Vehicle trajectory audit analysis system and method based on spatiotemporal feature clustering

Abstract

The present application relates to the field of intelligent transportation technology, and discloses a vehicle travel trajectory audit and analysis system and method based on spatiotemporal feature clustering. The system includes: a topology construction module, configured to: construct a highway network toll attribute graph based on node and edge sets to form a topological network of node reachability relationships; a feature processing module, configured to: standardize the spatiotemporal features of historical travel trajectories to eliminate dimensional differences; a clustering analysis module, configured to: perform adaptive clustering based on standardized spatiotemporal features to determine the division of time clusters and space clusters; a trajectory reasoning module, configured to: generate candidate paths based on spatiotemporal cluster features, calculate comprehensive scores through spatiotemporal similarity calculation and weight distribution, and output the optimal reasoning path. This method corresponds to the system. The adoption of this application has greatly improved the efficiency of network toll auditing work and effectively maintained the national network toll order.

Description

Technical Field

[0001] The present application relates to the field of intelligent transportation technology, and specifically to a vehicle travel trajectory audit and analysis system and method based on spatiotemporal feature clustering. Background Art

[0002] Traditional highway network toll collection and audit management has significant limitations. Traditional data analysis methods and reliance on manual verification are not only inefficient but also prone to missing evasion behaviors and susceptible to subjective factors. Although automated audit models are currently in use, most of these models are based on explicit rule-based engines. Data analysis methods focus solely on comparing surface features and lack the ability to effectively and deeply mine implicit information such as inherent patterns and correlations in the data. Therefore, they struggle to address the complex and ever-changing nature of evasion behaviors. Furthermore, data limitations and untimely updates further limit the ability to identify evasion patterns.

[0003] Therefore, in order to adapt to the new situation of nationwide networked toll collection operations on expressways, it is urgent to develop an auditing technology that can integrate multi-source data, is more intelligent and adapts to business development. Summary of the Invention

[0004] The purpose of this application is to provide a vehicle traffic trajectory audit and analysis system and method based on spatiotemporal feature clustering to solve the technical problems raised in the above-mentioned background technology, greatly improve the efficiency of networked toll auditing work, and effectively maintain the national networked toll collection order.

[0005] To achieve the above objectives, this application discloses the following technical solutions:

[0006] In a first aspect, the present application discloses a vehicle trajectory audit and analysis system based on spatiotemporal feature clustering, comprising:

[0007] The topology building module is configured to: construct a highway network toll attribute graph based on a set of nodes and edges, forming a topological network of node reachability relationships;

[0008] The feature processing module is configured to: standardize the spatiotemporal features of historical traffic trajectories to eliminate dimensional differences;

[0009] The cluster analysis module is configured to: perform adaptive clustering based on the standardized spatiotemporal features to determine the temporal and spatial cluster divisions;

[0010] The trajectory inference module is configured to: generate candidate paths by combining spatiotemporal cluster features, calculate comprehensive scores through spatiotemporal similarity calculation and weight allocation, and output the optimal inference path.

[0011] Preferably, the node and edge sets include a node set and a directed edge set;

[0012] The node set is ,in, is the toll station node set, is the gantry node set;

[0013] The directed edge set includes the edge from the toll booth to the gantry , door frame to door frame edge and the gantry to the side of the toll booth , through the adjacency matrix Represents the reachability relationship between nodes, and the matrix elements are 0 or 1 to indicate whether they are connected.

[0014] Preferably, the adjacent matrix Includes toll booth to toll booth relationship blocks , Toll station to gantry relationship block , gantry to toll station relationship block and gantry to gantry relationship blocks ; Among them, the toll station to toll station relationship block For elements all 0 Matrix, the toll booth to gantry relationship block The matrix dimensions are , the gantry to the toll station relationship block The matrix dimensions are , the gantry to gantry relationship block The matrix dimensions are .

[0015] Preferably, the feature processing module is further configured to construct a vehicle traffic record matrix based on historical traffic trajectories, specifically including:

[0016] Extract historical travel trajectory sequence , where each complete passage trajectory flow data for , For the entrance water, , Water flows for the door frame, , For export water, ;

[0017] The trajectory information is converted into a directed edge matrix to complete the conversion of vehicle traffic information from cross-section information to interval traffic status information. The expression of the directed edge matrix is:

[0018]

[0019] in, , is the pass mark of the nth directed edge of the i-th flow, is the directed edge identifier, is the entry time, For departure time, For travel time, For the mileage, is the average speed of traffic.

[0020] Preferably, the spatiotemporal features include time features and space features;

[0021] The time characteristics include travel time, peak traffic period ratio and average speed, and the time characteristics are normalized using Z-score;

[0022] The spatial features include mileage, number of lanes, node degree, longitude and latitude, traffic density and adjacent edge correlation, and the spatial features are normalized using a hybrid of Min-Max normalization and one-hot encoding.

[0023] Preferably, the time feature vector is ,in, is the average travel time, is the travel time variance, is the coefficient of variation of travel time, is the proportion of traffic flow during peak hours, is the average speed over the entire journey, is the maximum travel time, The minimum travel time.

[0024] As an advantage, the cluster analysis module is specifically configured as follows: dynamically determine the optimal number of clusters through the elbow rule and GapStatistic, calculate the Euclidean distance and iteratively update the cluster center, and output the time cluster division respectively. and spatial clustering .

[0025] Preferably, the generating of candidate paths by combining spatiotemporal cluster features specifically includes:

[0026] Missing type definition and preprocessing: The missing position of the flow data is defined as the inlet flow missing scenario, the outlet flow missing scenario and the gantry flow missing scenario. The expression corresponding to the inlet flow missing scenario is: , the expression corresponding to the outlet water loss scenario is , the expression corresponding to the gantry water flow missing scenario is ;

[0027] Generate candidate path set: If the entrance flow is missing, extract the first known portal , reversely retrieve all possible entry candidate sets , forming a candidate path set ,in, , ; If the outlet flow is missing, extract the known gantry at the end , forward search for all possible exit nodes , forming a candidate path set ,in, , ; If the gantry sequence is missing, extract the missing starting endpoint ,from Forward search All possible nodes in the interval, where the set of directed edges during the period is , ; Output candidate path set .

[0028] Preferably, the calculating of the comprehensive score and outputting the optimal reasoning path specifically includes:

[0029] Feature alignment: For each candidate path , extracting time features and spatial characteristics , aligned with the historical trajectory feature results, where , ;

[0030] Trajectory spatiotemporal similarity calculation: Calculate the distance between the time feature of each edge in the path and the center of its time cluster to obtain the temporal similarity , the expression of time similarity is:

[0031]

[0032] in, Candidate path edge The time feature vector of For the historical trajectory Time cluster The central eigenvector, is the standard deviation of historical time characteristics;

[0033] Calculate the distance between the spatial features of each edge in the path and the center of the spatial cluster to which it belongs, and calculate the spatial similarity , the expression of spatial similarity is:

[0034]

[0035] in, Candidate path edge The spatial eigenvectors of For the historical trajectory Space cluster The central eigenvector, is the standard deviation of spatial characteristics;

[0036] Comprehensive score calculation: Combine time and space similarity to calculate the score and calculate the comprehensive score , the calculation formula is:

[0037]

[0038] in, For the target vehicle on the directed edge The weighted high-frequency edge weights that appear in the historical trajectory, is the time similarity weight coefficient, is the spatial similarity weight coefficient;

[0039] Output the optimal reasoning path: Output the top-K paths in descending order according to the comprehensive score.

[0040] In a second aspect, the present application discloses a vehicle trajectory audit and analysis method based on spatiotemporal feature clustering, which uses the vehicle trajectory audit and analysis system based on spatiotemporal feature clustering as described above. The method includes the following steps:

[0041] Construct a highway toll collection attribute graph based on node and edge sets to form a topological network of node reachability relationships;

[0042] Standardize the temporal and spatial characteristics of historical travel trajectories;

[0043] Adaptive clustering is performed based on the standardized spatiotemporal features to obtain temporal and spatial clusters;

[0044] Candidate paths are generated by combining spatiotemporal cluster features, and a comprehensive score is calculated through spatiotemporal similarity calculation and weight distribution to output the optimal reasoning path.

[0045] Beneficial effects: The vehicle traffic trajectory audit and analysis system and method based on spatiotemporal feature clustering of the present application lays the foundation for the rationality of the accessible paths for vehicles by constructing a networked highway toll collection attribute map; by extracting and clustering the spatiotemporal data features of historical data, the sensitivity of vehicle traffic features is effectively increased, and the noise interference of data analysis is significantly reduced; by utilizing temporal local feature clustering and spatial local feature clustering, the complexity of data retrieval is greatly reduced, and the accuracy of candidate paths is improved when generating candidate paths for selection and screening; by using a weight calculation that balances historical behavior matching and path rationality, combined with a scoring function and dynamic constraints, the path is quantitatively completed through multiple indicators, and the effective promotion and restoration of missing paths in the feature flow is realized. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0047] Figure 1 This is a structural block diagram of the vehicle traffic trajectory audit and analysis system based on spatiotemporal feature clustering provided in an embodiment of the present application. DETAILED DESCRIPTION

[0048] The following is a clear and complete description of the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0049] In this document, the term "comprising" is intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising..." does not preclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0050] In the first aspect, this embodiment discloses a method such as Figure 1 The vehicle traffic trajectory audit and analysis system based on spatiotemporal feature clustering shown includes a topology construction module, a feature processing module, a cluster analysis module and a trajectory reasoning module.

[0051] Specifically

[0052] The topology building module is configured to: build a highway network toll attribute graph based on a set of nodes and edges to form a topological network of node reachability relationships.

[0053] The feature processing module is configured to: standardize the spatiotemporal features of historical travel trajectories to eliminate dimensional differences.

[0054] The cluster analysis module is configured to perform adaptive clustering based on the standardized spatiotemporal features to determine the temporal cluster and spatial cluster divisions.

[0055] The trajectory inference module is configured to: generate candidate paths by combining spatiotemporal cluster features, calculate comprehensive scores through spatiotemporal similarity calculation and weight allocation, and output the optimal inference path.

[0056] In one embodiment, the node and edge sets include a node set and a directed edge set.

[0057] The node set is ,in, is the toll station node set, It is a gantry node set. The node attribute definition includes but is not limited to: node number, name, belonging area, use status, toll station type, main line pile number, up sign, up sign, provincial boundary sign, latitude and longitude, etc. The toll station node set is , gantry node set: .

[0058] The directed edge set includes the edge from the toll booth to the gantry , door frame to door frame edge and the gantry to the side of the toll booth , through the adjacency matrix Represents the reachability relationship between nodes, and the matrix elements are 0 or 1 to indicate whether they are connected. , , , the directed edge set is .

[0059] The adjacent matrix Includes toll booth to toll booth relationship blocks , Toll station to gantry relationship block , gantry to toll station relationship block and gantry to gantry relationship blocks ; Among them, the toll station to toll station relationship block For elements all 0 Matrix, the toll booth to gantry relationship block The matrix dimensions are , the gantry to the toll station relationship block The matrix dimensions are , the gantry to gantry relationship block The matrix dimensions are .

[0060] Adjacency Matrix The relevant introduction is as follows:

[0061]

[0062] : Matrix, edges from toll booth to toll booth (no connection, all 0).

[0063] : Matrix, representing the connection relationship between the toll booth and the gantry.

[0064]

[0065] : Matrix, representing the connection relationship from the gantry to the toll station.

[0066]

[0067] : Matrix, representing the gantry-to-gantry edges.

[0068]

[0069] In one embodiment, the feature processing module is further configured to construct a vehicle traffic record matrix based on historical traffic trajectories, specifically including:

[0070] Extract historical travel trajectory sequence , where each complete passage trajectory flow data for , For the entrance water, , Water flows for the door frame, , For export water, ;

[0071] The trajectory information is converted into a directed edge matrix to complete the conversion of vehicle traffic information from cross-section information to interval traffic status information. The expression of the directed edge matrix is:

[0072]

[0073] in, , is the pass mark of the nth directed edge of the i-th flow, is the directed edge identifier, is the entry time, For departure time, For travel time, For the mileage, is the average speed of traffic.

[0074] In one embodiment, the spatiotemporal features include time features and space features;

[0075] The time characteristics include travel time, peak traffic period ratio and average speed, and the time characteristics are normalized using Z-score;

[0076] The spatial features include mileage, number of lanes, node degree, longitude and latitude, traffic density and adjacent edge correlation, and the spatial features are normalized using a hybrid of Min-Max normalization and one-hot encoding.

[0077] Furthermore, the cluster analysis module is specifically configured as follows: dynamically determining the optimal number of clusters through the elbow rule and GapStatistic, calculating the Euclidean distance and iteratively updating the cluster center, and outputting the time cluster division respectively. and spatial clustering .

[0078] Specifically, temporal local feature clustering analysis includes:

[0079] 1. Temporal feature standardization and enhanced extraction: Based on the directed edge matrix of vehicle traffic flow , extract each directed edge , extract the local features of time, perform Z-score normalization, eliminate the dimension difference, and form the feature vector The time feature vector is , is the maximum travel time, The minimum travel time.

[0080] in

[0081] It is the average travel time, which is used to reflect the overall travel efficiency and traffic congestion of the directional side road segment.

[0082]

[0083] n: the number of vehicles passing a directed edge e during the observation period, : The actual time taken by the i-th vehicle to travel on this directed edge.

[0084] is the travel time variance, which is used to measure the volatility of travel time. A larger value indicates a more unstable travel time.

[0085]

[0086] is the coefficient of variation of travel time, which is dimensionless by comparing the standard deviation to the mean, making it easier to compare across directed edges.

[0087]

[0088] It is the proportion of traffic flow during peak hours.

[0089]

[0090] is the peak time distribution set for edge e, The travel time of the i-th vehicle belongs to the peak time, Used to indicate the function to count the total number of vehicles passing through during peak hours. is the total traffic volume.

[0091] The average speed for the entire journey.

[0092]

[0093] is the mileage to edge e, is the average travel time to edge e.

[0094] 2. Adaptive time cluster number determination and clustering:

[0095] (1) Input parameters: directed edge set E, feature matrix , the number of classification clusters K, the convergence threshold ϵ.

[0096] (2) Algorithm steps:

[0097] Initialization: Randomly select K initial cluster centers ;

[0098] Assign clusters: For each edge , calculate its Euclidean distance to each center;

[0099]

[0100] in, is the normalized eigenvalue of the directed edge e, is the center value of the K-th cluster, is the cluster with the smallest Euclidean distance of directed edge e.

[0101] Update Center: Recalculate cluster means:

[0102]

[0103] in, is the number of edge sets of cluster k in the tth iteration, It is the center point of cluster K calculated at the t+1th iteration.

[0104] Iteration: Repeat the steps of assigning clusters and updating centers until:

[0105]

[0106] in, To set the threshold, is the center point of cluster K in this iteration calculation, is the center point of cluster K calculated in the last iteration.

[0107] (3) Output results: time cluster division

[0108] Spatial local feature cluster analysis, including:

[0109] 1. Spatial feature standardization and extraction: Based on the attribute graph, extract each directed edge , extract local features of the space, adopt a hybrid normalization strategy, and form a feature vector ;

[0110]

[0111] in, is the directed edge mileage after Min-Max normalization; is the number of lanes after Min-Max normalization; is the directed edge type (0 is from toll station to gantry, 1 is from gantry to gantry, 2 is from gantry to exit toll station), which is a one-hot encoding; and are the out-degree and in-degree (the number of other connected edges) of the start and end nodes after Min-Max normalization; and is the longitude and latitude of the midpoint of the standardized edge; is the standardized daily average traffic density; is the normalized correlation between adjacent edges (based on the historical conversion probability).

[0112] 2. Adaptive spatial cluster number determination and clustering

[0113] (1) Input parameters: directed edge set E, feature matrix , the number of classification clusters Y, the convergence threshold ϵ;

[0114] (2) Processing using algorithm steps such as temporal local feature clustering analysis;

[0115] (3) Output: Spatial cluster division .

[0116] In one embodiment, the generation of candidate paths by combining spatiotemporal cluster features specifically includes:

[0117] Missing type definition and preprocessing: The missing position of the flow data is defined as the inlet flow missing scenario, the outlet flow missing scenario and the gantry flow missing scenario. The expression corresponding to the inlet flow missing scenario is: , the expression corresponding to the outlet water loss scenario is , the expression corresponding to the gantry water flow missing scenario is ;

[0118] Generate candidate path set: Based on the topological structure of the attribute graph M=(U,A,E) and spatiotemporal constraints (time and distance thresholds), generate a candidate path set. If the entrance flow is missing, extract the first known portal , reversely retrieve all possible entry candidate sets , forming a candidate path set ,in, , ; If the outlet flow is missing, extract the known gantry at the end , forward search for all possible exit nodes , forming a candidate path set ,in, , ; If the gantry sequence is missing, extract the missing starting endpoint ,from Forward search All possible nodes in the interval, where the set of directed edges during the period is , ; Output candidate path set .

[0119] Furthermore, the calculation of the comprehensive score and output of the optimal reasoning path specifically includes:

[0120] Feature alignment: For each candidate path , extracting time features and spatial characteristics , aligned with the historical trajectory feature results, where , ;

[0121] Trajectory spatiotemporal similarity calculation: Calculate the distance between the time feature of each edge in the path and the center of its time cluster to obtain the temporal similarity , the expression of time similarity is:

[0122]

[0123] in, Candidate path edge The time feature vector of For the historical trajectory Time cluster The central eigenvector, is the standard deviation of historical time characteristics;

[0124] Calculate the distance between the spatial features of each edge in the path and the center of the spatial cluster to which it belongs, and calculate the spatial similarity , the expression of spatial similarity is:

[0125]

[0126] in, Candidate path edge The spatial eigenvectors of For the historical trajectory Space cluster The central eigenvector, is the standard deviation of spatial characteristics;

[0127] Comprehensive score calculation: Combine time and space similarity to calculate the score and calculate the comprehensive score , the calculation formula is:

[0128]

[0129] in, For the target vehicle on the directed edge The weighted high-frequency edge weights that appear in the historical trajectory, is the time similarity weight coefficient, is the spatial similarity weight coefficient;

[0130] Output the optimal reasoning path: Output the Top-K paths in descending order according to the comprehensive score, where Top-K is the comprehensive score The first K pushed paths are used as the scope of key trajectory audit for target vehicles with missing traffic information

[0131] Based on the above, the vehicle traffic trajectory audit and analysis system based on spatiotemporal feature clustering of this embodiment lays the foundation for the rationality of the accessible paths for vehicles by constructing a networked highway toll attribute map; by extracting and clustering the spatiotemporal data features of historical data, it effectively increases the sensitivity of vehicle traffic features and significantly reduces the noise interference of data analysis; by utilizing temporal local feature cluster division and spatial local feature cluster division, when generating candidate paths for selection and screening, it greatly reduces the complexity of data retrieval and improves the accuracy of candidate paths; by using a weight calculation that balances historical behavior matching and path rationality, combined with scoring functions and dynamic constraints, the path is quantitatively completed through multiple indicators, and the effective promotion and restoration of missing paths in the feature flow is realized.

[0132] In a second aspect, this embodiment provides a vehicle trajectory audit and analysis method based on spatiotemporal feature clustering, which is applied to the vehicle trajectory audit and analysis system based on spatiotemporal feature clustering as described above. The method includes the following steps:

[0133] Construct a highway toll collection attribute graph based on node and edge sets to form a topological network of node reachability relationships;

[0134] Standardize the temporal and spatial characteristics of historical travel trajectories;

[0135] Adaptive clustering is performed based on the standardized spatiotemporal features to obtain temporal and spatial clusters;

[0136] Candidate paths are generated by combining spatiotemporal cluster features, and a comprehensive score is calculated through spatiotemporal similarity calculation and weight distribution to output the optimal reasoning path.

[0137] It should be noted that the vehicle traffic trajectory audit and analysis method based on spatiotemporal feature clustering of this embodiment corresponds to the aforementioned vehicle traffic trajectory audit and analysis system based on spatiotemporal feature clustering. Therefore, the parts that are not described in detail in the vehicle traffic trajectory audit and analysis method based on spatiotemporal feature clustering of this embodiment (including but not limited to specific technical means, technical effects, etc.) can refer to the relevant description in the aforementioned vehicle traffic trajectory audit and analysis system based on spatiotemporal feature clustering, and this text will not go into details here.

[0138] In the embodiments provided herein, it should be understood that the embodiments described herein can be implemented using hardware, software, firmware, middleware, code, or any appropriate combination thereof. For hardware implementation, the processor may be implemented in one or more of the following: an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field-programmable gate array (FPGA), a processor, a controller, a microcontroller, a microprocessor, or other electronic units designed to implement the functionality described herein, or any combination thereof. For software implementation, some or all of the processes of the embodiments may be performed by a computer program instructing the relevant hardware. During implementation, the program may be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein communication media includes any medium that facilitates the transmission of a computer program from one location to another. The storage medium may be any available medium that can be accessed by a computer. Computer-readable storage media may include, but are not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing the desired program code in the form of instructions or data structures and accessible by a computer.

[0139] Finally, it should be noted that the above is only a preferred embodiment of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or make equivalent replacements for some of the technical features therein. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. A vehicle trajectory audit and analysis system based on spatiotemporal feature clustering, characterized by: include: The topology building module is configured to: construct a highway network toll attribute graph based on a set of nodes and edges, forming a topological network of node reachability relationships; The feature processing module is configured to: standardize the spatiotemporal features of historical traffic trajectories to eliminate dimensional differences; The cluster analysis module is configured to: perform adaptive clustering based on the standardized spatiotemporal features to determine the temporal and spatial cluster divisions; The trajectory inference module is configured to: generate candidate paths by combining spatiotemporal cluster features, calculate comprehensive scores through spatiotemporal similarity calculation and weight allocation, and output the optimal inference path; The calculation of the comprehensive score and output of the optimal reasoning path specifically includes: Feature alignment: For each candidate path , extracting time features and spatial characteristics , aligned with the historical trajectory feature results, where , , is the average travel time, is the travel time variance, is the coefficient of variation of travel time, is the proportion of traffic flow during peak hours, is the average speed over the entire journey, is the maximum travel time, is the minimum travel time, is the directed edge mileage after Min-Max normalization; is the number of lanes after Min-Max normalization; is the directed edge type; and are the out-degree and in-degree of the starting and ending nodes after Min-Max normalization; and is the longitude and latitude of the midpoint of the standardized edge; is the standardized daily average traffic density; To normalize the correlation between adjacent edges, is a directed edge in the candidate path; Trajectory spatiotemporal similarity calculation: Calculate the distance between the time feature of each edge in the path and the center of its time cluster to obtain the temporal similarity , the expression of time similarity is: in, Candidate path edge The time feature vector of For the historical trajectory Time cluster The central eigenvector, is the standard deviation of historical time characteristics; Calculate the distance between the spatial features of each edge in the path and the center of the spatial cluster to which it belongs, and calculate the spatial similarity , the expression of spatial similarity is: in, Candidate path edge The spatial eigenvectors of For the historical trajectory Space cluster The central eigenvector, is the standard deviation of spatial characteristics; Comprehensive score calculation: Combine time and space similarity to calculate the score and calculate the comprehensive score , the calculation formula is: in, For the target vehicle on the directed edge The weighted high-frequency edge weights that appear in the historical trajectory, is the time similarity weight coefficient, is the spatial similarity weight coefficient; Output the optimal reasoning path: Output the top-K paths in descending order according to the comprehensive score.

2. The vehicle traffic trajectory audit and analysis system based on spatiotemporal feature clustering according to claim 1 is characterized in that: The node and edge sets include a node set and a directed edge set; The node set is ,in, is the toll station node set, is the gantry node set; The directed edge set includes the edge from the toll booth to the gantry , door frame to door frame edge and the gantry to the side of the toll booth , through the adjacency matrix Represents the reachability relationship between nodes, and the matrix elements are 0 or 1 to indicate whether they are connected.

3. The vehicle traffic trajectory audit and analysis system based on spatiotemporal feature clustering according to claim 2 is characterized in that: The adjacent matrix Includes toll booth to toll booth relationship blocks , Toll station to gantry relationship block , gantry to toll station relationship block and gantry to gantry relationship blocks ; Among them, the toll station to toll station relationship block For elements all 0 Matrix, the toll booth to gantry relationship block The matrix dimensions are , the gantry to the toll station relationship block The matrix dimensions are , the gantry to gantry relationship block The matrix dimensions are .

4. The vehicle traffic trajectory audit and analysis system based on spatiotemporal feature clustering according to claim 3 is characterized in that: The feature processing module is further configured to construct a vehicle traffic record matrix based on historical traffic trajectories, specifically including: Extract historical travel trajectory sequence , where each complete passage trajectory flow data for , For the entrance water, , Water flows for the door frame, , For export water, ; The trajectory information is converted into a directed edge matrix to complete the conversion of vehicle traffic information from cross-section information to interval traffic status information. The expression of the directed edge matrix is: in, , is the pass mark of the nth directed edge of the i-th flow, is the directed edge identifier, is the entry time, For departure time, For travel time, For the mileage, is the average speed of traffic.

5. The vehicle traffic trajectory audit and analysis system based on spatiotemporal feature clustering according to claim 1 or 4, characterized in that: The spatiotemporal features include time features and space features; The time characteristics include travel time, peak traffic period ratio and average speed, and the time characteristics are normalized using Z-score; The spatial features include mileage, number of lanes, node degree, longitude and latitude, traffic density and adjacent edge correlation, and the spatial features are normalized using a hybrid of Min-Max normalization and one-hot encoding.

6. The vehicle traffic trajectory audit and analysis system based on spatiotemporal feature clustering according to claim 5 is characterized in that: The time feature vector is ,in, is the average travel time, is the travel time variance, is the coefficient of variation of travel time, is the proportion of traffic flow during peak hours, is the average speed over the entire journey, is the maximum travel time, The minimum travel time.

7. The vehicle traffic trajectory audit and analysis system based on spatiotemporal feature clustering according to claim 1 is characterized in that: The cluster analysis module is specifically configured as follows: dynamically determine the optimal number of clusters through the elbow rule and GapStatistic, calculate the Euclidean distance and iteratively update the cluster center, and output the time cluster division respectively. and spatial clustering .

8. The vehicle traffic trajectory audit and analysis system based on spatiotemporal feature clustering according to claim 4 is characterized in that: The generation of candidate paths by combining spatiotemporal cluster features specifically includes: Missing type definition and preprocessing: The missing position of the flow data is defined as the inlet flow missing scenario, the outlet flow missing scenario and the gantry flow missing scenario. The expression corresponding to the inlet flow missing scenario is: , the expression corresponding to the outlet water loss scenario is , the expression corresponding to the gantry water flow missing scenario is ; Generate candidate path set: If the entrance flow is missing, extract the first known portal , reversely retrieve all possible entry candidate sets , forming a candidate path set ,in, , ; If the outlet flow is missing, extract the known gantry at the end , forward search for all possible exit nodes , forming a candidate path set ,in, , ; If the gantry sequence is missing, extract the missing starting endpoint ,from Forward search All possible nodes in the interval, where the set of directed edges during the period is , ; Output candidate path set .

9. A vehicle trajectory audit and analysis method based on spatiotemporal feature clustering, using the vehicle trajectory audit and analysis system based on spatiotemporal feature clustering according to any one of claims 1 to 8, characterized in that: The method comprises the following steps: Construct a highway toll collection attribute graph based on node and edge sets to form a topological network of node reachability relationships; Standardize the temporal and spatial characteristics of historical travel trajectories; Adaptive clustering is performed based on the standardized spatiotemporal features to obtain temporal and spatial clusters; Candidate paths are generated by combining spatiotemporal cluster features, and a comprehensive score is calculated through spatiotemporal similarity calculation and weight distribution to output the optimal reasoning path.

Images