A road network pattern recognition method and system based on unsupervised graph representation learning

Abstract

The present invention provides a road network pattern recognition method and system based on unsupervised graph representation learning. First, the road network is modeled through a spatial dual graph, and the graph node features are designed based on cognitive inspiration. Then, the proposed graph neural network model is based on GAE, which can be trained in an unsupervised manner. At the same time, subgraph isomorphism count (SIC) is introduced in the road segment embedding learning stage and global context attention mechanism (GCA) is introduced in the graph embedding generation stage to enhance the model representation performance. Finally, the geometric similarity of graph-level embedding is used to identify road network patterns. Experimental results show that the present invention outperforms the classic road network recognition method in all indicators in the experimental area, and the classification accuracy is improved by more than 12%, which is even slightly better than the supervised baseline graph neural network method.

Description

Technical Field

[0001] The present invention relates to the technical field of road network pattern recognition, and in particular to a road network pattern recognition method and system based on unsupervised graph representation learning. Background Art

[0002] As a basic geographical element, understanding the spatial structure and morphological characteristics of the road network helps to explore the relationship between the earth's surface environment and human activities.

[0003] Classic road network pattern recognition methods are primarily based on geometric calculations and graph theory. Geometric calculation methods can be categorized into linear and planar feature-based methods, depending on the recognition unit. Linear feature methods identify road network patterns based on the geometric characteristics of road segments. Planar feature methods identify road network patterns based on the spatial morphology of road network cells. These geometric calculation-based road network pattern recognition methods primarily rely on the shape characteristics of the road network, ignoring the topological connectivity properties of the road network as a spatially connected network. Graph theory, in contrast, offers an effective modeling approach that integrates both geometric shape and topological connectivity. Generally speaking, traditional algorithms mostly focus on describing the characteristics of specific patterns, requiring separate algorithms for different patterns and resulting in limited generalization. Furthermore, road network pattern recognition is highly subjective and complex due to the influence of spatial perception and visual psychology. Traditional algorithms lack intelligent reasoning and decision-making processes, making them difficult to achieve human-level recognition. Therefore, there is a need to explore recognition methods with strong generalization and closer to human perception to effectively identify and classify diverse road network patterns.

[0004] In recent years, deep learning technology, especially graph neural networks (GNNs), has made significant progress in processing graph data. Deep learning methods based on GNNs can automatically extract road network features without the need for manually designed feature extractors. Compared with traditional algorithms, this method reduces manual intervention and improves the objectivity and accuracy of road network pattern recognition. However, it relies on a large amount of labeled training data, and the annotation of road network data is usually complex and expensive. Model training and inference also consume a lot of computing resources, especially when the data scale is large. In addition, the embedded representations generated by GNNs models mainly serve specific tasks (such as pattern recognition) and may no longer be applicable or perform poorly when applied to new tasks, showing their lack of generalization ability and limited flexibility. Summary of the Invention

[0005] In view of the problems that existing GNNs-based road network feature extraction solutions rely on a large amount of labeled training data and the generated embedded representation has insufficient generalization ability, the present invention provides a road network pattern recognition method and system based on unsupervised representation learning.

[0006] In one aspect, the present invention provides a road network pattern recognition method based on unsupervised graph representation learning, comprising:

[0007] Step 1: Take the stroke segments in the target road network data as the nodes of the graph, and the connection relationship between adjacent segments and the spatial relationship added for spatially similar segments as the edges of the graph to construct the spatial dual graph of the target road network data, and generate the adjacency matrix of the target road network data based on the spatial dual graph;

[0008] Step 2: Extract the segment features of each stroke segment in the target road network data to generate a feature matrix of the target road network data; wherein the segment features include geometric attribute features and shape context SC attribute features of the segment;

[0009] Step 3: Based on the spatial dual graph, adjacency matrix, and feature matrix of the target road network data, a preset graph neural network model is used to perform road graph representation learning to generate a graph-level embedding for the target road network data. The graph neural network model includes a GAE model and a structure-aware graph representation learning module. The structure-aware graph representation learning module uses the node-level embedding output by the encoder in the GAE model as input to generate a graph-level embedding.

[0010] Step 4: Measure the similarity between the graph embedding of the target road network data and the graph embedding of multiple preset standard anchor points, and take the road network pattern of the standard anchor point with the highest similarity value as the road network pattern of the original road network data.

[0011] Furthermore, step 1 also includes: extracting spatially similar road sections and adding spatial relationships for the spatially similar road sections; wherein, the spatially similar road sections refer to two road sections that meet preset spatial requirements but have no connection relationship; the preset spatial requirements refer to that the distance similarity and angle similarity of the two road sections are both greater than 0.

[0012] Furthermore, the distance similarity between two road sections is calculated according to the following formula: and angle similarity

[0013]

[0014]

[0015] Among them, sp_dist(s i ,s j ) indicates calculating two road segments s i and s j One-way Hausdorff distance between, ag_dist(s i ,s j ) indicates calculating two road segments s i and sj The absolute angular distance between ds and δ as Indicates two preset distance thresholds.

[0016] Furthermore, the geometric attribute features of the road section include position features extracted from the macro scale, connection features extracted from the meso scale, and length features, shape features and angle features extracted from the micro scale.

[0017] Furthermore, the process of extracting the SC attribute features of the road section includes:

[0018] Establish a polar coordinate system at the midpoint of the target road section, and set the maximum radius of the polar coordinate system to r = γ × L(s); where L(s) is the length of the target road section and γ is the scale factor;

[0019] Divide the polar coordinate system into p concentric circles at equal distances and q sectors at equal angles to obtain p×q sectors, and define the value of each sector as the ratio of the total length of the road segments in the region to the total length of the road segments in the context range;

[0020] The values ​​of p×q sectors are expressed as a vector, which is the SC attribute feature of the target road segment.

[0021] Furthermore, a subgraph isomorphism counting (SIC) module is provided in the encoder, and the SIC module is used to generate structure identifiers for nodes in the space dual graph. The encoder performs neighborhood information transfer based on the structure identifier, adjacency matrix and feature matrix of the node to generate node-level embedding.

[0022] Furthermore, the method further includes: training the encoder and decoder in the GAE model; during the training process, the decoder reconstructs the inter-node association based on the node-level embedding output by the encoder to generate a reconstructed adjacency matrix, and uses the following cross-entropy loss function to calculate the distance between the reconstructed adjacency matrix A′ and the true adjacency matrix A:

[0023]

[0024] Where N represents the total number of nodes in the space-dual graph.

[0025] Furthermore, the process of generating node-level embedding specifically includes:

[0026] Define a subgraph collection For subgraph collections For each subgraph H in the space dual graph G, first find its isomorphic subgraph G in the space dual graph G S , for G S Each node in Find the orbital Orb of its mapping f(v) in the subgraph H H (f(v)), Orb H (*) represents the preset orbital mapping function, where orbit refers to the position or role of a node in H; then, the position of each node v in different isomorphic subgraphs G is calculated. S The track times, and obtain the structural characteristics of node v on the i-th track of subgraph H represents the i-th track in the subgraph H; f(v) represents the node mapping function that maps node v to the node in the subgraph H; finally, the combined subgraph set Count the different subgraph structures and orbits in the node to get the structure identifier of the node in, is subgraph H i the number of mid-tracks;

[0027] The adjacency matrix, feature matrix, and structure identifier are updated according to the following propagation rules between adjacent convolutional layers:

[0028]

[0029] in, A is the adjacency matrix, is the identity matrix, yes degree matrix of ; is the convolution signal matrix of the lth layer, representing the features of each node in the lth layer, m l is the dimension of the node feature at layer l; the input layer H 0 =M, Represents the feature matrix, d1 represents the dimension of the node's geometric attribute feature, d2 represents the dimension of the node's SC attribute feature; X v is the structure identifier of the node, and is the parameter training weight matrix and common bias of layer l, σ(·) is the nonlinear activation function, and N represents the total number of nodes in the space-dual graph G.

[0030] Furthermore, the structure-aware graph representation learning module takes the node-level embedding output by the encoder in the GAE model as input to generate graph-level embedding, specifically including:

[0031]

[0032] Among them, E S represents the graph-level embedding, Z represents the node-level embedding output by the encoder, is the embedding of the nth node, N represents the total number of nodes; is a learnable weight matrix, and tanh(·) is the activation function; represents the attention weight of the node, and σ(·) is a nonlinear activation function.

[0033] In another aspect, the present invention provides a road network pattern recognition system based on unsupervised graph representation learning, comprising:

[0034] The spatial dual graph construction module is used to construct the spatial dual graph of the target road network data by using the stroke road segments in the target road network data as the nodes of the graph and the connection relationships between adjacent road segments and the spatial relationships added for spatially similar road segments as the edges of the graph;

[0035] An adjacency matrix generation module, configured to generate an adjacency matrix of target road network data according to the space dual graph;

[0036] A feature matrix generation module is used to extract the segment features of each stroke segment in the target road network data to generate a feature matrix of the target road network data; wherein the segment features include geometric attribute features and shape context SC attribute features of the segment;

[0037] A graph-level embedding generation module is configured to generate a graph-level embedding for the target road network data by learning a road graph representation based on the spatial dual graph, adjacency matrix, and feature matrix of the target road network data using a preset graph neural network model; wherein the graph neural network model includes a GAE model and a structure-aware graph representation learning module; the structure-aware graph representation learning module uses the node-level embedding output by the encoder in the GAE model as input to generate a graph-level embedding;

[0038] The road network pattern recognition module is used to measure the similarity between the graph embedding of the target road network data and the graph embedding of multiple preset standard anchor points, and take the road network pattern of the standard anchor point with the highest similarity value as the road network pattern of the original road network data.

[0039] Beneficial effects of the present invention:

[0040] (1) The road network pattern recognition method and system of the present invention address the issue of insufficient spatiality in road network modeling by introducing spatial edges to improve road graph construction, resulting in a spatial dual graph. This spatial dual graph can be used to better learn embedded representations of road network data. Furthermore, the graph-level embedding of road network data is learned based on a Graph Autoencoder (GAE) model, eliminating the need for large amounts of labeled training data, and the generated graph-level embedding is independent of specific downstream tasks.

[0041] (2) When extracting the initial node features of the road graph, the present invention follows the Gestalt theory and extracts visual variables such as position, connection, length, shape and angle from macro, meso and micro perspectives as the geometric features of the road segment. At the same time, the SC attribute features of the node are extracted to describe the spatial structure pattern of the neighborhood of the target road segment. The geometric attributes of each road segment and the SC attributes formed by its neighborhood are integrated as the features of the corresponding graph nodes, thereby obtaining a more effective road graph embedding representation.

[0042] (3) During the encoder's road segment embedding learning phase, this paper introduces subgraph isomorphism counting (SIC) to enhance the model's expressiveness. During the graph-level embedding generation phase, a global context attention (GCA) mechanism is added to improve the representation of road network data. Through SIC and GCA, the model can better capture and understand the local structure and global topological relationships in the road network, thereby generating a more representative, robust, and structure-aware road graph-level embedding representation that better captures road network structural information and significantly improves the performance of road network pattern recognition tasks.

[0043] (4) This paper uses graph-level embedded geometric similarity to achieve road network pattern recognition. Experiments were conducted on real road network data, comparing the performance of classic road network pattern recognition algorithms and multiple graph neural network recognition algorithms. The experiments verified the effectiveness of the recognition method. Compared with classic road network pattern recognition methods, SUGAR improves the pattern classification accuracy by more than 12%. Compared with supervised baseline graph neural network methods, SUGAR can achieve comparable or even slightly better performance without the need for sample annotation.

[0044] (5) The dimensionality reduction visualization technology was used to evaluate the road graph embedding performance. The pattern recognition performance of graph neural network models using different modules (SIC, GCA) was analyzed from the two aspects of road network samples and graph embedding representation. It was found that the SUGAR model using both SIC and GCA had clear boundaries between categories, clear clustering of regular patterns, and was not confused with irregular patterns, achieving the best classification effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A schematic flow chart of a road network pattern recognition method based on unsupervised graph representation learning provided by an embodiment of the present invention;

[0046] Figure 2 The process of constructing a space dual graph provided by an embodiment of the present invention;

[0047] Figure 3 A schematic diagram of the calculation process of the geometric attribute characteristics of a road segment provided in an embodiment of the present invention;

[0048] Figure 4 Examples of SC attributes for road sections in different road network modes provided in an embodiment of the present invention;

[0049] Figure 5 The generation process of the adjacency matrix A and the characteristic matrix M provided in the embodiment of the present invention;

[0050] Figure 6 The graph representation learning process of the graph neural network model with GAE as the main framework provided by the embodiment of the present invention;

[0051] Figure 7 Examples of subgraphs and their tracks used by the road network subgraph isomorphism counting (SIC) provided in an embodiment of the present invention;

[0052] Figure 8 The subgraph isomorphism count provided by the embodiment of the present invention (taking triangle and chord cycle as examples);

[0053] Figure 9 Road network pattern recognition based on graph embedding geometric similarity provided by an embodiment of the present invention;

[0054] Figure 10 The road network pattern recognition results provided by the embodiment of the present invention (including some correct recognition examples and all incorrect recognitions) are shown. DETAILED DESCRIPTION

[0055] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly described below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0056] As a basic geographical element, the road network can be analyzed and understood by studying its vector representation method and its application in pattern recognition. Traditional road network pattern recognition methods mainly rely on the geometric characteristics and topological properties of the road network, but such methods are computationally complex and lack the ability of intelligent reasoning, and their performance degrades under complex road network structures. In recent years, recognition methods based on graph neural networks have become a research hotspot. Such methods reduce manual intervention and improve the objectivity and accuracy of recognition, but they rely on a large amount of labeled data. Moreover, the above methods are algorithms designed for specific tasks and have limited generalization capabilities. Based on this, the present invention mainly studies the road network vector representation method and its application in road network pattern recognition, mapping the complex road network structure into a road network structure pattern vector representation suitable for analysis and calculation, realizing the analysis and understanding of the road network spatial structure, and providing a computational method for the information representation and processing of the digital twin of the earth system.

[0057] The present invention uses graph neural networks to learn the representation of road network data and infers the similarity relationship between spatial structures and topological connections between different road network data based on graph-level embedding, thereby realizing pattern recognition of road networks. The basic process of the method of the present invention is as follows: Figure 1 As shown, it mainly includes 4 steps:

[0058] S101: Road graph model construction.

[0059] Specifically, the present invention aims to use graph embedding to measure the geometric similarity of road networks for pattern recognition. Given that undirected graphs are more suitable for expressing the symmetry and interconnectivity of road networks, thereby effectively capturing the geometric similarity between road networks, the embodiment of the present invention selects undirected graphs to construct a road graph model.

[0060] There are two common ways to construct a road graph: one is to use nodes and edges to represent the endpoints and segments of a road segment; the other is to use the idea of ​​a dual graph, using road segments as graph nodes and connection relationships as edges. In the first modeling method, the nodes only express the connection relationship between road segments, and it is difficult to directly consider the attributes of road segments, but these attributes are particularly important when expressing the spatial pattern of the road network. In contrast, the dual graph modeling method using road segments as nodes has the advantage of representing the attribute characteristics of road segments and their topological relationships. The construction process of the dual graph includes the following key steps: Given a road network ( Figure 2 .a), First, we pre-process and filter out noise data such as small length and isolated points and perform stroke segment extraction ( Figure 2 .b); Then, each stroke segment is regarded as a node in the dual graph, and the node is usually located at the midpoint of the stroke segment; Finally, in the dual graph, the road segments connected to the original road network are connected to the corresponding nodes through edges ( Figure 2 .c).

[0061] The road network dual graph construction focuses on the topological structure of the graph, which enables the embedding of the target node to be learned by aggregating the embeddings of neighboring nodes when applying the GNNs model on the road graph. However, this approach ignores the spatial structure of the road network. For example, Figure 2 .d, although s1 and s2 are close in space and have similar directions, they are Figure 2 The dual graph of the road network in .e is cut into two parts; similarly, although s3 and s4 are very similar in position and angle, there is no directly connected edge in the dual graph, and effective information transmission cannot be achieved. To solve this problem, the embodiment of the present invention adds a spatial relationship for spatially similar road segments. For example, in Figure 2 In .f, a red line segment is added between s1 and s2, and a red line segment is added between s3 and s4.

[0062] Based on the above, the process of constructing a space-dual graph can be summarized as follows: Stroke segments are extracted from the road network data and used to construct a graph model describing the relationships between segments. The stroke segments in the road network data are used as nodes in the graph, and the connections between adjacent segments and the spatial relationships added for spatially similar segments are used as edges to construct a space-dual graph for the road network data. Next, an adjacency matrix for the road network data is generated based on this space-dual graph.

[0063] It should be noted that, for the convenience of subsequent description, the edges corresponding to the connection relationship are called topological edges, and the edges corresponding to the spatial relationship are called spatial edges. Different from the dual graph that only considers the topological connection relationship of the road segment, the spatial dual graph ( Figure 2 g) Contains two types of undirected edges: topological edges and spatial edges, represented by solid lines (blue) and dashed lines (red), respectively. Spatial edges capture the spatial similarity between road segments, helping the model identify spatially close but not directly connected road segments in the original road network. This modeling approach enables the model to effectively capture spatial similarities that cannot be directly expressed in traditional dual graph models, optimizes the integration and transmission of information between graph nodes, and provides a comprehensive understanding of the road network through topological and spatial structures, thereby improving the accuracy of road network pattern recognition.

[0064] S102: Road map feature extraction.

[0065] Specifically, the geometric attribute features of stroke segments in road network data are extracted, and the shape context (SC) attribute features of stroke segments are introduced as the neighborhood features of the segments. The segment geometry and its neighborhood features are integrated to form a feature matrix as the initial feature description of the road graph nodes.

[0066] S103: Road graph representation learning.

[0067] Specifically, based on the spatial dual graph, adjacency matrix and feature matrix of the road network data, a preset graph neural network model is used to perform road graph representation learning to generate a graph-level embedding for the target road network data. The graph neural network model includes a GAE model and a structure-aware graph representation learning module; the structure-aware graph representation learning module uses the node-level embedding output by the encoder in the GAE model as input to generate a graph-level embedding. GAE refers to Graph Autoencoder, a graph autoencoder model for graph data, and a neural network model for unsupervised learning. Node-level embedding and graph-level embedding are two embedding methods in the graph embedding technology in this field.

[0068] S104: Road network pattern recognition based on geometric similarity.

[0069] Specifically, for different road network data, different graph-level embeddings are generated through steps S101 to S103 above. The graph-level embeddings are then used to calculate the geometric similarity between the different road network data. Finally, different road network patterns are classified based on the similarity, achieving multi-pattern recognition of road networks. Specifically, a similarity measurement is performed between the graph embedding of the target road network data and the graph embeddings of multiple preset standard anchor points. The road network pattern belonging to the standard anchor point with the highest similarity value is used as the road network pattern of the original road network data.

[0070] The road network pattern recognition method provided by an embodiment of the present invention addresses the problem of insufficient spatiality in road network modeling by introducing spatial edges to improve road graph construction, resulting in a spatial dual graph. This spatial dual graph can be used to better learn the embedded representation of road network data. Furthermore, learning graph-level embeddings for road network data based on the GAE model eliminates the need for relying on large amounts of labeled training data, and the generated graph-level embeddings are independent of specific downstream tasks. Finally, the present invention leverages the geometric similarity of graph-level embeddings to achieve road network pattern recognition.

[0071] Based on the above embodiments, the present invention provides a process for extracting spatially similar road segments. First, spatially similar road segments are defined as two road segments that meet preset spatial requirements but are not connected; the preset spatial requirements are that the distance similarity and angle similarity of the two road segments are both greater than 0.

[0072] As an implementation method, the distance similarity between two road segments is calculated according to the following formula: and angle similarity

[0073]

[0074] Among them, sp_dist(s i ,s j ) indicates calculating two road segments s i and s jOne-way Hausdorff distance ag_dist(s i ,s j ) indicates calculating two road segments s i and s j The absolute angular distance between i .radian-s j .radian|), δ ds and δ as Represents two preset distance thresholds, which set the distance similarity to zero when the spatial distance and angular distance are too large. and angle similarity The value of is set to 0 (because cosπ / 2=0).

[0075] In order to introduce spatial neighborhood information into the aggregation process, the embodiment of the present invention further proposes a spatial similarity adjacency matrix, which records the similarity between road segments according to the spatial attributes of the road segments (such as position, angle, etc.). The matrix elements The value of is shown as follows:

[0076]

[0077] It can be understood that the matrix element A s ij =1 means that the road section s i and s j When the spatial similarity between them meets the preset spatial requirements (i.e., the spatial distance is close and the spatial direction is similar), a spatial edge is added to the dual graph, such as Figure 2 As shown in .f.1; if the spatial similarity requirement is not met, or only one of the spatial distance relationship and spatial direction relationship requirements is met, no spatial edge connection is added (A s ij =0), such as Figure 2 .f.2.

[0078] On the basis of the above embodiments, after the road graph is constructed, its space dual graph is represented as G=V,A t ,A s , where V={v0,v1,…v N} is a set of nodes, N represents the total number of nodes in G; It is a topological adjacency matrix, which represents the topological connection relationship between nodes. If there is a connection v i and v j The edge e ij , then A t ij =1; otherwise A t ij=0, indicating that there is no such edge. Spatial similarity adjacency matrix Represents the spatial similarity between road segments. When the spatial distance and angular distance between two road segments are less than a certain threshold, A s ij =1, otherwise A s ij =0.

[0079] In the road network space dual graph, each node represents a road segment and carries a series of descriptive features. Because road network patterns are influenced by the morphological characteristics of roads and the connectivity of neighboring roads, relying solely on the geometric features of a single road segment is insufficient for accurate pattern recognition. Therefore, to achieve a more effective road graph embedding representation, embodiments of the present invention extract the geometric attributes of each road segment and the SC attributes formed by its neighborhood as features of the corresponding graph node.

[0080] (1) Geometric properties of a single stroke segment

[0081] The embodiment of the present invention follows the Gestalt theory and extracts visual variables such as position, connection, length, shape and angle as the geometric features of the road section from the macro, meso and micro perspectives. Specifically, for each graph node, the visual variable method is used to extract the road section attributes through specific visual features to realize the holistic perception principle in the Gestalt theory, thereby obtaining better road network spatial cognition and graphical representation. Therefore, node features are extracted from three levels: macro, meso and micro: the macro scale involves the overall layout and pattern of the entire road network, the meso scale focuses on the connection between road sections and the hierarchical relationship of different road grades, and the micro scale focuses on the specific features of a single road section. Based on these different scales, in Table 1, the embodiment of the present invention lists 6 different geometric attribute features extracted from aspects such as position, connection relationship, length, shape and angle, forming a feature matrix Calculation method as Figure 3 shown. Figure 3 In the figure, point O is the center coordinate, which represents the arithmetic mean of the midpoints of all road sections. (x p ,y p ) is the coordinate of the midpoint of the road segment, l and w are the length and width of the minimum circumscribed matrix, and P1 and P2 are the two endpoints of the longest side of the minimum circumscribed matrix.

[0082] Table 1 Six road segment features extracted from global, meso and micro perspectives

[0083]

[0084] (2) SC attributes of road section neighborhood

[0085] The embodiment of the present invention uses SC descriptors to describe the spatial structure pattern of the neighborhood of the target road segment. First, a polar coordinate system is established at the midpoint of the road segment, with a maximum radius of r = γ × L(s), where L(s) is the length of the target road segment (red); γ is a scaling factor that determines the receptive field of the road segment context feature. Then, p concentric circles are divided at equal distances, and q sectors are divided at equal angles to obtain p × q sectors. The value of each sector is defined as the ratio of the total length of the road segment in the area to the total length of the road segment in the context range; for example, the value V of the i-th sector is i Calculated as Finally, the values ​​of p×q sectors are expressed as a vector to describe the SC attribute of the road segment. For a graph with n nodes, the neighborhood features are obtained through the SC descriptor

[0086] Figure 4 The SC descriptions of four road segments with four different road network patterns (grid, complex interchange, parallel, and irregular) are shown, with parameters p, q, and γ set to 4, 8, and 1.20, respectively. It can be seen that the SC descriptions of road segments with the same road network pattern are more similar.

[0087] Figure 5 The overall process of constructing the adjacency matrix A and feature matrix M of road network data is given. Given a road network, first build its space dual graph and extract the topological adjacency matrix A t And the spatial similarity matrix A s , the two are added together to form the final adjacency matrix A; then, the geometric features M of the node segments are extracted f and SC attribute feature M sc , the 6 types of road segment features and the 32-dimensional SC attribute features are combined and spliced ​​into a 38-dimensional vector, which is used as the input feature vector M of the corresponding graph node. Finally, the road network graph is represented as the adjacency matrix A and the feature matrix M for subsequent analysis.

[0088] Based on the above embodiments, Figure 6 As shown in the figure, the graph neural network model provided by the embodiment of the present invention is a structure-aware road network graph representation learning model (SUbgraph-aware GraphAutoencoder of Road network, SUGAR) that does not rely on specific downstream tasks. The graph representation learning process is as follows: the encoder uses the subgraph isomorphism counting SIC module as the core and the road graph (M, A) as input to calculate the node-level embedding Where N is the total number of nodes, m represents the dimension of node features, and the nth row of Z is the embedding of node n. The decoder takes the node-level embedding Z as input to reconstruct the adjacency matrix The goal is to minimize the difference between the reconstructed adjacency matrix A′ and the original adjacency matrix A. In the graph representation learning phase, a simple mean readout function is used to convert the node-level embedding Z into a simple intermediate graph-level embedding Then GCA is used to calculate an attention weight for each node, and finally the structure-aware graph-level embedding is obtained by weighted average. It is used for cosine similarity calculation. The functions of each module will be introduced in detail below.

[0089] (1) Encoder

[0090] Neural networks can automatically identify repetitive patterns in data to understand complex structures. For example, when processing images, convolutional neural networks can identify different areas in the image (such as the facial features); or when processing text, they can identify different parts of a sentence (such as the subject and predicate). However, existing GNNs cannot distinguish between the different "structural roles" of nodes, that is, all nodes are treated in the same way. To this end, an embodiment of the present invention introduces a SIC module in the encoder, explicitly encodes the "structural role" as part of the message passing, and directly tells each node in the road graph the "structural role" to capture richer topological properties.

[0091] The road network spatial dual graph G is composed of a large number of subgraphs consisting of nodes and edges, where each node has a different "structural role". The "structural role" can be encoded into features by counting the number of occurrences of certain repeated subgraphs.

[0092] The embodiment of the present invention defines a set of small, connected, and recurring subgraphs for the road network pattern recognition task. They represent triangle, tailed-triangle, 4-cycle, chordal-cycle, 2-triangle and 4-clique respectively. Figure 7 shown.

[0093] For each subgraph First, find its isomorphic subgraph G in graph G. S For G S Each node in Find the orbital Orb of its mapping f(v) in H H (f(v)) to infer its role relative to H, Orb H (*) indicates the preset orbital mapping function, where orbit refers to the position or role of a node in H. Figure 7 As shown, nodes with the same color in the same subgraph are in the same orbit. Then, calculate the SThe number of occurrences of a specific orbital in H is used to obtain its structural characteristics on the i-th orbital in H. in, represents the i-th track in H; f(v) represents the node mapping function that maps node v to the node in H. In this paper, the automorphism group Aut(H) function is selected (i.e., the node transformation set that keeps the graph structure unchanged). Finally, by combining the subgraph set The counts of different subgraphs and tracks in the , and the total feature vector The dimensions is subgraph H i The number of tracks. Figure 8 Shows the use of triangle and chordal-cycle isomorphic subgraphs to count the node's structural identifier X v .

[0094] While traditional GNNs can also distinguish nodes by introducing unique identifiers, the model output changes when the node order changes, thus losing permutation equivariance. SIC, on the other hand, generates structural identifiers that are permutation equivariant because the number of subgraphs remains constant and changes in node order do not affect the structural identifier. Furthermore, using "structural roles" as identifiers is not necessarily unique; multiple nodes can share the same structural identifier, which allows this approach to strike a good balance between uniqueness and generalization.

[0095] The encoder in the present invention can be implemented using a variety of networks, such as MPNN and GIN. This embodiment of the present invention uses GCNN and introduces SIC, so messages from neighboring nodes now also contain structural information. Each node v updates its state by combining its previous state with the aggregated neighborhood message. Each convolutional layer only processes first-order neighborhood information, and by stacking multiple convolutional layers, multi-order neighborhood information can be transferred. The propagation rules of the convolutional layer are as follows:

[0096]

[0097] in, Represents the adjacency matrix of the undirected graph G plus self-connection (that is, each vertex and itself plus an edge), is the identity matrix, yes The degree matrix of is the convolution signal matrix of the lth layer, representing the features of each node in the lth layer, m l is the dimension of the node feature at layer l, the input layer H 0 =M, d1 represents the dimension of the node's geometric attribute features, and d2 represents the dimension of the node's SC attribute features.v The structure identifier of the node. and is the parameter training weight matrix and common bias of layer l. σ(·) is the nonlinear activation function. The hidden layer uses the linear rectifier unit function Relu(·), and the output layer uses the logistic sigmoid function.

[0098] (2) Decoder

[0099] During the training process of the SUGAR model, the node-level embedding Z uses the SC descriptor, which enhances the similarity between adjacent nodes while reducing the similarity between non-adjacent nodes. This embedding method makes adjacent nodes highly overlap in the embedding space, which helps to accurately reconstruct the connections between nodes in the graph.

[0100] Specifically, if two graph nodes have a high degree of similarity in the embedding space, the probability that there is an edge between them is relatively large. Therefore, the accuracy of the node-level embedding Z can be evaluated by comparing the distance between the reconstructed adjacency matrix A′ and the true adjacency matrix A. In this setting, the labels are not given externally but directly derived from the structure of the graph (i.e., the adjacency matrix). This allows the model to learn complex relationships between nodes without external supervision labels and optimize the node-level embedding by minimizing the difference between the predicted and true adjacency matrices.

[0101] Furthermore, in this embodiment, the cross entropy loss function is used to calculate the distance between the reconstructed adjacency matrix A′ and the true adjacency matrix A:

[0102]

[0103] Where N is the total number of nodes. The logarithmic loss of the predicted connection probability is calculated for each pair of nodes, and the sum is averaged to form the loss value of the entire graph.

[0104] (3) Structure-aware graph representation learning module

[0105] The encoder output is a node-level embedding Z, but the geometric similarity measure between road networks requires a graph-level embedding. Typically, the conversion from node-level embedding to graph-level embedding is achieved through a readout function such as average, maximum, or pooling.

[0106] The embodiment of the present invention uses the average operator to embed the node level Convert to simple graph-level embedding Then the GCA module is introduced. The specific operation is as follows: First, a nonlinear transformation is performed on the graph-level embedding E. Get the global context in, is the embedding of the nth node; N represents the total number of nodes; is a learnable weight matrix, and its parameters are also updated by graph reconstruction to reweight and adjust the various features of the global context vector; while tanh(·) provides the necessary nonlinearity, enabling the model to capture more complex graph structural features. The global context c provides global structure and feature information of the graph adapted to the given similarity metric through the learned weight matrix. Then, we calculate an attention weight for each node based on c, which is calculated as It is expressed as the inner product of node embedding and global context c, making its attention a n It can perceive the global context information. At the same time, the nonlinear activation function σ(·) (the sigmoid function is used here) is used to process the inner product result to ensure that the attention weight is in the range of (0,1). Finally, the graph embedding is obtained is the weighted sum of node embeddings Z, i.e. The following formula summarizes the process of generating structure-aware graph-level embeddings based on GCA:

[0107]

[0108] The embodiment of the present invention proposes a structure-aware road network graph representation learning model SUGAR that does not rely on specific downstream tasks. It combines topological edges and spatial edges to construct a road graph, uses an encoder-decoder architecture with SIC as the core to perform unsupervised learning of graph node embedding, and generates structure-aware graph-level embedding based on the GCA attention mechanism.

[0109] Based on the above embodiments, the embodiment of the present invention provides a geometric similarity measurement method based on road network graph level embedding. The road graph level embedding obtained by attention weighting is denoted as Because important nodes are given higher weights, the embedding representation maintains the main structure of the road network. The two road networks input in the test phase are denoted as E S1 and E S2 Therefore, the geometric similarity measure between road networks is regarded as the similarity between two high-dimensional vectors ( s im), the similarity is calculated by cosine similarity, and the calculation formula is as follows:

[0110]

[0111] After obtaining the similarity, we use the geometric similarity of graph embeddings to identify road network patterns. Specifically, we first select four standard road network patterns: grid, parallel, complex interchange, and radial, as standard anchor points. We then perform a geometric similarity measurement between the test set's road network and these four standard anchor points, obtaining four different similarity values. The pattern corresponding to the highest similarity value is used as the classification for the road network. However, if the highest similarity value is less than 0.5, it is classified as an irregular pattern.

[0112] Furthermore, in order to reduce the bias that may be caused by a single anchor point and enhance the generalization ability of the model to the road network pattern, this embodiment selects four representative anchor points for each road network pattern to cover the typical features and changes of the category as much as possible. Correspondingly, the similarity measure is converted into the average similarity measure. The overall process is as follows: Figure 9 shown.

[0113] Corresponding to the above method, an embodiment of the present invention also provides a road network pattern recognition system based on unsupervised graph representation learning, including a spatial dual graph construction module, an adjacency matrix generation module, a feature matrix generation module, a graph-level embedding generation module and a road network pattern recognition module.

[0114] The spatial dual graph construction module is used to take the stroke segments in the target road network data as the nodes of the graph, and the connection relationship between adjacent segments and the spatial relationship added for spatially similar segments as the edges of the graph to construct the spatial dual graph of the target road network data; the adjacency matrix generation module is used to generate the adjacency matrix of the target road network data according to the spatial dual graph; the feature matrix generation module is used to extract the segment features of each stroke segment in the target road network data to generate the feature matrix of the target road network data; wherein the segment features include the geometric attribute features and shape context SC attribute features of the segment; the graph level embedding generation module is used to generate the feature matrix of the target road network data based on the target road network data; The spatial dual graph, adjacency matrix and feature matrix of the marked road network data are used to perform road graph representation learning using a preset graph neural network model to generate a graph-level embedding of the target road network data; wherein, the graph neural network model includes a GAE model and a structure-aware graph representation learning module; the structure-aware graph representation learning module uses the node-level embedding output by the encoder in the GAE model as input to generate a graph-level embedding; the road network pattern recognition module is used to measure the similarity between the graph embedding of the target road network data and the graph embedding of multiple preset standard anchor points, and the road network pattern to which the standard anchor point with the highest similarity value belongs is used as the road network pattern of the original road network data.

[0115] It should be noted that the road network pattern recognition system provided in the embodiment of the present invention is for implementing the above method. Its specific functions can be referred to the above method embodiments and will not be described in detail here.

[0116] In order to verify the effectiveness of the solution of the present invention, the present invention also conducted relevant experiments.

[0117] (1) Experimental data and settings

[0118] (1) Experimental dataset

[0119] The experimental data for road network pattern recognition comes from road networks downloaded from OpenStreetMap (OSM) for ten cities, including Beijing, Xi'an, and Ansbach, Germany. The original road network was preprocessed by eliminating pseudo-nodes and deleting short dangling line segments. Five road network patterns were collected: grid, complex interchange, radial, parallel, and irregular. 220 samples were generated for each pattern, for a total of 1,100 samples. The first four types of road networks all have typical spatial structures and road connectivity, and frequently appear in road networks. All other road network patterns are labeled as irregular. To ensure effective model evaluation on an independent dataset and sufficient training data, a 6:2:2 ratio was used to divide the training, validation, and test sets into training, validation, and test sets.

[0120] (2) Experimental setup

[0121] In the experiment, multiple SUGAR models were constructed by varying the model depth (i.e., the number of convolutional layers), named SUGAR-X, where X represents the model depth. Each convolutional layer contains 64 first-order Chebyshev polynomial kernels. The hidden layer uses the linear rectifier unit function Relu(·), the output layer uses the logistic sigmoid function, and the optimization algorithm uses Adam. During the training process, after multiple tests, hyperparameters such as batch size, learning rate, number of iterations, dropout value, and embedding dimension were set to 32, 0.001, 2000, 0.3, and 128, respectively. The training and validation processes were unsupervised, with annotations only on the test set: y∈{0,1,2,3,4}, representing grid, parallel, complex interchange, radial, and irregular patterns, respectively.

[0122] The SUGAR model is compared with four road network pattern recognition models. First, two rasterization models for complex interchange pattern recognition are compared, using AlexNet and [For each road network test sample, a 224×224 pixel image was generated as input, with a rendering width of 2 pixels. For the AlexNet model, the learning rate, batch size, maximum number of iterations, and dropout value were set to 0.02, 16, 10,000, and 0.5, respectively. For the GoogLeNet model, these parameters were set to 0.02, 16, 20,000, and 0.8. A multi-layer perceptron (MLP) model was also compared. The MLP model consists of a hidden layer with 64 neurons and uses ReLU (·) as the activation function. Finally, the proposed state-of-the-art model GCNN-4 for grid pattern recognition and complex overpass pattern recognition is compared. This model consists of one input layer, four graph convolutional layers, and one output layer. It uses third-order Chebyshev polynomial convolution kernels, with each convolutional layer containing 64 kernels and a maximum number of iterations of 2,000. All of these models were trained on a random selection of 10% of the node labels.

[0123] The proposed SUGAR was also compared with three other GAE models with ablation. The first model, GCNN_GAE-3, omitted SIC and GCA, using only the basic GCNN convolutional layers. The second model, GCA_GAE-3, used only GCA. The third model, SIC_GAE-3, used only SIC. All three models achieved the best pattern recognition results using the three convolutional layers, with all other parameters identical to those of the SUGAR model.

[0124] (3) Evaluation indicators

[0125] The experiment uses classification accuracy, which is the ratio of correctly classified samples to the total number of samples, to evaluate pattern classification results. This is the most direct evaluation metric and calculates the ratio of correctly classified road network patterns to the total number of road network samples.

[0126]

[0127] In addition, three quantitative indicators are used: precision, recall and F1-score. For each pattern type result, the three indicators are calculated as follows:

[0128]

[0129] For each pattern category, precision is calculated as the ratio of the number of networks correctly predicted to be of that pattern category to the total number of networks predicted to be of that pattern category, while recall is calculated as the ratio of the number of networks correctly predicted to be of that pattern category to the total number of networks actually belonging to that pattern category. TP, FP, and FN represent the number of true positives, false positives, and false negatives, respectively. Taking the results for the grid pattern as an example, true positives refer to network samples that have been manually labeled and predicted to be of the grid pattern type. Network samples predicted to be of the grid pattern but manually labeled as another type are considered false positives. Network samples manually labeled as the grid pattern but predicted to be of another pattern type are considered false negatives. The F1-score is the harmonic mean of precision and recall, a metric that comprehensively considers both precision and recall. The F1-score can be calculated for each category, or the total F1-score can be calculated as an overall performance evaluation.

[0130] (2) Experimental results and analysis

[0131] Table 2 lists the pattern classification accuracy of SUGAR models with different depths and 7 comparison models in the test sample set.

[0132] Table 2 Classification accuracy of different models for test sample sets

[0133]

[0134] As can be seen, the SUGAR model's test accuracy ranges from 84.55% to 93.18%, significantly outperforming the AlexNet and GoogLeNet models. This is because the SUGAR model considers the connectivity between road segments, while the grid-based processing makes it difficult for AlexNet and GoogLeNet models to directly utilize this structural information, affecting recognition performance. Furthermore, the SUGAR model outperforms the MLP model because it effectively aggregates information about nodes and their neighbors through graph convolution, capturing both local and global structural features between nodes. The MLP model, however, only processes data based on the independent features of each node and performs poorly on graph data. The state-of-the-art GCNN-4 model achieved an accuracy of 89.55% on our test set, while the unsupervised GCNN_GAE-3 model achieved an accuracy of 85.45%, a decrease of 4.10%. This decrease is due to the GCNN_GAE-3 model's inability to directly utilize labels for optimization, as is the case with supervised learning. However, its performance still demonstrates the potential of unsupervised learning for road network pattern recognition.

[0135] Comparing SUGAR models of varying depths, classification accuracy initially increases and then decreases with depth. This phenomenon may be due to two factors: first, due to limited sample size, deep models are prone to overfitting, resulting in poor performance on the test set; second, the continuous multiplication effect during backpropagation can cause gradients in shallow convolutional layers to vanish or explode, hindering model learning. Therefore, although the SUGAR model has higher overall accuracy than the GCNN-4 model, its test accuracy is lower when the number of convolutional layers is one or more than four.

[0136] Furthermore, with the same number of convolutional layers, the SUGAR model outperformed the GCNN_GAE, GCA_GAE, and SIC_GAE models, further demonstrating the importance of SIC and GCA. SIC enhances the model's ability to capture local structural features, while GCA improves the model's understanding of overall topological relationships. Specifically, SIC provides rich topological information for each graph node by identifying and counting subgraph tracks, helping the model more accurately describe and distinguish road network patterns. GCA dynamically adjusts the importance of nodes, ensuring that the model focuses on the areas most relevant to the task, thereby improving classification and recognition accuracy.

[0137] As shown in Table 2, the SUGAR-3 model achieved the highest classification accuracy, reaching 93.18%. Therefore, we selected it as the baseline model for road network pattern recognition. Table 3 shows the confusion matrix and evaluation metrics for the recognition results on the test set. The grid, complex interchange, and parallel patterns performed relatively well in terms of precision and recall, achieving 93.62% and 100%, 97.62% and 93.18%, and 97.67% and 95.46%, respectively. This may be due to the unique spatial structures of these three pattern types, such as the four-way perpendicular orthogonal distribution of the grid pattern, the two-way continuous distribution of the parallel pattern, and the central convergence distribution of the complex intersection, making these patterns easier to distinguish. However, some misidentifications occurred, likely due to confusion caused by structural similarities between different patterns. For example, the diversity and complexity of the irregular pattern can easily lead to confusion with other patterns based on certain features.

[0138] Table 3 Confusion matrix and classification results evaluation of the test dataset using the SUGAR-3 model

[0139]

[0140] Figure 10The pattern recognition results of some correctly identified road network samples and all incorrectly identified samples are shown. Overall, the model's recognition effect is good, but there are errors in the recognition of some patterns. Specifically, the model mistakenly identifies some parallel patterns (20, 21) and irregular patterns (30) as grid patterns. This may be because there are some small burrs when the road network samples are intercepted, which makes the samples have certain grid characteristics, resulting in misjudgment. In addition, the model mistakenly identifies multiple irregular patterns (32, 33, 34, 35) as radial patterns. This is because irregular patterns are mostly distributed in a divergent manner, and some samples may have characteristics similar to radial patterns. In some cases, the model identifies irregular pattern (30) as a grid pattern and irregular pattern (31) as a complex interchange pattern. This is because these samples have certain grid characteristics and complex interchange characteristics, respectively, which interfere with the model's judgment. It is worth noting that due to the subjective intention of the road network pattern, some ambiguous situations will occur during recognition. For example, although samples 13, 14, 26, 27, and 28 are classified as irregular patterns, they also have some characteristics of complex interchange patterns and radial patterns. Therefore, the similarity between these patterns and these patterns is high, but because it does not exceed the threshold, it is ultimately misclassified as an irregular pattern. A similar situation occurs in sample 12, which has characteristics of both a complex interchange pattern and a radial pattern. Therefore, the similarity between these two patterns is very high during recognition, resulting in a misclassification. The results show that the model is more likely to identify radial patterns as irregular patterns. This may be because the radial pattern samples themselves have some characteristics of irregular patterns. When the radial centers are not concentrated at a single point, the model has difficulty in correctly classifying them. Grid patterns are difficult to misclassify because their characteristics are obvious and very different from radial and irregular patterns. Their vertical orthogonality is also easy to extract through the direction of the road segment.

[0141] This paper proposes a recognition method for road network pattern recognition based on geometric similarity graph representation learning. This method utilizes spatial dual graph modeling, cognitively inspired graph node feature extraction, the introduction of subgraph isomorphism counting (SIC) during the segment embedding learning phase, and the global contextual attention (GCA) mechanism during the graph embedding generation phase. This method establishes a mapping between the input road graph and the output road network pattern, enhancing the model's representation of the road network and, consequently, improving road network pattern recognition. Experiments on road network pattern recognition using real-world data demonstrate that the proposed method achieves a classification accuracy of 93.18%, surpassing grid-based AlexNet and GoogLeNet models, as well as supervised neural network-based MLP and GCNN models. Under the same parameter settings, it also outperforms ablation models such as GCNN_GAE, SIC_GAE, and GCA_GAE. Finally, comparative analysis of road network pattern samples and graph-level embedding representations reveals that the SUGAR model effectively clusters the generated graph embeddings for the same pattern, with clear boundaries between different clusters. This validates the effectiveness of SIC and GCA in improving recognition performance.

[0142] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A road network pattern recognition method based on unsupervised graph representation learning, characterized in that: include: Step 1: Take the stroke segments in the target road network data as the nodes of the graph, and the connection relationship between adjacent segments and the spatial relationship added for spatially similar segments as the edges of the graph to construct the spatial dual graph of the target road network data, and generate the adjacency matrix of the target road network data based on the spatial dual graph; Step 2: Extract the segment features of each stroke segment in the target road network data to generate a feature matrix of the target road network data; wherein the segment features include geometric attribute features and shape context SC attribute features of the segment; Step 3: Based on the spatial dual graph, adjacency matrix, and feature matrix of the target road network data, a preset graph neural network model is used to perform road graph representation learning to generate a graph-level embedding for the target road network data. The graph neural network model includes a GAE model and a structure-aware graph representation learning module. The structure-aware graph representation learning module uses the node-level embedding output by the encoder in the GAE model as input to generate a graph-level embedding. Step 4: Measure the similarity between the graph embedding of the target road network data and the graph embedding of multiple preset standard anchor points, and take the road network pattern of the standard anchor point with the highest similarity value as the road network pattern of the original road network data.

2. A road network pattern recognition method based on unsupervised graph representation learning according to claim 1, characterized in that: Step 1 also includes: extracting spatially similar road sections and adding spatial relationships for the spatially similar road sections; wherein, the spatially similar road sections refer to two road sections that meet preset spatial requirements but have no connection relationship; the preset spatial requirements refer to that the distance similarity and angle similarity of the two road sections are both greater than 0.

3. A road network pattern recognition method based on unsupervised graph representation learning according to claim 2, characterized in that: Calculate the distance similarity between two road segments according to the following formula: and angle similarity Among them, sp_dist(s i ,s j ) indicates calculating two road segments s i and s j One-way Hausdorff distance between, ag_dist(s i ,s j ) indicates calculating two road segments s i and s j The absolute angular distance between ds and δ as Indicates two preset distance thresholds.

4. The road network pattern recognition method based on unsupervised graph representation learning according to claim 1, characterized in that: The geometric attribute features of the road section include position features extracted from the macro scale, connection features extracted from the meso scale, and length features, shape features and angle features extracted from the micro scale.

5. The road network pattern recognition method based on unsupervised graph representation learning according to claim 1, characterized in that: The process of extracting the SC attribute features of the road section includes: Establish a polar coordinate system at the midpoint of the target road section, and set the maximum radius of the polar coordinate system to r = γ × L(s); where L(s) is the length of the target road section and γ is the scale factor; Divide the polar coordinate system into p concentric circles at equal distances and q sectors at equal angles to obtain p×q sectors, and define the value of each sector as the ratio of the total length of the road segments in the region to the total length of the road segments in the context range; The values ​​of p×q sectors are expressed as a vector, which is the SC attribute feature of the target road section.

6. The road network pattern recognition method based on unsupervised graph representation learning according to claim 1, characterized in that: A subgraph isomorphism counting (SIC) module is provided in the encoder, and the SIC module is used to generate structure identifiers for nodes in the space dual graph. The encoder performs neighborhood information transfer based on the structure identifier, adjacency matrix and feature matrix of the node to generate node-level embedding.

7. The road network pattern recognition method based on unsupervised graph representation learning according to claim 1, characterized in that: The method further includes: training the encoder and decoder in the GAE model; during the training process, the decoder reconstructs the inter-node association based on the node-level embedding output by the encoder to generate a reconstructed adjacency matrix, and uses the following cross-entropy loss function to calculate the distance between the reconstructed adjacency matrix A′ and the true adjacency matrix A: Where N represents the total number of nodes in the space-dual graph.

8. The road network pattern recognition method based on unsupervised graph representation learning according to claim 6, characterized in that: The process of generating node-level embeddings specifically includes: Define a subgraph set For subgraph collections For each subgraph H in the space dual graph G, first find its isomorphic subgraph G in the space dual graph G S , for G S Each node in Find the orbital Orb of its mapping f(v) in the subgraph H H (f(v)), Orb H (*) represents the preset orbital mapping function, where orbit refers to the position or role of a node in H; then, the position of each node v in different isomorphic subgraphs G is calculated. S The track times, and obtain the structural characteristics of node v on the i-th track of subgraph H represents the i-th track in the subgraph H; f(v) represents the node mapping function that maps node v to the node in the subgraph H; finally, the combined subgraph set Count the different subgraph structures and orbits in the node to get the structure identifier of the node in, is subgraph H i the number of mid-tracks; The adjacency matrix, feature matrix, and structure identifier are updated according to the following propagation rules between adjacent convolutional layers: in, A is the adjacency matrix, is the identity matrix, yes degree matrix of ; is the convolution signal matrix of the lth layer, representing the features of each node in the lth layer, m l is the dimension of the node feature at layer l; the input layer H 0 =M, Represents the feature matrix, d1 represents the dimension of the node's geometric attribute feature, d2 represents the dimension of the node's SC attribute feature; X v is the structure identifier of the node, and is the parameter training weight matrix and common bias of layer l, σ(·) is the nonlinear activation function, and N represents the total number of nodes in the space-dual graph G.

9. The road network pattern recognition method based on unsupervised graph representation learning according to claim 1, characterized in that: The structure-aware graph representation learning module takes the node-level embedding output by the encoder in the GAE model as input to generate graph-level embeddings, specifically including: Among them, E S represents the graph-level embedding, Z represents the node-level embedding output by the encoder, is the embedding of the nth node, N represents the total number of nodes; is a learnable weight matrix, and tanh(·) is the activation function; represents the attention weight of the node, and σ(·) is a nonlinear activation function.

10. A road network pattern recognition system based on unsupervised graph representation learning, characterized in that: include: The spatial dual graph construction module is used to construct the spatial dual graph of the target road network data by using the stroke road segments in the target road network data as the nodes of the graph and the connection relationships between adjacent road segments and the spatial relationships added for spatially similar road segments as the edges of the graph; An adjacency matrix generation module, configured to generate an adjacency matrix of target road network data according to the space dual graph; A feature matrix generation module is used to extract the segment features of each stroke segment in the target road network data to generate a feature matrix of the target road network data; wherein the segment features include geometric attribute features and shape context SC attribute features of the segment; A graph-level embedding generation module is configured to generate a graph-level embedding for the target road network data by learning a road graph representation based on the spatial dual graph, adjacency matrix, and feature matrix of the target road network data using a preset graph neural network model; wherein the graph neural network model includes a GAE model and a structure-aware graph representation learning module; the structure-aware graph representation learning module uses the node-level embedding output by the encoder in the GAE model as input to generate a graph-level embedding; The road network pattern recognition module is used to measure the similarity between the graph embedding of the target road network data and the graph embedding of multiple preset standard anchor points, and take the road network pattern of the standard anchor point with the highest similarity value as the road network pattern of the original road network data.

Images