (3.3) Based on the input dominant force-surface set, extract the boundary line from its surface features. For any line segment AB in the boundary line, assign a regular element at point A, making its principal feature vector as follows: This generates a smooth tensor field that adapts to the boundary curve; and then obtains the basic tensor field domains of different modes generated under the constraints of the set of dominant forces.

[0090] (3.4) The influence range and magnitude of the force of each basic tensor field are controlled according to the parameter rules. The generated basic tensor fields are fused by the decaying radial basis function to obtain a tensor field domain T(p) that changes continuously and smoothly among various constraints.

[0091] The formula for calculating the attenuation radial basis function is as follows:

[0092]

[0093] Where T(p) is the fused tensor field, Ti(p) is the basic tensor field corresponding to the constraint, d is the decay constant, p is a point on the entire field where the design site is located, and pi is the position of the constraint; the closer to the constraint, the higher the weight. The larger the value, the greater the influence of the constraints; conversely, the greater the distance, the weaker the influence.

[0094] (3.5) By dividing the entire field space of the design site into a two-dimensional network model, based on the generated tensor field, using the tensors of the four grid points of any grid in the two-dimensional network model, the tensor at any position inside the grid is calculated by bilinear interpolation, and then the mutually perpendicular primary and secondary eigenvectors of the tensor at any position in the entire field of the design site are obtained.

[0095] (3.6) Through tensor field visualization, using topology, for the basic pattern of tensor field, extract the topology along the main eigenvector and secondary eigenvector of tensor at any position in the field to obtain two different sets of path lines, and further combine them into different patterns of the basic tensor field, including grid type, radial type and organic type multi-mode backbone network.

[0096] (4) Based on the main road network generated in step three, the site is divided into enclosed areas, and three types of control points are identified and generated: connection points, external control points, and internal control points. Specifically, these include:

[0097] (4.1) Identify the points on the main road of the enclosed area that connect with other roads, except for the vertex of the area, as connection points Pi(x,y). The latitude and longitude are determined by the projection coordinate system in which they are located, where x represents longitude and y represents latitude.

[0098] (4.2) Between two intersections on the main road of the enclosed area, points are randomly generated according to the secondary road spacing index (350-500m) in a clockwise generation order from the upper left area to the lower right area, and identified as external control points Pj(x,y). The latitude and longitude are determined by the projection coordinate system in which they are located, where x represents longitude and y represents latitude.

[0099] (4.3) Identify the area of ​​the enclosed area Bn, calculate the number of intersections to be set according to the intersection density parameters, and randomly generate a rectangular dot matrix according to the calculated number. The dot matrix spacing is 350-500m. The generated points are identified as internal control Pn(x,y), and their latitude and longitude are determined by the projection coordinate system in which they are located, where x represents longitude and y represents latitude.

[0100] (4.4) Rotate the generated rectangular dot matrix to the same direction based on the horizontal direction, and delete the internal control points located outside the area or on the edge of the area; delete the internal control points with a vertical distance of less than 350m from the edge of the area; add, adjust and delete internal control points as appropriate according to the morphological control factors.

[0101] (5) Based on the three types of control points identified in step four, select the internal control points and connect them to the external control points of the edge lines. Connect the control points based on the spline fitting model to form a connected secondary trunk road network, specifically including:

[0102] (5.1) For each internal control point, select an external control point that meets the following three requirements: It should be as close as possible to the target internal control point; if the connection between the two points forms a transverse road, select an external control point with the smallest possible latitude difference; if the connection between the two points forms a longitudinal road, select an external control point with the smallest possible longitude difference; the road connecting the two points should be as perpendicular as possible to the edge of the plot. The external control point selection is performed using a goal programming method, and the objective function is expressed as follows:

[0103]

[0104] (z row and z col Let w represent the objective functions for selecting external points for the lateral and longitudinal paths, respectively, with the variables being the coordinates of the external control points; w represents the weights of each variable, and d represents the weights of the external control points. IO d represents the distance between the outermost internal control point and the target's external control point. y and d x These represent the latitude and longitude differences between the internal control point and the external control point of the target, respectively, and θ is the angle between the internal control point and the external control point of the target relative to the edge of the land parcel.

[0105] (5.2) Connect the selected internal and external control points using a spline fitting model to generate the secondary trunk road network; let P S =(p0,p1,…,p q-1 () represents q control points in region Bn, sorted by longitude or latitude. Assume control point p... m and p m+1 The polynomial function of the fitted curve between (m=0,1,...,q-2) is as follows:

[0106]

[0107] in, ω=(ω0,ω1,…,ω β );ω i β represents the polynomial coefficients; β is the order of the polynomial function.

[0108] (6) Input the urban land use functions of the blocks divided by the generated main and secondary roads, discretize the block boundaries, select the best generation nodes for the branch road grid and generate the control grid, and then gradually generate the global grid from the block boundaries inward. After completion, perform grid adjustment to obtain the branch road network, which specifically includes:

[0109] (6.1) Divide the blocks into blocks based on the generated main and secondary road network, and input the urban land use function for each block;

[0110] (6.2) Discretize the block boundaries. The discretization length depends on the functional type of the block, as shown in the table below.

[0111] Table 2: Boundary Discrete Length and Generated Mesh Segment Length Control

[0112]

[0113]

[0114] (6.3) Measure the interior angles of the block boundary points and determine the optimal generation point P1 with the smallest included angle between all discrete line segments. Use discrete line segment P1P2 as the base edge. If the included angle θ is less than 110 degrees, take the existing boundary point P on the adjacent edge and take boundary point G so that PG1 is parallel to P1P2. The length of PG1 must meet the control of the length of the generated grid road segment (see Table 2). If the included angle θ is greater than or equal to 110 degrees, take the boundary point G1 on the angle bisector of the included angle θ. The length of P1G1 must meet the control of the length of the generated grid road segment. Then select boundary point G2 again. G1G2 needs to be parallel to P1P2.

[0115] (6.4) Generate the first control mesh based on the interior angles of the optimal generation point and the coordinates of the corresponding boundary points;

[0116] (6.5) Based on the generated control grid, the generation point Pn is inserted sequentially, and the global grid is generated iteratively from the block boundary to the inside. The grid is then adjusted to obtain the adjusted branch network.

[0117] (7) Output the city road plan and perform interactive optimization, specifically including:

[0118] (7.1) After the scheme evaluation and comparison and the overall improvement of the road network, the generated road network is input into the smart city design platform to generate a visualized three-dimensional model, which is then displayed on the holographic sand table in combination with the intelligently generated three-dimensional model of the land parcel.

[0119] (7.2) Utilize VR / AR technology throughout the entire urban design process to adjust the road network layout in real time, perform interactive optimization, and simultaneously display the adjusted road network structure and information of each road segment.

[0120] 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, 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.

[0121] 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 intelligently generating urban road networks based on step-by-step iteration of parameter rules, characterized in that: include: Collect relevant vector data of design site conditions and input them into the geographic information platform for format standardization processing to construct a digital sand table of the design site and form a design site dataset. The design site condition dataset is categorized into dominant force-point set, dominant force-line set, and dominant force-surface set, which serve as inputs for different constraints in tensor field generation. A parameter rule library is constructed based on preset parameter rules according to relevant specifications. Using the parameter rule library, a tensor field algorithm is employed to transform the different input constraints into multiple tensor fields and fuse them to obtain a tensor field that continuously and smoothly varies between different constraints, generating feature vectors for any point within the design site. A topology extraction algorithm is then used to generate main roads based on these feature vectors, resulting in a multi-mode main road network. The main roads generated from the tensor field divide the site into enclosed areas, identifying and generating three types of control points: connection points, external control points, and internal control points. Internal control points are selected and connected to external control points along the boundary lines. Control points are then connected using a spline fitting model to form a connected secondary arterial road network. The urban land use functions of the blocks divided by the generated primary and secondary arterial roads are input, and the block boundaries are discretized. The optimal generation nodes for the branch road grid are selected and control grids are generated. A global grid is then gradually generated from the block boundaries inwards. After completion, the grid is adjusted to obtain the branch road network. The progressively generated urban road network is output, with local adjustments to improve the overall form. The generated overall road network is then interactively optimized in conjunction with the entire urban design process. The site is divided into enclosed areas, and three types of control points are identified and generated: connection points, external control points, and internal control points. Specifically, this includes: dividing the site into enclosed areas Bn based on the main roads generated by the tensor field; identifying connection points Pi(x,y) on the main roads of the enclosed areas, excluding the area vertices, connecting to other roads, with latitude and longitude determined by the projected coordinate system, where x represents longitude and y represents latitude; between two intersections on the main roads of the enclosed areas, points are randomly generated within the areas in a clockwise order from the upper left to the lower right, based on the secondary road spacing index of 350-500m, and identified as external control points Pj(x,y), with latitude and longitude determined by the projected coordinate system, where x represents longitude and y represents latitude; identifying the area of ​​the enclosed area Bn, calculating the required number of intersections based on the intersection density parameter, and randomly generating a rectangular point matrix with a point spacing of 350-500m based on the calculated number, identifying the generated points as internal control points Pn(x,y). x, y), the latitude and longitude are determined by the projection coordinate system, where x represents longitude and y represents latitude; the generated rectangular point array is rotated to the same direction based on the horizontal direction, and internal control points located outside the area or on the edge of the area are deleted; internal control points with a vertical distance of less than 350m from the edge of the area are deleted; internal control points are appropriately added, adjusted and deleted according to the morphological control factors; external control points are selected to connect internal control points to the edge, and control points are connected based on spline fitting model to form a connected secondary trunk road network, specifically including: for each internal control point, external control points that meet the following three requirements are selected: (1) as close as possible to the target internal control point; (2) if the two points are connected to form a transverse road, then the external control point with the smallest latitude difference is selected, and if the two points are connected to form a longitudinal road, then the external control point with the smallest longitude difference is selected; (3) the road is as perpendicular as possible to the edge of the plot after the two points are connected; the external control points are selected using the target planning method, and the objective function is expressed as: z row and z col Let represent the objective functions for selecting external points for the lateral and longitudinal paths, respectively, with the variables being the coordinates of the external control points; w represents the weights of each variable, and d represents the weights of the external control points. IO d represents the distance between the outermost internal control point and the target's external control point. y and d x These represent the latitude and longitude differences between the internal control point and the target's external control point, respectively. Define the angle between the target internal control point and the external control point relative to the edge of the plot; use a spline fitting model to connect the selected internal and external control points to generate a secondary trunk road network.

2. The intelligent generation method for urban road networks based on step-by-step iteration of parameter rules according to claim 1, characterized in that, Relevant design site vector data is collected and input into a geographic information platform for format standardization processing to construct a digital sand table of the design site, forming a design site condition dataset. Specifically, this includes: collecting relevant design site vector data through the OpenStreetMap open data platform; inputting the collected design site vector data into the geographic information platform; using a format conversion tool to uniformly convert the above design site vector data into a Shapefile to obtain design site boundary data, design site slope data, and design site natural boundary data; the relevant design site vector data includes the geographic coordinate information, plot height information, plot shape information, and plot area information of the design site; according to the planning intent of the design site, through annotation and drawing on the geographic spatial information platform, point element vector data containing geographic coordinate information is obtained to represent the nodes of the design site; line element vector data is used to represent the urban spatial axis of the design site; and area element vector data is used to represent the different urban functional zones of the design site, including commercial areas, residential areas, and industrial areas; the above vector data is standardized in terms of dimensions and coordinates and integrated into the Youli holographic sand table; the digital sand table of the design site is constructed through spatial alignment to obtain the design site condition dataset.

3. The intelligent generation method for urban road networks based on step-by-step iteration of parameter rules according to claim 2, characterized in that, The design site condition dataset is categorized into dominant force-point set, dominant force-line set, and dominant force-surface set, which serve as different constraint inputs for tensor field generation. A parameter rule library is constructed based on preset parameter rules according to relevant specifications. Specifically, the design site condition dataset is categorized according to the different geometric shapes of its elements to obtain: a dominant force-point set, used to control the formation of a radial road network; a dominant force-line set, used to guide and control the orientation of road forms; and a dominant force-surface set, used to assist in road form control, ensuring that the road form conforms to the natural boundaries of the design site. These serve as different constraint inputs for generating the tensor field of the main road network. A parameter rule library for generating road networks at all levels is constructed based on the "Urban Road Traffic Planning and Design Code" and the "Urban Road Intersection Planning Code." 4. The intelligent generation method for urban road networks based on step-by-step iteration of parameter rules according to claim 3, characterized in that, Based on the parameter rule base, the different input constraints are transformed into multiple tensor fields and fused using the tensor field algorithm to obtain a tensor field that changes continuously and smoothly between different constraints, and to generate the feature vector of any point in the design site. Using a topology extraction algorithm, based on the feature vectors, a main road network is generated, resulting in a multi-mode main road network. Specifically, this includes: generating different constraint parameter sets and parameter rule bases based on the input dominant force-point set, dominant force-line set, and dominant force-surface set; generating basic tensor fields of different modes defined on the design site; controlling the influence range and force magnitude of each basic tensor field according to the parameter rules; and fusing the generated basic tensor fields using decaying radial basis functions to obtain a tensor field that continuously and smoothly varies among multiple constraints. The formula for calculating the attenuation radial basis function is as follows: in, For the fused tensor field, Let be the fundamental tensor field corresponding to the constraint conditions. It is the attenuation constant. For the design site, at a point within the entire site, The closer to the constraint, the higher the weight. The larger the value, the greater the influence of the constraints; conversely, the farther the distance, the weaker the influence. By dividing the entire field space of the design site into a two-dimensional network model, based on the generated tensor field, using the tensors of the four grid points of any grid in the two-dimensional network model, the tensor at any position inside the grid is calculated by bilinear interpolation, thereby obtaining the mutually perpendicular principal and secondary eigenvectors of the tensor at any position in the entire field of the design site. Through tensor field visualization, using a topological approach, for the basic pattern of the tensor field, the topology is extracted along the principal and secondary eigenvectors of the tensor at any position in the field, resulting in two different sets of path lines. These are further combined into different patterns of the fused basic tensor field, including a multi-mode trunk network of grid, radial, and organic types.

5. The intelligent generation method for urban road networks based on step-by-step iteration of parameter rules according to claim 4, characterized in that, Based on the different constraint parameter sets of the input dominant force-point set, dominant force-line set, and dominant force-surface set, basic tensor fields of different modes defined on the entire field of the design site are generated respectively. Specifically, this includes: according to the given point elements in the input dominant force-point set... Generate a tensor whose primary eigenvectors are circular and whose secondary eigenvectors emanate from the center, thus obtaining the basic tensor field of the node pattern, defined as follows: 𝑇0(𝑝)= in, , , Let u be any point within the field; the direction (u) is defined by the given line elements in the input dominant force-line union set. x , u y ), generating a tensor whose principal and secondary eigenvectors are mutually perpendicular, thus obtaining the basic tensor field of the grid pattern, which has the following definition: 𝑇1(𝑝)= in, , Based on the input dominant force-surface set, extract the boundary line from its surface features. For any line segment AB within the boundary line, assign a regular element at point A, making its principal feature vector... This generates a smooth tensor field that adapts to the boundary curve; and then obtains the basic tensor field domains of different modes generated under the constraints of the set of dominant forces.

6. The intelligent generation method for urban road networks based on step-by-step iteration of parameter rules according to claim 5, characterized in that, The process involves inputting the urban land use functions of the blocks divided by the existing primary and secondary arterial roads, discretizing the block boundaries, selecting the optimal generation nodes for the branch road grid, generating a control grid, and then gradually generating a global grid from the block boundaries inwards. After completion, the grid is adjusted to obtain the branch road network. Specifically, this includes: dividing blocks from the existing primary and secondary arterial road network; inputting the urban land use functions for each block; and discretizing the block boundaries, with the discretization length depending on the block's function type, as shown in the table below: Boundary Discrete Length and Generated Grid Segment Length Control. Measure the interior angles of the block boundary points to determine the optimal generation point P1 with the smallest included angle among all discrete line segments. Use discrete line segment P1P2 as the base edge. If the included angle θ is less than 110 degrees, take the existing boundary point P on the adjacent edge and take boundary point G such that PG1 is parallel to P1P2. The length of PG1 must meet the control of the generated grid segment length. If the included angle θ is greater than or equal to 110 degrees, take the boundary point G1 on the angle bisector of the included angle θ. The length of P1G1 must meet the control of the generated grid segment length. Then select boundary point G2 again, which must be parallel to P1P2. Generate the first control grid based on the interior angles of the optimal generation point and the coordinates of the corresponding boundary points. Based on the generated control grid, insert generation points Pn sequentially, iteratively generating the global grid from the block boundary inwards, and perform grid adjustments to obtain the adjusted branch network.

7. The intelligent generation method for urban road networks based on step-by-step iteration of parameter rules according to claim 6, characterized in that, Output urban road plans and perform interactive optimization, specifically including: after evaluating and comparing the plans and improving the overall shape of the road network, inputting the generated road network into the smart city design platform to generate a visualized 3D model, which is then displayed on a holographic sand table in combination with the intelligently generated 3D model of the land parcels; using VR / AR technology to assist in the entire urban design process, adjusting the road network layout in real time, performing interactive optimization, and synchronously displaying the adjusted road network structure and information of each road segment.

Citation Information

Patent Citations

  • Intelligent vehicle speed prompting driving system based on mobile phone

    CN103680177A