Method and device for automatically generating topological relation based on track time sequence

Abstract

The invention discloses a method and a device for automatically generating a topological relation based on track time sequence, and aims to solve the problems of an existing topological relation generation method. According to the method, automatic and precise construction and optimization of a topological relation are realized by deeply mining time sequence characteristics and spatial distribution rules of trajectory data and fusing technologies such as a graph theory and machine learning. The method comprises the following specific steps: preprocessing track data, removing noise, abnormal and repeated points, and sorting according to timestamps; track segmentation and feature extraction: segmenting and extracting time sequence, space and other features according to a multi-dimensional rule; space-time clustering analysis is carried out, and a DBSCAN algorithm is used for clustering track segments so as to reduce the subsequent analysis complexity; the method comprises the following steps: constructing a topological relation based on time sequence, establishing a connection condition according to time sequence continuity and spatial proximity to construct a preliminary topological connection graph, and optimizing through lane endpoint clustering and lane track smoothing; the device integrates a data acquisition module, a preprocessing module, a feature extraction module, a space-time clustering module, a topological connection identification module, a topological relation construction and optimization module and other modules for cooperative work. The method has remarkable effects in the aspects of data processing efficiency, topological relation accuracy, adaptability, expandability and labor cost reduction, and provides powerful technical support for the fields of intelligent transportation, geographic information systems and the like.

Description

Technical Field

[0001] This invention relates to the fields of data processing, geographic information systems, and big data analysis, and aims to automatically extract effective topological relationships from massive trajectory data. In this regard, an innovative method and apparatus for automatically generating topological relationships based on trajectory temporal sequence is invented. Background Technology

[0002] With the rapid development of mobile internet and IoT technologies, smart terminals such as positioning devices and smart wearable devices are rapidly becoming widespread. This significant change has made the large-scale, high-precision collection of trajectory data a practical reality. This trajectory data is like a treasure trove containing countless secrets, not only recording in detail the spatial location of a moving object at a specific time, but also revealing rich and diverse behavioral patterns and dynamic changes that evolve over time.

[0003] In the fields of data analytics and intelligent transportation, accurately extracting valuable information from massive amounts of trajectory data has always been a core task. Among these tasks, constructing a topological map that accurately reflects the actual road network structure and real-time traffic flow characteristics is of paramount importance. The actual road network structure is complex and varied, encompassing various special cases such as one-way streets, roundabouts, and roads with different numbers of lanes. Traffic flow characteristics can also fluctuate dramatically due to factors such as morning and evening rush hours, holidays, and traffic accidents.

[0004] In the past, the construction of topological relationships mainly relied on manual annotation or algorithm design based on static maps. Manual annotation not only requires a significant investment of manpower and time, but is also extremely inefficient when dealing with massive and constantly updated trajectory data. Algorithms based on static maps struggle to adapt to dynamically changing traffic environments and cannot reflect real-time changes such as temporary road closures and newly opened road sections. In recent years, some studies have attempted to automate the generation of topological relationships using machine learning and data mining techniques. However, most of these studies have focused on the extraction and matching of spatial features, neglecting the crucial temporal information in trajectory data. For example, the frequency and speed variations of vehicles passing through the same road segment at different times of the day are essential for understanding traffic flow characteristics, but this information has not been fully explored in previous studies. Therefore, how to cleverly and fully utilize the temporal characteristics of trajectory data to achieve automated and intelligent generation of topological relationships has become a key problem that urgently needs to be solved in this field. Summary of the Invention

[0005] This invention focuses on solving the problems existing in current topology relation generation methods, and innovatively proposes a method and apparatus for automatically generating topology relations based on trajectory temporal characteristics. This solution deeply mines the temporal characteristics and spatial distribution patterns in trajectory data, and organically integrates cutting-edge technologies such as graph theory and machine learning to achieve automated, accurate construction and optimization of topology relations, providing strong technical support for fields such as intelligent transportation and geographic information systems. The specific steps are as follows: I. Methods and Steps (I) Step 1: Trajectory Data Preprocessing First, a comprehensive and meticulous cleaning and preprocessing procedure is carried out on the raw trajectory data, with the core objective of ensuring high accuracy and reliability of the data. The specific steps are as follows: Noise data identification and removal: Signal interference and complex environmental factors often lead to noise data being mixed into the original trajectory. This type of noise data usually manifests as abnormal jumps in position coordinates or drastic fluctuations in key parameters such as velocity and direction. This invention, by setting a scientifically reasonable threshold system, such as velocity threshold and acceleration threshold, can accurately identify and efficiently remove this noise data. It should be noted that the threshold setting is closely related to the timestamp interval of the trajectory points and should be flexibly adjusted according to the needs of the actual application scenario; this invention does not impose fixed limitations on it.

[0006] Precise detection and removal of anomalies: Anomalies exhibit behavior patterns in trajectory data that are drastically different from the vast majority of normal data points. For example, vehicle trajectory points may appear in unlikely areas, such as inside green belts, or form trajectory segments that deviate significantly from other points. This invention utilizes advanced distance-based detection methods, such as the mature noise point identification technology in the DBSCAN clustering algorithm, to accurately locate and effectively remove these anomalies.

[0007] Efficient removal of duplicate entries: Due to excessively high sampling frequency of the equipment or repeated transmission during data transmission, a large number of duplicate points are easily found in the original trajectory data. This invention, through a rigorously designed comparison mechanism, compares the position coordinates and timestamps of adjacent points one by one, accurately identifying and thoroughly eliminating these duplicate points, thereby greatly reducing data redundancy and improving data processing efficiency.

[0008] Timestamps are arranged in order: The trajectory data that has undergone the above processing is then strictly sorted in ascending order according to timestamps. This operation ensures that subsequent analysis closely follows the temporal characteristics of the data, laying a solid foundation for uncovering the inherent logical relationships within the trajectory data.

[0009] (II) Step Two: Trajectory Segmentation and Feature Extraction Intelligent trajectory segmentation: The long trajectory is divided into multiple shorter trajectory segments according to carefully designed rules. This operation aims to greatly simplify the subsequent analysis process, significantly reduce the amount of computation, and accurately extract key information from the trajectory. The specific segmentation rules are based on multi-dimensional factors: Based on dynamic changes in velocity or acceleration: When the velocity or acceleration of a segment in the trajectory fluctuates significantly, exceeding a pre-set threshold range, the system automatically segments the trajectory at that location. This method can accurately detect key dynamic events such as acceleration, deceleration, and stopping in the trajectory.

[0010] Based on significant changes in steering angle: If the steering angle of a segment in the trajectory exceeds a preset threshold, it indicates a significant change in trajectory direction, and the system will segment the trajectory at that location. This rule is particularly suitable for application scenarios that are highly sensitive to changes in trajectory direction.

[0011] Precise positioning based on semantic information: The system intelligently segments the trajectory based on semantic information contained within it, such as vehicle stopping points. For example, when the system detects a vehicle's stopping point, it automatically divides the trajectory into two parts at that point.

[0012] Comprehensive feature extraction: For each short trajectory segment, the system comprehensively extracts rich temporal and spatial features, specifically including the following key dimensions: Temporal characteristics: Detailed analysis of velocity changes: Accurately calculate the velocity of each point within the trajectory segment and deeply analyze its changing trends to gain insight into the dynamic evolution of velocity over time.

[0013] Accurate acceleration calculation: Based on the velocity difference between adjacent points, the acceleration is accurately calculated and the trend of acceleration change is analyzed in detail, providing key basis for evaluating the dynamic characteristics of the trajectory.

[0014] Precise time interval statistics: Precisely count the time interval between adjacent points in the trajectory segment, thereby accurately grasping the sampling frequency of the trajectory and the moving speed of the target object.

[0015] Spatial characteristics: Precise location coordinate recording: The latitude and longitude coordinates of each point in the trajectory segment are recorded in detail, providing basic data for accurately depicting the spatial location of the trajectory.

[0016] Accurate distance calculation: Depending on the actual needs, you can choose to calculate the straight-line distance between the start and end points of the trajectory segment, or calculate the actual driving distance through complex algorithms to meet the distance information needs of different application scenarios.

[0017] Precise steering angle measurement: In-depth analysis of the changes in direction within the trajectory segment, and calculation of the steering angle through precise algorithms to capture subtle changes in trajectory direction.

[0018] Accurate description of trajectory shape: Using advanced pattern recognition technology, the shape characteristics of trajectory segments are accurately described, such as determining whether they are straight lines, curves, polylines, or other different types.

[0019] Other features: Accurate extraction of dwell time: In segmented scenarios based on semantic information, the dwell time of dwell points is accurately extracted, providing key information for analyzing the static state of target objects.

[0020] (III) Step 3: Spatiotemporal Cluster Analysis This invention employs the classic DBSCAN clustering algorithm to conduct in-depth spatiotemporal clustering analysis on the trajectory segments processed above. This algorithm can efficiently group trajectory segments with similar characteristics and spatiotemporal distributions into the same category. This clustering operation not only significantly reduces the complexity of subsequent analysis processes but also greatly improves the accuracy and efficiency of topology construction, providing a high-quality data foundation for subsequent topology construction work.

[0021] (iv) Step 4: Constructing topological relationships based on temporal order Connection conditions are scientifically established: Based on rigorous temporal continuity and spatial proximity criteria, this invention scientifically establishes the conditions for determining whether a connection exists between two trajectory segments: Strict control over temporal continuity: By accurately comparing the timestamp information of trajectory segments, the two trajectory segments are required to be closely continuous in the time dimension, that is, the end time of one trajectory segment should be very close to the start time of another trajectory segment, so as to ensure the continuity of the time series.

[0022] Precise measurement of spatial proximity: Utilizing advanced distance calculation algorithms, such as the Euclidean distance algorithm, the spatial distance between the start and end points of a trajectory segment is accurately calculated. Only when this distance is within a reasonable range, indicating that the two trajectory segments are spatially close enough, can they possibly belong to the same lane.

[0023] Efficiently constructing the initial topology connection graph: The system performs a full traversal of all trajectory segments, rigorously checking each pair of segments to ensure they meet the established connection conditions. Once the conditions are met, the system quickly connects the nodes corresponding to the two trajectory segments, thus efficiently constructing a preliminary topology graph.

[0024] Topology connection depth optimization and adjustment: Considering that the initially constructed topology connections may have practical problems such as disconnection and offset, this invention designs a deep optimization and adjustment mechanism: Intelligent clustering of lane endpoints: This invention again utilizes the DBSCAN clustering algorithm to perform deep spatiotemporal clustering analysis on lane endpoints. During implementation, it is necessary to scientifically set two key parameters: the neighborhood radius (epsilon) and the minimum number of points (min_samples). The epsilon parameter is used to precisely control the cluster density; its value must be adjusted reasonably based on the distribution characteristics of the actual data to ensure accurate capture of the spatial relationships between lane endpoints. The min_samples parameter determines the minimum number of neighboring points necessary for a point to become a core point; it also needs to be optimized based on the actual data to ensure the formation of stable and reliable clusters.

[0025] During the algorithm's operation, it accurately identifies core points (high-density points with a sufficient number of neighboring points), boundary points (points located within the neighborhood of core points but with relatively low density), and noise points (isolated points that are neither core points nor boundary points). Through a unique density reachability relationship, the algorithm cleverly connects core points and boundary points into clusters, with each cluster representing a group of lane endpoints that are closely adjacent in the spatiotemporal dimension.

[0026] Lane trajectory optimization and smoothing: This invention employs an advanced optimal control method to perform fine-grained smoothing of lane trajectories. By constructing accurate dynamic and trajectory optimization models, the system can efficiently calculate the optimal vehicle control input and driving trajectory. This method fully considers the vehicle's dynamic constraints, such as maximum acceleration limits and turning radius limits, thereby achieving optimal trajectory smoothing and effectively improving the accuracy and practicality of topology relationships.

[0027] II. Equipment Design To efficiently implement the aforementioned innovative method, this invention designs a device for automatically generating topological relationships based on trajectory temporal sequence. This device integrates multiple powerful modules, which work collaboratively to ensure the efficient and accurate operation of the entire topological relationship generation process. Data acquisition module: It has strong compatibility and can quickly and stably collect trajectory data from various devices, providing a rich data source for subsequent processing.

[0028] Data preprocessing module: Performs deep cleaning and sorting preprocessing operations on the collected raw trajectory data to effectively remove noise, outliers and duplicates, ensuring data quality and laying a solid foundation for subsequent analysis.

[0029] Feature extraction module: accurately extracts key temporal and spatial features from preprocessed data, providing core data support for the construction of topological relationships.

[0030] Spatiotemporal clustering module: Utilizes advanced clustering algorithms to perform efficient spatiotemporal clustering analysis on trajectory segments, reducing data complexity and improving data processing efficiency.

[0031] Topology connection identification module: Based on the extracted temporal features and spatial location information, it accurately identifies the potential connection relationships between trajectory segments, providing key clues for the construction of topological relationships.

[0032] Topology Relationship Construction and Optimization Module: Responsible for constructing the initial topology network and using advanced graph theory algorithms to deeply optimize it, ensuring that the generated topology relationships are accurate and reliable. Attached Figure Description

[0033] Figure 1 This is a schematic diagram of the basic process for automatically generating topological relationships based on trajectory temporality in some embodiments of the present invention. Detailed Implementation

[0034] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Example

[0035] refer to Figure 1 This invention presents a method for automatically generating topological relationships based on trajectory temporal characteristics. It deeply mines the temporal features and spatial distribution of trajectory data and organically integrates graph theory, machine learning, and other related technologies to achieve automatic construction and optimization of topological relationships. The method mainly includes the following key steps: Step 1: Trajectory Data Preprocessing First, a comprehensive cleaning and preprocessing process is performed on the raw trajectory data to remove noise, outliers, and duplicates, ensuring high accuracy and reliability. Simultaneously, the trajectory data is sorted according to timestamps to guarantee that subsequent analysis strictly adheres to temporal sequence. Specific operational details are as follows: Noise Data Identification and Removal: Noise data is typically generated due to signal interference, environmental factors, etc., and manifests as abrupt changes in position coordinates or abnormalities in key parameters such as velocity and direction. By setting a series of reasonable thresholds, such as velocity thresholds and acceleration thresholds, this noise data can be effectively identified and removed. It is important to note that the specific selection of the thresholds is closely related to the timestamp interval of the trajectory points, and can be flexibly set according to specific needs in practical applications.

[0036] Anomaly detection: Anomalies exhibit behavior patterns that significantly differ from the vast majority of data points. For example, they may appear in illogical areas (such as vehicle trajectory points appearing in a green belt) or form trajectory segments that deviate considerably from other points. Using distance-based methods (such as the mature noise point identification technique in DBSCAN clustering), these anomalies can be accurately detected and removed.

[0037] Duplicate point removal: Due to the high sampling frequency of the device or repeated transmission during data transmission, the original trajectory data often contains a large number of duplicate points. By carefully comparing the position coordinates and timestamps of adjacent points, these duplicate points can be accurately identified and removed, thereby effectively reducing data redundancy.

[0038] Step 2: Trajectory Segmentation and Feature Extraction The long trajectory is divided into multiple shorter trajectory segments according to specific rules to facilitate more detailed and in-depth analysis. Subsequently, rich temporal and spatial features, such as velocity changes, acceleration, turning angle, and trajectory shape, are extracted for each trajectory segment. These features will become an important basis for constructing the topological relationships in the later stages.

[0039] Trajectory segmentation: Trajectory segmentation is based on factors such as time interval, spatial distance, velocity change, acceleration, and turning angle. The specific segmentation method is as follows: Segmentation based on velocity or acceleration changes: If a significant change in velocity or acceleration occurs within a segment of the trajectory (i.e., exceeding a pre-set threshold), then a segment is created at that location. This method helps to accurately identify key events such as acceleration, deceleration, and stopping within the trajectory.

[0040] Segmentation based on steering angle: When the steering angle of a certain segment of the trajectory exceeds a preset threshold, it is segmented at that location. This method is particularly suitable for application scenarios that are highly sensitive to changes in trajectory direction.

[0041] Segmentation based on semantic information: Segmentation is performed based on the semantic information contained in the trajectory (such as stop points). For example, when a vehicle's parking position is detected, the trajectory can be divided into two parts at that point.

[0042] Feature extraction: After completing trajectory segmentation, the following types of features are extracted: Temporal characteristics: Velocity variation: Accurately calculate the velocity of each point within the trajectory segment and analyze its changing trends in depth to gain insight into the dynamic evolution of velocity over time.

[0043] Acceleration: Acceleration is accurately calculated based on the velocity difference between adjacent points, and the changes in acceleration are analyzed in detail, providing key basis for in-depth evaluation of the dynamic characteristics of the trajectory.

[0044] Time interval: Carefully count the time interval between adjacent points in the trajectory segment to understand the sampling frequency of the trajectory and the moving speed of the target object.

[0045] Spatial characteristics: Location coordinates: The latitude and longitude coordinates of each point in the trajectory segment are recorded in detail, providing a solid data foundation for accurately depicting the position of the trajectory in space.

[0046] Distance: Depending on the actual needs, the straight-line distance between the start and end points of the trajectory segment can be flexibly selected, or the actual driving distance can be calculated through a more complex algorithm to meet the diverse needs of different application scenarios for distance information.

[0047] Steering Angle: By deeply analyzing the changes in direction within the trajectory segment and calculating the steering angle using precise algorithms, subtle changes in trajectory direction can be keenly captured.

[0048] Trajectory Shape: Using advanced pattern recognition technology, the shape characteristics of the trajectory segment are accurately described, such as determining whether it is a straight line, curve, polyline, or other different types.

[0049] Other features: Dwell time: In the process of segmentation based on semantic information, the dwell time of dwell points is accurately extracted, providing key information for in-depth analysis of the static state of the target object.

[0050] Velocity distribution: Using statistical methods, comprehensively analyze the Gaussian distribution or other representative statistical characteristics of velocity in the trajectory segment to reveal the distribution pattern of velocity.

[0051] Acceleration distribution: Similarly, an in-depth analysis of the distribution characteristics of acceleration provides a multi-dimensional perspective for a more comprehensive understanding of the dynamic characteristics of the trajectory.

[0052] Step 3: Spatiotemporal clustering analysis The classic DBSCAN clustering algorithm is used to perform in-depth spatiotemporal clustering analysis on the trajectory segments processed above. This algorithm can effectively group trajectory segments with similar characteristics and spatiotemporal distribution into the same category. This clustering operation not only significantly reduces the complexity of subsequent analysis processes but also greatly improves the accuracy and efficiency of topology construction, laying a solid data foundation for subsequent topology construction work.

[0053] Step 4: Constructing topological relationships based on temporal order Based on spatiotemporal clustering, a preliminary topological connectivity graph is constructed according to the temporal continuity and spatial proximity between trajectory segments. The specific implementation process is as follows: Determine connection conditions: Temporal continuity: Two trajectory segments should be continuous in the time dimension, that is, the end time of one trajectory segment should be close to the start time of the other trajectory segment. This condition can be achieved by carefully comparing the timestamps of the trajectory segments.

[0054] Spatial proximity: Two track segments should be spatially close enough to indicate that they may belong to the same lane. This condition can be measured by calculating the spatial distance (such as Euclidean distance) between the start and end points of the track segments.

[0055] Constructing a preliminary topology graph: Traverse all trajectory segments one by one, and for each pair of trajectory segments, strictly check whether they meet the connection conditions established above. Once the conditions are met, quickly connect the nodes corresponding to the two trajectory segments, thereby efficiently constructing a preliminary topology graph.

[0056] Optimization and Adjustment: The initially constructed topology connections often fail to ensure that the endpoints of two connected lanes completely overlap, potentially leading to disconnections or offsets. Therefore, optimization and adjustment are necessary to further improve its accuracy and usability. Specific optimization methods are as follows: Lane endpoint clustering: The DBSCAN algorithm is used again to perform deep spatiotemporal clustering analysis on lane endpoints.

[0057] Parameter settings: The two key parameters, neighborhood radius (epsilon) and minimum number of points (min_samples), should be set scientifically and reasonably. The neighborhood radius (epsilon) determines the cluster density, and its value needs to be flexibly adjusted based on the actual data distribution to ensure accurate capture of the reasonable spatial relationships between lane endpoints. The minimum number of points (min_samples) defines the minimum number of neighboring points a point must have to become a core point; this also needs to be optimized based on actual data to ensure the formation of stable and reliable clusters.

[0058] Clustering Operation: Based on the predefined epsilon and min_samples parameters, DBSCAN clustering is performed on the lane endpoint data. During the clustering process, the algorithm accurately identifies core points (i.e., those points with high density and a sufficient number of neighboring points), boundary points (points located within the neighborhood of core points but with relatively low density), and noise points (isolated points that belong neither to core points nor boundary points). Through a unique density reachability relationship, the algorithm cleverly connects core points and boundary points into clusters, each cluster representing a group of lane endpoints that are spatially and temporally close.

[0059] Lane trajectory smoothing: An advanced optimal control method is employed to perform fine-grained smoothing of the lane trajectory. By constructing accurate dynamic and trajectory optimization models, the system can efficiently calculate the optimal vehicle control input and driving trajectory. This method fully considers the vehicle's dynamic constraints, such as maximum acceleration and turning radius, thereby achieving optimal trajectory smoothing and significantly improving the accuracy and practicality of topology relationships.

[0060] Through the comprehensive and detailed description of the above implementation methods, the specific implementation process and technical points of the method and apparatus for automatically generating topological relationships based on trajectory temporality of the present invention can be more clearly understood. In practical applications, the technical solution of the present invention can provide a more efficient and accurate method for generating topological relationships in related fields, and has significant application value and promotional significance.

Claims

1. A method for automatically generating topological relationships based on trajectory temporal sequence, characterized in that, Includes the following steps: Basic data acquisition: Utilizing a highly compatible data acquisition module, trajectory data containing the spatiotemporal location information of moving objects is collected from multiple devices to provide data support for subsequent processing. Data preprocessing: The data preprocessing operation is completed through four steps: noise removal, outlier removal, duplicate point cleaning, and timestamp sorting. Feature grouping: Feature extraction is performed on the data through trajectory segmentation, temporal features, spatial features, and other features. Data clustering: The DBSCAN algorithm is used to cluster the trajectory segments in time and space, grouping those with similar features and distributions into one class. Then, all trajectory segments are traversed, and the DBSCAN algorithm is used to cluster the lane endpoints to complete the topology construction.

2. The method for automatically generating topological relationships based on trajectory temporal sequence according to claim 1, characterized in that, In the basic data acquisition step, the data acquisition module collects trajectory data from various devices. To ensure the comprehensiveness and accuracy of the data, it performs preliminary integration and verification of the collected data, providing a reliable data foundation for subsequent steps.

3. The method for automatically generating topological relationships based on trajectory temporal sequence according to claim 1, characterized in that, The velocity threshold, acceleration threshold, and their correlation with the timestamp interval set in the data preprocessing step are derived from the analysis and verification of a large amount of actual trajectory data. They can effectively identify and process noisy data, and meet the needs of subsequent data processing and topology generation.

4. The method for automatically generating topological relationships based on trajectory temporal sequence according to claim 1, characterized in that, In the feature grouping step, trajectory segmentation is performed based on changes in velocity or acceleration, changes in steering angle, and semantic information because these factors can reflect the feature changes of the trajectory in different dimensions, which helps to accurately extract key information and provide effective support for subsequent topological relationship construction.

5. The method for automatically generating topological relationships based on trajectory temporal sequence according to claim 1, characterized in that, In spatiotemporal clustering analysis, the DBSCAN clustering algorithm is used. By reasonably setting parameters such as radius ε and minimum point threshold MinPts according to the characteristics of trajectory data and actual application scenarios, the algorithm can better capture the density-based natural clustering characteristics between trajectory segments, avoid unreasonable grouping, and ensure that the clustering effect meets the actual needs.

6. The method for automatically generating topological relationships based on trajectory temporal sequence according to claim 1, characterized in that, In the time-series-based topology construction steps, the established judgment conditions for temporal continuity and spatial proximity are derived from in-depth research on the actual road network structure and traffic flow characteristics. These conditions can accurately reflect the potential connection relationships between trajectory segments, laying the foundation for constructing accurate topology relationships.

7. The method for automatically generating topological relationships based on trajectory temporal sequence according to claim 1, characterized in that, In lane endpoint clustering, the scientific setting of the neighborhood radius (epsilon) and the minimum number of points (min_samples) is determined by analyzing a large amount of actual lane endpoint data and combining the spatiotemporal distribution characteristics of trajectory data. This can effectively improve the accuracy of topology construction and accurately capture the spatial correlation between lane endpoints.

8. The method for automatically generating topological relationships based on trajectory temporal sequence according to claim 1, characterized in that, The topology diagram generated through the above steps can comprehensively and accurately reflect the actual road network structure and real-time traffic flow characteristics. It can be displayed on the screen through a corresponding visualization system, providing intuitive data support for decision-making and management in fields such as intelligent transportation and geographic information systems.

9. The method for automatically generating topological relationships based on trajectory temporal sequence according to claim 1, characterized in that, The method also includes evaluating the generated topological relationships. Evaluation indicators include, but are not limited to, the completeness, accuracy, and matching degree of the topological relationships with the actual road network. By comparing and analyzing with actual road network data, the advantages of this method in generating topological relationships and the quality of data processing are verified.

Images