Construction method of computable road network

Abstract

The invention discloses a method for constructing a computable road network. Road data in the vector map are extracted, line string data or line segments obtained by splitting the line string data are used as connecting edges, geographic feature codes containing Geohash information are given to starting points and ending points of the connecting edges, and therefore a basic road network map is constructed. And extracting a middle vertex set of the basic road network diagram and an edge attribute set of middle vertex connection edges, and carrying out road network vertex optimization based on edge attributes to obtain an optimized computable road network. Geographic feature codes adopted by the method comprise spatial geographic information and have the characteristics of graph vertex coding, and the constructed road network graph can be used for graph calculation and verification and restoration of road connection relations. Furthermore, in combination with a representation point acquisition scheme provided by the invention, a road network map fusing facilities and equipment can be constructed, so that the application range of the computable road network is expanded. The computable road network constructed by the method is high in operation efficiency and suitable for different production environments.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent transportation technology, and specifically relates to a method for constructing a computable road network. Background Art

[0002] Computable road networks are built based on map data, and road elements and attributes can be used to calculate graphical road networks. With the rapid development of intelligent transportation technology, computable road networks are widely used in transportation business areas, including but not limited to map data management, path planning, traversal retrieval, centrality analysis, accessibility analysis, connectivity analysis, and regional coverage analysis.

[0003] There are usually some problems in constructing a computable road network based on vector map data (especially high-precision maps), such as: high-precision maps have high geometric accuracy, but when converting them into a computable road network, it is necessary to balance the relationship between geometric accuracy and graphical expression; map data focuses on expressing the accuracy of geographic location but is not good at expressing the connectivity of roads; how to use the geographic accuracy of high-precision maps to calculate and correct connectivity when constructing the map network so that the converted computable road network is fully connected is an issue that needs to be solved urgently; in addition, when using high-precision maps to convert computable road networks, directly graphical road networks often contain excessive vertices and edges, which affects computational efficiency, and it is necessary to consider optimizing them for easy computational use. Summary of the Invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a method for constructing a computable road network.

[0005] In order to achieve the above technical objectives, the present invention adopts the following technical solutions: A method for constructing a computable road network, comprising: Extracting road data from map data, including at least road reference line vector data; Using line string data in the vector data or line segment data obtained by splitting the line string data as connecting edges, constructing a connecting edge dataset, assigning attribute information of the original road to the corresponding connecting edges, using the starting and ending points of each connecting edge as vertices and assigning geographic feature codes to the vertices to construct a basic road network map; wherein the geographic feature codes include Geohash code information; Extract the set of intermediate vertices of the basic road network graph and the set of edge attributes of the edges connecting the intermediate vertices; Extracting a first set of edge attributes from the edge attribute set, in which the predecessor and successor edge attributes must remain consistent when merging edges; Extracting intermediate vertices whose predecessor edges and successor edges have the same attribute values in the first set from the intermediate vertex set to form a vertex set to be optimized; The predecessor edges and successor edges of the vertices in the set of vertices to be optimized are merged into new edges, the road network graph optimization is completed, and the computable road network is obtained.

[0006] In some embodiments of the present invention, the line string data of the road reference line is split into minimum straight line segments to obtain a minimum line segment data set, and the minimum line segment data is used as a connecting edge to construct a basic road network graph.

[0007] In some embodiments of the present invention, the geographic feature code comprises a Geohash code calculated based on vertex geometric coordinates, or a combination of a Geohash code and an elevation code; the elevation code is a code containing elevation information; When the geographic feature code adopts a combination of Geohash code and elevation code, the Geohash code and elevation information are sequentially concatenated to form a final code. Optionally, the elevation code is expressed in numerical characters and includes a sign bit representing the positive and negative values of the elevation.

[0008] In some embodiments of the present invention, the construction method further includes verifying the accuracy of the road connection relationship data using geographic feature coding, including: Obtain road connection relationship data, and calculate the geographic feature code value for the end point of the predecessor road and the starting point of the successor road for any connected road in the road connection relationship data. If the calculated geographic feature code value is exactly the same or the previous one in the Geohash code is the same, n If the positions are consistent, the roads are connected.

[0009] In some embodiments of the present invention, the accuracy verification also includes, when the connection relationship verification fails, making a correction judgment on the connection relationship. If the connection relationship needs to be corrected, a virtual road connection line is constructed using the geometric coordinates and geographic feature codes of the missing connection points and appended to the connection edge dataset.

[0010] In some embodiments of the present invention, the construction method further includes, after constructing the basic road network map, using the basic road network map to review road connection problems and correcting the road reference line vector data and the basic road network map, including: Calculate the in-degree and out-degree of the vertices of the basic road network graph; Filter the vertices to be checked based on the in-degree and out-degree characteristics of the graph vertices; Perform Geohash decoding on the vertex to be checked to obtain the vertex coordinate value; Display the vertices to be checked on the GIS map, analyze and judge them in combination with the original vector map data, and repair the vector map data with problems; The repaired basic road network map is regenerated using the repaired vector map data.

[0011] In some embodiments of the present invention, the construction method further includes, after constructing the basic road network map, embedding the facilities and equipment in the transportation system as vertices into the road network map to obtain a road network map that integrates the facilities and equipment information.

[0012] In some embodiments of the present invention, when optimizing a road network graph that integrates facility and equipment information, it is necessary to remove vertices embedded with facilities and equipment from the set of vertices to be optimized.

[0013] In some embodiments of the present invention, embedding the facilities and equipment in the transportation system as vertices in the road network graph includes: According to the geographical location relationship between the facilities and equipment and the road network, the vertices related to the facilities and equipment in the road network are obtained as the facility and equipment road network representation points, and the attribute information of the facilities and equipment is embedded in the corresponding road network representation points; or, According to the geographical location relationship between facilities and equipment and the road network, the road network representation point of the facilities and equipment is calculated, the attribute information of the facilities and equipment is bound to the representation point, and the representation point is inserted into the road network graph as a new vertex. A new connecting edge is constructed and the old edge is deleted after inheriting the edge attributes.

[0014] In some embodiments of the present invention, the method for obtaining or calculating the road network representation point includes: a. Calculate the coordinate point or geometric center point of the facility equipment and the vertex with the smallest distance from the vertex set of the basic road network graph as the candidate road network representation point; or b. Calculate the nearest edge between the coordinate point or geometric center point of the facility and the edge set of the underlying road network graph, and calculate the foot of the perpendicular to the edge as the candidate road network representation point; or c. extracting vertices contained in the facilities as candidate road network representation points; Or d. Based on the geometric shape of the facility, construct a buffer zone of a preset distance; use the buffer zone as the calculation range, and extract the vertices contained in the calculation range as candidate road network representation points; Or e. For regional facilities covered by a road network, use the vertices corresponding to all upstream and downstream junctions and confluences in the covered area as candidate road network representation points for the facility; or f. Based on the characteristics of the facilities and equipment and the application requirements, calculate candidate road network representation points that accurately express the spatial relationship between the facilities and equipment and the road network.

[0015] In some embodiments of the present invention, vertices in the vertex set to be optimized are deleted one by one in a recursive manner, and new edges are added to the graph during deletion, while the predecessor and successor edges of the deleted vertices are removed, thereby completing the road network graph optimization.

[0016] The present invention has the following beneficial effects: (1) The present invention uses vectorized maps to construct road networks, and can directly use coordinates or coordinate encoding to express and construct road network maps without relying on connection relationships. In order to solve the problem of large scale and high computational complexity of basic road network maps, and to take into account the scale, accuracy and computational performance of road network maps during engineering, the road network map optimization method proposed in the present invention can reduce the number of vertices and retain the reduced vertex information on the edge attributes. While maintaining the topological relationship, the road geographic information is retained, and the scale of the road network map is greatly reduced, the computational efficiency when using graph algorithms is improved, and the final optimized road network map is suitable for production environments with high response requirements.

[0017] (2) The present invention assigns geographic feature codes containing Geohash coding information to vertices. This code can be used as a vertex number in the construction and use of the road network map, and also carries the geographic spatial location information of the point. It can be decoded to obtain longitude, latitude and altitude, so that the road network map can be used in specific application scenarios. For example, when solving the straight-line distance and elevation difference between two vertices, the decoded vertex position is used to calculate it using a geographic geometry algorithm; in addition, the assigned code can also reconstruct the connection relationship based on the coordinates. It can be used to simultaneously express the actual geographic coordinate value and the vertex code of the calculable road network; the length of the Geohash is used to control the expression accuracy of the geographic location. At the same time, the proximity attribute of the code can be used to autonomously determine the connection relationship between (predecessor and successor) roads. The road connection relationship is verified based on the geographic feature code, and when the connection relationship verification fails, the constructed data set is corrected, including geographic coordinate distance calculation, attribute comparison and connection relationship completion. This process can effectively solve the common topological discontinuity problem in high-precision map data and ensure the connectivity and completeness of the road network.

[0018] (3) This invention innovatively proposes a geographic feature coding system that combines Geohash coding with elevation information. By concatenating Geohash values with elevation values to form a string, it breaks through the limitation of traditional Geohash that can only express planar positions and achieves accurate expression of road topological relationships in three-dimensional space. This coding method not only intuitively retains the geometric accuracy of high-precision maps, but also enables rapid graph vertex construction and edge connection relationship expression, facilitating subsequent road network calculations and geographic information restoration. For example, when using the road network, elevation data can also be used as needed, such as when calculating slopes and slope change points.

[0019] (4) This paper proposes a method for calculating multiple facility network representation points for facility embedding, solving the problem of representing facilities and equipment with different functions and providing a data foundation for analyzing the relationships between facilities. The vertex attributes of facility and equipment embedding use an extensible data structure, supporting the flexible expansion of the facility type set. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is the technical roadmap of the present invention.

[0021] Figure 2 This is the display effect of facilities embedded in the road network map superimposed on the GIS (partial); the yellow line in the figure is the original road vector data, and the red and green points are the diverging points and converging points of various types of connection points extracted from the basic road network map. Green is the diverging point, and red is the converging point. The numbers before and after "->" correspond to the in-degree and out-degree of the point; the orange and blue icons represent certain types of facilities, and the corresponding text is the facility number.

[0022] Figure 3 This is the display effect (partial) of the optimized facility embedded in the road network map superimposed in GIS; the black line is the optimized road network map; the text is the starting point geographic feature code.

[0023] Figure 4 The figure shows the result of the computable road network when applied (facility topology connectivity analysis); the purple line is the relationship expression of facility topology connectivity. DETAILED DESCRIPTION

[0024] The technical solution of the present invention is further described below in conjunction with specific embodiments.

[0025] The terms involved in this embodiment are explained as follows:

[0026] High-precision maps are high-fidelity maps with an absolute accuracy of less than 100 cm and a relative accuracy of less than 10 cm. They are collected and produced by qualified surveying and mapping companies. High-precision road maps are thematic maps for the transportation sector, typically containing a variety of information, including multi-level roads and lanes (lane groups), bridges, tunnels, facilities and equipment, signs and markings, surrounding greenery, terrain, and water bodies. Due to these characteristics, high-precision maps are primarily used in scenarios such as autonomous driving and simulation.

[0027] A road network graph is a graph object or model used to describe road connectivity. Directed graphs are often used to describe road attributes and connectivity. Road network graphs can be used with graph algorithms to solve various transportation problems, such as path planning, accessibility analysis, maximum flow, and site selection.

[0028] Geohash is a geographic coordinate encoding method that uses a string to represent geographic coordinates (latitude and longitude). The length of the string determines the accuracy of the location. For example, 8-bit encoding corresponds to 19-meter accuracy, 10-bit encoding corresponds to 1.2m accuracy, 12-bit encoding can reach approx. 3.7 cm accuracy. This encoding method facilitates the expression and storage of geographic location information. It also enables neighborhood searches with a given accuracy through prefix matching, and can be used for geospatial indexing and efficient queries.

[0029] Unless otherwise specified, some steps in the embodiments can be implemented in existing ways.

[0030] Example 1

[0031] The present invention mainly includes the following processes:

[0032] A) Construction of a basic road network map; B) Embedding transportation facilities and equipment into the road network map (optional); C) Optimization and simplification of the road network map.

[0033] A) Construction of basic road network map

[0034] A1 map data preprocessing

[0035] A1-1 Load and extract road-related data from original map data, such as roads (road reference line vector data), lane groups, and road connection relationships.

[0036] A1-2 pre-processes the data, including quality inspection and cleaning, data conversion and standardization.

[0037] The original map data may be any map data, and high-precision map data may be used.

[0038] Preparation for A2 road network construction

[0039] Calculation of geographic feature coding values for the starting and ending points of A2-1 road lines

[0040] The present invention proposes a combined coding based on Geohash, which adds elevation information to the original coding and is used to express geographic information on road network maps. The construction process of the combined coding includes:

[0041] 1) Set the bit length of Geohash according to the required accuracy;

[0042] 2) Set the bit length of the elevation code according to the accuracy requirement, and express it in numerical characters. , For floating point precision, when the elevation value is negative, The first character of is recorded as "-", and the total length remains unchanged;

[0043] 3) Concatenate the Geohash and elevation codes to form the final code.

[0044] Taking the current common high-precision map as an example, the Geohash value can be 12 bits, and the elevation value can be 6 bits (including 2 bits of floating point precision) to form an 18-bit string as the geographic feature code value of the point coordinates. The geographic feature code values of the starting and ending points of the road are calculated in the above way and additionally expanded to the extracted road data table.

[0045] A2-2 Verification and repair of road connection relationships

[0046] As an optional implementation, the road connection relationship data can be used in combination with the A2-1 code to verify and repair the road connection, specifically:

[0047] For any roads represented by the road connection relationship data, obtain the geographical coordinates of the end point of the predecessor road and the starting point of the successor road, and use the A2-1 method to calculate their geographical feature code values. If the two values are in the front, If the coordinates of the connection relationship are equal, it is considered that the connection relationship is consistent with the coordinate expression on the vector data, that is, it can also be effectively expressed on the constructed road network map; otherwise, it is considered that the coordinates of the connection relationship are inconsistent on the vector data; The closer the value is to It is considered that the smaller the allowable error range is, the more accurate the absolute value of the specific error is and the precision range of the Geohash digit is.

[0048] If there are inconsistent coordinates in the connection relationship on the vector data, it can be judged and corrected based on the following rules:

[0049] 1) Calculate the actual distance between the two endpoints in the geodetic coordinate system, using an appropriate threshold as the primary criterion. The threshold selection can be determined based on the accuracy of the vector map.

[0050] 2) Optionally, refer to the consistency of other attributes of the road before and after comparison as a secondary judgment criterion;

[0051] 3) Optionally, decide whether to make corrections based on the actual situation;

[0052] 4) If correction is required, use the two points as the starting and ending points according to the road connection sequence, use their geometric coordinates and geographic feature codes to construct a new connecting line, and append it to the data set output by A2-1.

[0053] A2-3 Construct the minimum straight line segment based on the road line

[0054] As an optional implementation, road reference lines (line string data) are split into segments to obtain minimum straight line segment data, which consists of only two endpoints. For each minimum straight line segment, the geographic feature codes for the origin and destination points are calculated using the A2-1 encoding method. The reference information (such as the road number and name) is inherited from the original road table and expanded horizontally into the minimum line segment table. This implementation can be used to resolve the problem of road data not being disconnected at junctions and merging points. By effectively disconnecting the road data using minimum straight line segments, Geohash encoding is used to autonomously determine the road connectivity at junctions and merging points when constructing the road network map.

[0055] A3 builds a basic road network map

[0056] The basic road network diagram is constructed based on the data obtained from A1 / A2. The basic road network diagram is defined as ,in is a set of vertices, is the edge set.

[0057] The set of all start and end point codes of a road line or minimum line segment As the vertex set V of the graph:

[0058]

[0059] Road lines or minimal line segments As the edge set of a graph :

[0060]

[0061] Other attributes in the minimum segment data are attached to the graph as attributes of the edge superior.

[0062] A4 Use the basic road network diagram to review road connections

[0063] As an optional implementation, use Use GIS tools to review road connection issues in the road network and revise road line data and road network maps.

[0064] A4-1 Calculating the in-degree and out-degree of a vertex

[0065] In-degree is the number of nodes pointing to a vertex. The number of all directed edges of , expressed as:

[0066]

[0067] in, Indicates that from the vertex point to ;

[0068] The out-degree is the The number of directed edges pointing to other vertices, expressed as:

[0069]

[0070] in, Indicates that from the vertex point to The directed edges of

[0071] For the vertices in the basic road network graph, calculate the corresponding in-degree and out-degree combinations:

[0072]

[0073] The reasonable vertices in the road network graph correspond to the entrances, exits, divergences, confluences or intermediate points in the real road. The in / out degree combination of each type of vertex has different characteristics in terms of value. For example: the road network entry point is , the road network exit point is , they are usually at the border of the map / graph; for example: the diversion point is usually The confluence point is usually , the midpoint is .

[0074] The abnormal road connection in the road network graph will be reflected by the in-degree and out-degree characteristics of the vertices. For example, when a road that should be connected is disconnected in the data and graph, the disconnection will have an in-degree and out-degree of and When the roads at the diverging point are not connected in the data and the map, an abnormal vertex with an in-and-out degree of and The abnormal vertex; the same applies to the confluence point.

[0075] A4-2 Outlier Screening and GIS Viewing

[0076] For the road connection problem to be detected, the in-and-out degree characteristics of the graph vertices are used for screening;

[0077] Perform Geohash decoding on the selected problem points to obtain their coordinate values;

[0078] Use GIS tools to display problem points on the GIS map, combine them with the original vector data for analysis, and repair the problematic vector data.

[0079] After this process is completed, the repaired vector data can be used to regenerate the repaired basic road network map based on steps A1–3.

[0080] B Facility and equipment embedding

[0081] This step integrates facility and equipment information with the road network map, determines associated vertices based on geographic location relationships, or constructs new representation vertices. The specific process is as follows:

[0082] B1 For each type of facility and equipment, calculate and obtain the road network representation points (geographic feature code values):

[0083] The original data of facilities and equipment usually does not contain elevation values. When using geographic feature codes containing elevation for the vertices of the road network, the geographic feature code values of the facilities and equipment can be calculated using the equipment location data integrated with the high-precision map (disclosed in the applicant's previous application CN 119066619A).

[0084] The calculation method for the representation points of facilities and equipment on the road network is determined based on the type, purpose, geometry, spatial location, attribute information, etc. of the facilities and equipment, combined with the needs of use. The geographical geometry of facilities and equipment is usually expressed as points, lines, polygons, surfaces, etc. The processing methods for different types are slightly different. This invention provides a variety of calculation methods for the representation points of facilities and equipment on the road network, as follows:

[0085] 1) Spatial nearest neighbor (SJN) method: Calculate the coordinate point (or geometric center point) of the facility equipment and the vertex with the smallest distance in the basic road network vertex set, and use this point as the representative point. This method has a fast calculation speed and is suitable for point-type facilities and equipment.

[0086] 2) Spatial Nearest Neighbor Interpolation (SJI): Calculates the coordinates (or geometric center) of a facility and the nearest edge (i.e., the nearest road) in the underlying network edge set. The foot of the facility point on this edge and its geographic feature code are calculated. The latter is inserted into the network as a representation point, constructing a new connecting edge and deleting the old one, while inheriting the properties of the deleted edge. This method is slow to compute and is suitable for facilities that require more precise positioning.

[0087] 3) Spatial Intersection Relationship (SRI): For facilities with polygonal geometries, extract the vertices contained in the facility as representation points. If this is not possible, consider the SJN or SJI methods.

[0088] 4) Spatial Buffer Intersection (SBI): Based on the geometric shape of the facility, a buffer zone of a certain distance is constructed; the SRI method is used as the reference range calculation for the spatial relationship method based on the buffer zone;

[0089] 5) TDC method: For regional facilities covered on the road network, all upstream and downstream tributary points (corresponding vertices) in the coverage area can be used as the representation points of regional facilities;

[0090] 6) Other methods: Depending on the characteristics and application requirements of specific facilities and equipment, projection methods, weight center methods, or composite algorithms can also be used to obtain the representation points of the facilities and equipment on the road network to ensure that the spatial relationship between the facilities and equipment and the road network can be accurately expressed.

[0091] It should be noted that the same facility and equipment may have different vertex calculation methods in different business application scenarios and may exist at the same time.

[0092] B2 Facility and Equipment Embedding: The core purpose of this step is to embed the results of B1 (various facility and equipment representation points and corresponding attribute information) into the road network directed graph to achieve the integration of facilities and equipment with the road network. The specific method is as follows:

[0093] B2-1 Facility Characterization Point Data Integration

[0094] Integrate the data of each facility characterization point into a dataset that includes the following key information:

[0095] 1) Basic information such as the unique identifier, name, category code, coordinates, and geographic feature code of the facility and equipment;

[0096] 2) The geographic feature code value of the characterization point corresponding to the facility and equipment;

[0097] 3) Description of the relationship between facilities and equipment and the characterization points, distance, acquisition method and other information.

[0098] B2-2 Representation Point Embedding

[0099] Each representative point in the obtained B2-1 dataset is a vertex in the base road network graph. These vertices are labeled with facility or equipment attributes, indicating that they represent facilities or equipment. Furthermore, the facility and equipment attribute data is encapsulated as a dictionary within the vertex attributes, completing the road network embedding of the facilities and their attributes. This processed road network graph not only contains road connection topology information but also facility location and attribute information, achieving a fusion of the graph and facilities.

[0100] C Road Network Optimization

[0101] Road network graph optimization aims to reduce graph complexity through structural simplification while preserving the network's topological relationships and geometric characteristics. The optimized road network graph can significantly reduce the number of vertices and edges, thereby reducing computational complexity and improving the efficiency of graph algorithms. To achieve this goal, the present invention provides a recursive and iterative edge merging method for optimizing road network graphs. This method is applicable to both the base road network graph obtained in process A and the facility-embedded road network obtained in process B.

[0102] C1 graph optimization preparation

[0103] The preparation work includes: 1) determining the set of vertices to be optimized as needed, and 2) determining the aggregation strategy of the predecessor and successor edge attributes of the vertices to be optimized as needed.

[0104] C1-1 For a given road network directed graph to be optimized , decompose the vertex set V into four mutually exclusive subsets: 、 、 and :

[0105]

[0106]

[0107]

[0108]

[0109] in,

[0110] is the entry vertex of the road network graph, which is represented by the set of vertices with zero in-degree on the road network graph;

[0111] is the exit vertex of the road network graph, which is represented as a set of vertices with zero out-degree on the road network graph;

[0112] It is a set of intermediate vertices (transfer vertices), which are represented on the road network graph as vertices with both in-degree and out-degree 1. The vertices to be optimized need to be further screened from this set;

[0113] A set of junction vertices, represented on a road network graph as vertices with non-zero in-degree and non-zero out-degree, and at least one of the in-degree and out-degree is greater than 1; it represents a junction, bifurcation, or convergence point, and is a key node with multiple inputs and multiple outputs, multiple inputs and a single output, or a single input and multiple outputs.

[0114] Furthermore, if G is a road network graph with facility embedding, then the facility vertex set among all vertices is extracted: :

[0115]

[0116] C1-2 obtains the attribute set A of all edges in the road network graph to be optimized, and further decomposes A into two mutually exclusive subsets:

[0117]

[0118] in, When merging edges, the predecessor and successor edge attributes need to maintain a consistent set of edge attributes. When merging edges, the predecessor and successor edge attributes do not need to maintain the same set of edge attributes.

[0119] The edge attributes described here typically include: road number, road type, milepost directionality, and number of lanes. Whether consistency is required or not can be determined based on business needs. For example, if the attributes of a predecessor edge X and a successor edge Y differ only in the number of lanes across all fields, they will not be merged if this attribute needs to be distinguished from the perspective. If not, they can be merged, and the attribute value will be updated according to the aggregation strategy during the merge.

[0120] based on ,Will Decompose into two mutually exclusive subsets: and :

[0121]

[0122]

[0123] in,

[0124] For the predecessor and successor edges A set of intermediate vertices with consistent attribute values;

[0125] For the predecessor and successor edges The set of intermediate vertices that have at least one inconsistent attribute value.

[0126] C1-3 Determine the vertex set to be optimized

[0127] When optimizing the basic road network diagram, use As the vertex set to be optimized; if the graph to be optimized is a road network graph with embedded facilities, use and The difference of is used as the vertex set to be optimized:

[0128] C1-4 Determine the processing strategy of edge attributes during optimization

[0129] The core of merging and simplifying edges in a road network graph lies in removing vertices and constructing new edges. In this process, appropriate merging strategies are used to define the properties of the new edges. The following strategies can be adopted according to usage requirements:

[0130] Merge strategy: If the attributes of the predecessor and successor edges have the same value, the attributes of the new edge inherit those of any edge.

[0131] Override strategy: When the attributes of the predecessor edge and the successor edge are different, the attributes of the new edge inherit the specified edge;

[0132] Accumulation strategy (add): The attribute of the new edge is the sum of the attribute values of the predecessor edge and the successor edge;

[0133] Removal strategy (remove): When the attributes of the predecessor edge and the successor edge are different, the attribute of the new edge takes an empty value.

[0134] Specifically, define the function , Vertex to be optimized The front side, For its successor edge, for any attribute , when edges are merged, the new edge Replace the old side ,property Must be consistent on the front and back edges and inherited to .

[0135] for , which can be accessed through specific functions when merging edges. deal with.

[0136]

[0137] The processing functions are defined as follows and can be defined and expanded according to actual needs:

[0138] 1) Accumulation function:

[0139] 2) Maximum function:

[0140] 3) Minimum function:

[0141] 4) Average function:

[0142] 5) Previous value function:

[0143] 6) Post-value function:

[0144] 7) Null value function:

[0145] Other functions: other processing functions defined according to actual needs.

[0146] C2 execution graph optimization

[0147] right The vertices in the graph are removed one by one recursively, while adding new edges to the graph. , and remove the predecessor and successor edges of the deleted vertex , thus completing the road network graph optimization. If the new edge is repeated with the existing one, the new edge will not be added. The codes of the removed vertices can be put into the vertex list and edge attributes in order. The specific execution steps are as follows:

[0148] C2-1 Initialize and optimize the environment

[0149] 1) The original road network diagram Copy and create the graph to be optimized

[0150] 2) Based on Create a queue of vertices to be processed.

[0151] C2-2 performs edge merging iterations

[0152] For each vertex to be optimized , perform the following processing:

[0153] 1) Determine With only front-wheel drive and the only successor edge ;

[0154] 2) Check all If the merging condition is not met, skip this vertex;

[0155] 3) Create new edges , merge attributes:

[0156] For attributes :

[0157] For attributes :

[0158] 4) Processing geographic feature code sequence: from and Extract geographic feature coding sequences and If it does not exist, set it to an empty set and build a new sequence ,in Vertex The geographic feature code value, and Assign as an attribute ;

[0159] 5) Update the road network map: add , remove 、 and possible vertices ;

[0160] 6) Based on the merge results, update the set of vertices to be optimized and continue iterating until there are no more vertices to be optimized.

[0161] C2-3 Optimization Result Verification

[0162] 1) Verify the topological integrity of the optimized road network and ensure that key nodes (such as intersection nodes) and facility nodes )reserve;

[0163] 2) Calculate optimization efficiency indicators, such as vertex and edge reduction rates;

[0164] 3) Establish a mapping relationship between nodes before and after optimization to facilitate subsequent application queries.

[0165] 3D model release and computable road network system

[0166] From the perspective of engineering applications, various computable road networks need to be organized and published according to the following mechanisms:

[0167] 1) Normalize model naming

[0168] 2) Persist the calculated model

[0169] 3) Calculate the file hash value of the persisted model and record it in the model list according to the version number.

[0170] 4) Before loading the corresponding model, the hash value needs to be checked for consistency

[0171] The computable road network generated by the present invention can be constructed into a systematic model library, the hierarchical relationship of which is as follows:

[0172] a) First level: basic road network map (used for preprocessing calculation);

[0173] b) First-level derivatives: optimized version of the basic road network graph (for production use, graph computing scenarios requiring no facility analysis);

[0174] c) Second-level derivative: facility embedding graph (for pre-processing calculations);

[0175] d) Second-level derivative: Facility-embedded graph optimized version (for production use, graph computing scenarios that require facility analysis).

[0176] Example 2

[0177] This example uses the expressways of Province A as an example to illustrate the method for implementing a computable road network. A map collection unit with surveying and mapping qualifications has collected maps of all expressways in the province, forming a high-precision map dataset. This example uses this high-precision map dataset and its ancillary facilities dataset as foundational data to construct a computable road network.

[0178] High-precision map data content items include road reference line vectorized data (including related attributes), road connection relationship data, lane centerline vectorized data and related attribute data, and lane centerline connection relationship data.

[0179] Highway facilities and equipment data are selected based on actual business needs and can be divided into point type, line type, and polygon type according to facility type and vectorization method. Including but not limited to:

[0180] 1) Point-type roadside facility data, such as checkpoints, gantries, and toll station entrances and exits;

[0181] 2) Line type roadside facility data, such as guardrails, medians, and curbs

[0182] 3) Polygonal roadside facility data, such as hubs, service areas, and toll booths

[0183] This embodiment uses the most commonly used demand benchmarks in the operation and management of expressways in the province and selects four types of facility data, including toll station entrances and exits (point type), gantries (point type), hubs (polygon type), and service areas (polygon type), for road network embedding.

[0184] To facilitate engineering use, the original data used to construct the computable road network are stored in the form of vectorized data in a spatial database. This embodiment uses two types of spatial databases: PostGIS and Spatialite.

[0185] The data loaded by road data includes:

[0186] Road reference line: road reference line vector data, including the following information: road data ID, route name, route code, road type, road function, number of lanes, mileage direction, and geographic geometry information (geom);

[0187] Road reference line connection: The connection relationship data between road reference lines, including the following information: index number, predecessor road ID, and successor road ID.

[0188] The construction of a computable road network based on the above data includes the following steps:

[0189] (1) Construct a basic road network map;

[0190] Using the geom information of the road reference line, the Geohash coordinate string is constructed in the order of the line string of the route and stored in the form of a list.

[0191] In this embodiment, to maximize the accuracy of the original coordinates, the encoding precision of Geohash is set to 12, the elevation information is encoded using 6 bits (where the floating point precision is 2), and finally an 18-bit composite code is used, for example: (WTTY4BQ6ZPV5009507, WTTY4BQ6V2K009508, WTTY4BQ6XKJ2009505, ...)

[0192] Based on the A2-3 method, all the coordinates of all the points on all the road reference lines are encoded and split pairwise. Therefore, any road is converted into a nested list representing all the minimum straight line segments within it, such as:

[0193] [(WTTY4BQ6ZPV5009507, WTTY4BQ6V2K009508), (WTTY4BQ6V2K009508, WTTY4BQ6XKJ2009505), …].

[0194] After flattening the list by the smallest straight line segment, a data table (road reference segment) is generated, in which the geographic location of the smallest straight line segment is represented by both longitude and latitude and geographic feature codes to facilitate one-to-one correspondence between GIS and road network maps; other field information in the original road reference line is inherited accordingly.

[0195] Based on the A2-2 method, a supplementation method is used to correct missing connections in vector data. This embodiment adopts a relatively strict judgment method, that is, in the preceding and following road connection relationship data, the 18-bit code of the predecessor road end point must be equal to the 18-bit code of the successor road end point, ensuring that the road connections are consistent in latitude, longitude and altitude, and supplementing inconsistent connection relationships.

[0196] Following the above steps, a minimum straight segment dataset was obtained based on the vector roads contained in the original road reference line vector dataset for Province A's expressways in this example. Next, data was obtained from the road connectivity table to verify road connectivity. Roads with spatial discontinuities were screened and supplemented as needed, completing the final construction of the basic expressway network map for Province A.

[0197] (2) Facility data embedding;

[0198] Depending on the type of traffic facility equipment and the required mapping method, the representation points of the facilities on the road network are calculated separately.

[0199] The calculation methods for representing points on the road network for transportation facilities and equipment vary depending on their type, purpose, geometry, spatial location, and attribute information. First, the geometric attributes of the facility itself must be clearly defined, such as point, line, or polygon. Second, the calculation method for its road network representation points must be determined based on the facility's function and business application. The same facility may use different calculation methods for its representation vertices in different business application scenarios. Available calculation methods include the SJN, SJI, SRI, and SBI methods described in Example 1.

[0200] This example uses the most common requirements for expressway operations and management in Province A as a benchmark, selecting the following six types of facility data for road network embedding: toll station entrances and exits, gantries, information boards, cameras, hubs, and service areas. Depending on the type and scenario, different methods are used to obtain the representation vertices of each facility and equipment on the underlying road network. Table 1 below summarizes the types of facilities and equipment and the calculation methods used to find the representation points in this example.

[0201] Table 1 Types of facilities and equipment, and calculation methods for finding representative points

[0202]

[0203] The facility embedding process is used to attach the facility information represented by the representation points to the directed graph object in the basic road network graph, so that the object can use its vertices to express and locate the corresponding facilities in the road network. The process is divided into two steps:

[0204] a) Integrate facility and equipment characterization point datasets

[0205] The characterization point data obtained for the six types of facilities mentioned above are merged to form an embedding dataset. This integrated dataset only contains basic information used for facility location and description, including: facility equipment ID (unique number), name, type, coordinates and Geohash, characterization point coordinates and Geohash, and the spatial relationship and distance between the facility equipment and the characterization point.

[0206] b) Obtaining the embedded road network map of facilities and equipment

[0207] The facility and equipment information is embedded into the basic road network map using the representation point dataset to obtain the facility embedded road network map. The process is as follows:

[0208] For each representation point, add a facility mark to the corresponding vertex of the graph (that is, define the is_fac variable as True), indicating that the vertex is both a road network vertex and a facility point.

[0209] Add a facility type set (fac_types) to the vertex attributes to indicate what types of facilities the vertex contains when multiple facilities share the same vertex. Add a data table to the attributes of the facility vertex to store detailed information about the facilities corresponding to the vertex. The data source is provided by the facility attribute data table.

[0210] The embedding process preferably includes a data integrity check. If a representative point is not included in the road network map, a warning will be issued and recorded. After embedding is complete, a road network map with facility information is obtained. This road network map not only contains the basic road network topology but also includes the location and attribute information of roadside facilities and equipment, which can be used for more complex and comprehensive road network analysis and calculations.

[0211] (3) Optimize road network;

[0212] This process optimizes the existing base network or facility-embedded network by reducing its size to meet the efficiency requirements of high-performance non-loaded computational analysis. The network constructed from a vector map contains a large number of intermediate vertices. By balancing attribute preservation with graph size, a merging strategy is employed to remove these intermediate vertices and sequentially record Geohash values as attributes within the merged edges, significantly compressing the network.

[0213] Taking the facility embedding road network graph as an example, we first extract the set of middle vertices of the road network graph to be optimized. ; Original image vertex , the number of edges indivual, The quantity is from Further remove the facilities vertices, and obtain the final set of vertices to be optimized ;total indivual.

[0214] Edge optimization is performed according to the method shown in Example 1. The example shows that the optimization method proposed by the present invention can significantly reduce the scale of the road network graph while retaining key topological structures and facility information:

[0215] Number of vertices: From Reduce to , scale reduction rate ;

[0216] Number of edges: From Reduce to , scale reduction rate .

[0217] Figure 2 、 3The comparison effects of the overlay of a local hub of the road network map on the GIS system with the high-precision map and facilities before and after optimization in this embodiment are respectively demonstrated.

[0218] (4) Data issuance.

[0219] Data of various road network map objects produced by the above process are issued for loading, verification and use in the business system during the production process.

[0220] The computable road network generated by the present invention is essentially a graph-objectified road network and facilities (and related attributes). Therefore, various general graph algorithms can be used based on its products to meet the analysis and calculation requirements of different businesses for road networks. Figure 4 This is an example of using a computable road network to perform facility topology connectivity analysis.

Claims

1. A method for constructing a computable road network, characterized in that: include; Extracting road data from map data, including at least road reference line vector data; Using line string data in the vector data or line segment data obtained by splitting the line string data as connecting edges, constructing a connecting edge dataset, assigning attribute information of the original road to the corresponding connecting edges, using the starting and ending points of each connecting edge as vertices and assigning geographic feature codes to the vertices to construct a basic road network map; wherein the geographic feature codes include Geohash code information; Extract the set of intermediate vertices of the basic road network graph and the set of edge attributes of the edges connecting the intermediate vertices; Extracting a first set of edge attributes from the edge attribute set, in which the predecessor and successor edge attributes must remain consistent when merging edges; Extracting intermediate vertices whose predecessor edges and successor edges have the same attribute values in the first set from the intermediate vertex set to form a vertex set to be optimized; The predecessor edges and successor edges of the vertices in the set of vertices to be optimized are merged into new edges, the road network graph optimization is completed, and the computable road network is obtained.

2. The construction method according to claim 1, characterized in that The line string data of the road reference line is split into the smallest straight line segments to obtain the smallest line segment data set, and the basic road network diagram is constructed using the smallest line segment data as the connecting edges.

3. The construction method according to claim 1, characterized in that The geographic feature code comprises a Geohash code calculated based on vertex geometric coordinates, or a combination of a Geohash code and an elevation code; the elevation code is a code containing elevation information; When the geographic feature code adopts a combination of Geohash code and elevation code, the coding sequence of the Geohash code and elevation information is spliced together to form a final code.

4. The construction method according to claim 1, wherein The construction method further includes verifying the accuracy of road connection relationship data using geographic feature coding, including: Obtain road connection relationship data, and calculate the geographic feature code value for the end point of the predecessor road and the starting point of the successor road for any connected road in the road connection relationship data. If the calculated geographic feature code value is exactly the same or the previous one in the Geohash code is the same n If the positions are consistent, the roads are connected.

5. The construction method according to claim 4, characterized in that The accuracy verification also includes, when the connection relationship verification fails, making a correction judgment on the connection relationship. If the connection relationship needs to be corrected, a road virtual connection line is constructed using the geometric coordinates and geographic feature codes of the missing connection points and appended to the connection edge dataset.

6. The construction method according to claim 1, characterized in that The construction method further includes, after constructing the basic road network map, using the basic road network map to review road connection problems and correcting the road reference line vector data and the basic road network map, including: Calculate the in-degree and out-degree of the vertices of the basic road network graph; Filter the vertices to be checked based on the in-degree and out-degree characteristics of the graph vertices; Perform Geohash decoding on the vertex to be checked to obtain the vertex coordinate value; Display the vertices to be checked on the GIS map, analyze and judge them in combination with the original vector map data, and repair the vector map data with problems; The repaired basic road network map is regenerated using the repaired vector map data.

7. The construction method according to claim 1, characterized in that The construction method further includes, after constructing the basic road network graph, embedding the facilities and equipment in the transportation system as vertices into the road network graph to obtain a road network graph integrating the information of the facilities and equipment; For a road network graph that integrates facility and equipment information, during vertex optimization, vertices embedded with facilities and equipment are removed from the vertex set to be optimized.

8. The construction method according to claim 7, characterized in that: The embedding of facilities and equipment in the transportation system as vertices into the road network graph includes: According to the geographical location relationship between the facilities and equipment and the road network, the vertices related to the facilities and equipment in the road network are obtained as the facility and equipment road network representation points, and the attribute information of the facilities and equipment is embedded in the corresponding road network representation points; or, According to the geographical location relationship between facilities and equipment and the road network, the road network representation point of the facilities and equipment is calculated, the attribute information of the facilities and equipment is bound to the representation point, and the road network representation point is inserted into the road network graph as a new vertex. New connecting edges are constructed and the old edges are deleted after inheriting the edge attributes.

9. The construction method according to claim 8, characterized in that: The method for obtaining or calculating the road network representation point includes: a. Calculate the coordinate point or geometric center point of the facility equipment and the vertex with the smallest distance from the vertex set of the basic road network graph as the candidate road network representation point; or b. Calculate the nearest edge between the coordinate point or geometric center point of the facility and the edge set of the underlying road network graph, and calculate the foot of the perpendicular to the edge as the candidate road network representation point; or c. extracting vertices contained in the facilities as candidate road network representation points; Or d. Based on the geometric shape of the facility, construct a buffer zone of a preset distance; use the buffer zone as the calculation range, and extract the vertices contained in the calculation range as candidate road network representation points; Or e. For regional facilities covered by a road network, use the vertices corresponding to all upstream and downstream junctions and confluences in the covered area as candidate road network representation points for the facility; or f. Based on the characteristics of the facilities and equipment and the application requirements, calculate candidate road network representation points that accurately express the spatial relationship between the facilities and equipment and the road network.

10. The construction method according to claim 1, characterized in that: For the vertices in the optimized vertex set, they are deleted one by one in a recursive manner. When deleting, new edges are added to the graph, and the predecessor and successor edges of the deleted vertices are removed, thereby completing the road network graph optimization.

Images