Big data-based similar route identification method and device

Abstract

The application relates to the technical field of route identification, and relates to a similar route identification method and device based on big data. The method comprises the following steps: acquiring a joint data set; obtaining vehicle running information and a vehicle driving route of multiple vehicles according to the joint data set; cutting the vehicle driving route according to the vehicle running information, so as to obtain at least one vehicle driving sub-route; respectively extracting trajectory key features of the at least one vehicle driving sub-route; acquiring a trajectory clustering result of the vehicle driving route of the multiple vehicles according to the trajectory key features of the at least one vehicle driving sub-route; respectively performing distributed calculation of route similarity on vehicle driving routes of different clusters according to the trajectory clustering result of the vehicle driving route of the multiple vehicles, so as to obtain a route similarity calculation result; and obtaining a similar route group according to the route similarity calculation result. The method can reduce the calculation amount and improve the calculation speed on the basis of ensuring the accuracy of similar route identification.

Description

Technical Field

[0001] This application relates to the field of route recognition technology, and in particular to a method and apparatus for similar route recognition based on big data. Background Technology

[0002] With the widespread application of the Global Positioning System (GPS) and the comprehensive coverage of vehicle-to-everything (V2X) terminals (Tboxes), massive amounts of V2X data are collected and stored every moment. Among these data, vehicle latitude and longitude information can effectively describe vehicle trajectories, and this trajectory data has extremely high data mining value. By identifying the route similarity of vehicles, it is possible to effectively support application scenarios such as group path preference, transportation route optimization, and road network prediction.

[0003] Currently, there are various methods for identifying similar routes, including point-based methods, shape-based methods, segmentation-based methods, and task-specific methods. However, trajectory information generated over long periods of transport by multiple vehicles produces massive amounts of data. Identifying similar routes using all available latitude and longitude data poses a significant challenge to the computing power and speed of current server and cluster computing methods, easily leading to memory overflow issues.

[0004] Therefore, how to reduce the amount of computation and improve the speed of computation while ensuring the accuracy of similar route identification is a problem that needs to be solved. Summary of the Invention

[0005] Therefore, it is necessary to provide a method and apparatus for similar route identification based on big data that can reduce the amount of computation and improve the computation speed while ensuring the accuracy of similar route identification, in order to address the above-mentioned technical problems.

[0006] Firstly, this application provides a method for identifying similar routes based on big data. The method includes:

[0007] Obtain a joint dataset, which includes vehicle network information and road network information for multiple vehicles;

[0008] Based on the joint dataset, vehicle operation information and vehicle travel routes for multiple vehicles are obtained;

[0009] Based on the vehicle operation information, the vehicle travel route is divided to obtain at least one vehicle travel sub-route;

[0010] Extract the key trajectory features of at least one of the vehicle's driving sub-routes;

[0011] Based on the key trajectory features of at least one of the vehicle driving sub-routes, obtain the trajectory clustering results of the vehicle driving routes of multiple vehicles.

[0012] Based on the trajectory clustering results of the vehicle travel routes of multiple vehicles, the route similarity is calculated in a distributed manner for the vehicle travel routes of different clusters to obtain the route similarity calculation results.

[0013] Based on the route similarity calculation results, a group of similar routes is obtained.

[0014] In one embodiment, the step of segmenting the vehicle travel route based on the vehicle operation information to obtain at least one vehicle travel sub-route includes:

[0015] Based on the vehicle operation information, obtain the current vehicle type information and vehicle GPS timestamp information. The vehicle type information is at least divided into passenger car type and commercial vehicle type.

[0016] If the current vehicle is a passenger car, then the vehicle's route is segmented based on the vehicle's GPS timestamp information to obtain at least one vehicle route.

[0017] In one embodiment, after obtaining the vehicle type information and vehicle GPS timestamp information of the current vehicle based on the vehicle operation information, the method further includes:

[0018] If the current vehicle type is not a passenger car, then the cargo information and POI information of the current vehicle are obtained by analyzing the vehicle operation information.

[0019] Based on the vehicle's GPS timestamp information, cargo information, and POI information, the vehicle's driving route is segmented to obtain at least one vehicle driving sub-route.

[0020] In one embodiment, the extraction of key trajectory features from at least one of the vehicle's driving sub-routes includes:

[0021] Based on the vehicle operation information, obtain the current vehicle's 1Hz positioning data;

[0022] Based on the current vehicle's 1Hz positioning data, at least one of the vehicle's driving sub-routes is sampled at equal intervals to obtain equal interval sampling results;

[0023] Based on the equidistant sampling results, the minimum description length algorithm is used to calculate the gain between sampling points, and the calculation results of the gain between sampling points are generated.

[0024] Based on the calculation results of the gain between the sampling points, the key trajectory features of at least one of the vehicle driving sub-routes are extracted.

[0025] In one embodiment, obtaining the trajectory clustering results of multiple vehicle routes based on the trajectory key features of at least one of the vehicle travel sub-routes includes:

[0026] Based on the key trajectory features of at least one of the vehicle driving sub-routes, the Quick bundles algorithm is used to cluster the trajectory of each vehicle driving route, generating a list of trajectory clustering results for multiple vehicle driving routes.

[0027] In one embodiment, the method further includes,

[0028] Based on the trajectory clustering results of multiple vehicle routes, the LCSS algorithm is used to calculate the length of the longest common subsequence of any two vehicle routes in the same cluster, and then normalization is performed to obtain the route similarity.

[0029] Secondly, this application also provides a similar route identification device based on big data. The device includes:

[0030] The information acquisition module is used to acquire a joint dataset, which includes vehicle network information and road network information of multiple vehicles;

[0031] The information processing module is used to obtain vehicle operation information and vehicle travel routes of multiple vehicles based on the joint dataset.

[0032] The information processing module is also used to divide the vehicle's driving route according to the vehicle's operating information to obtain at least one vehicle driving sub-route;

[0033] The information processing module is also used to extract the key trajectory features of at least one of the vehicle's driving sub-routes;

[0034] The information processing module is also used to obtain the trajectory clustering results of the vehicle routes of multiple vehicles based on the trajectory key features of at least one of the vehicle driving sub-routes.

[0035] The information processing module is also used to perform distributed calculation of route similarity for the vehicle routes of different clusters based on the trajectory clustering results of the vehicle routes of multiple vehicles, and obtain route similarity calculation results.

[0036] The information processing module is also used to obtain a group of similar routes based on the route similarity calculation results.

[0037] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the following steps:

[0038] Obtain a joint dataset, which includes vehicle network information and road network information for multiple vehicles;

[0039] Based on the joint dataset, vehicle operation information and vehicle travel routes for multiple vehicles are obtained;

[0040] Based on the vehicle operation information, the vehicle travel route is divided to obtain at least one vehicle travel sub-route;

[0041] Extract the key trajectory features of at least one of the vehicle's driving sub-routes;

[0042] Based on the key trajectory features of at least one of the vehicle driving sub-routes, obtain the trajectory clustering results of the vehicle driving routes of multiple vehicles.

[0043] Based on the trajectory clustering results of the vehicle travel routes of multiple vehicles, the route similarity is calculated in a distributed manner for the vehicle travel routes of different clusters to obtain the route similarity calculation results.

[0044] Based on the route similarity calculation results, a group of similar routes is obtained.

[0045] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, performs the following steps:

[0046] Obtain a joint dataset, which includes vehicle network information and road network information for multiple vehicles;

[0047] Based on the joint dataset, vehicle operation information and vehicle travel routes for multiple vehicles are obtained;

[0048] Based on the vehicle operation information, the vehicle travel route is divided to obtain at least one vehicle travel sub-route;

[0049] Extract the key trajectory features of at least one of the vehicle's driving sub-routes;

[0050] Based on the key trajectory features of at least one of the vehicle driving sub-routes, obtain the trajectory clustering results of the vehicle driving routes of multiple vehicles.

[0051] Based on the trajectory clustering results of the vehicle travel routes of multiple vehicles, the route similarity is calculated in a distributed manner for the vehicle travel routes of different clusters to obtain the route similarity calculation results.

[0052] Based on the route similarity calculation results, a group of similar routes is obtained.

[0053] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, performs the following steps:

[0054] Obtain a joint dataset, which includes vehicle network information and road network information for multiple vehicles;

[0055] Based on the joint dataset, vehicle operation information and vehicle travel routes for multiple vehicles are obtained;

[0056] Based on the vehicle operation information, the vehicle travel route is divided to obtain at least one vehicle travel sub-route;

[0057] Extract the key trajectory features of at least one of the vehicle's driving sub-routes;

[0058] Based on the key trajectory features of at least one of the vehicle driving sub-routes, obtain the trajectory clustering results of the vehicle driving routes of multiple vehicles.

[0059] Based on the trajectory clustering results of the vehicle travel routes of multiple vehicles, the route similarity is calculated in a distributed manner for the vehicle travel routes of different clusters to obtain the route similarity calculation results.

[0060] Based on the route similarity calculation results, a group of similar routes is obtained.

[0061] The above-mentioned method and apparatus for similar route identification based on big data have the following technical effects:

[0062] 1. Based on vehicle network and road network information during vehicle operation, key driving features are extracted from the vehicle's trajectory. Then, a clustering algorithm is used to cluster the trajectories. Finally, a distributed computing method distributes the routes from different clusters to different actuators for computation. By combining several algorithmic simplification techniques, the computational load is significantly reduced, computational efficiency is improved, and similar route identification based on big data is achieved.

[0063] 2. By using equidistant sampling to transform the data from dense to sparse, and then calculating the gain between data points based on the minimum description length algorithm, the key feature points of the route can be effectively extracted, reducing the sample size for subsequent calculations by hundreds of times.

[0064] 3. The LCSS algorithm is used to calculate the length of the longest common subsequence of any two vehicle routes in the same cluster, and then normalized. This makes the trajectory similarity measurement more robust. Attached Figure Description

[0065] Figure 1 This is an application environment diagram of a big data-based similar route identification method in one embodiment;

[0066] Figure 2 This is a flowchart illustrating a similar route identification method based on big data in one embodiment;

[0067] Figure 3 This is a schematic diagram illustrating the unfolding process of step S206 in another embodiment;

[0068] Figure 4 This is a schematic diagram illustrating the unfolding process of step S208 in another embodiment;

[0069] Figure 5 This is a structural block diagram of a similar route recognition device based on big data in one embodiment;

[0070] Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0071] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0072] First, a brief explanation of the terms used in the embodiments of this application will be provided:

[0073] Passenger vehicle types: including sedans, minibuses, and light passenger vehicles with no more than 9 seats. Passenger vehicles are further subdivided into basic passenger vehicles (sedans), multi-purpose vehicles (MPVs), sport utility vehicles (SUVs), special-purpose passenger vehicles, and crossover passenger vehicles.

[0074] Commercial vehicle types: All freight vehicles and buses with more than 9 seats are divided into five categories: freight trucks, semi-trailer tractors, incomplete passenger vehicles, and incomplete freight trucks.

[0075] 1Hz positioning data: refers to data that collects location information once per second. It is usually used in some low-speed motion application scenarios, and the accuracy is generally around 10 meters.

[0076] Route isometric sampling: In a Geographic Information System (GIS), this refers to sampling a route at equal intervals, distributing points evenly along the route to facilitate subsequent spatial analysis and processing. This sampling method determines the number and location of sampling points by calculating the route length and sampling interval, thus achieving isometric sampling of the route.

[0077] Minimum Description Length Algorithm: This refers to a general data compression algorithm used to calculate the gain between sampling points. Its basic idea is to use encoding to describe the characteristics of the data, thereby achieving data compression. The specific implementation process is as follows: 1. Encode the coordinates of the sampling points to generate an encoding sequence. 2. Calculate the encoding length using the encoding sequence, which is the minimum number of bits required to describe the data. 3. Calculate the gain between sampling points, which is the difference in encoding length between adjacent sampling points.

[0078] The similar route identification method based on big data provided in this application can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. The data storage system can store the combined data that server 104 needs to process. The data storage system can be integrated onto server 104, or it can be located in the cloud or on other network servers.

[0079] Taking the method flow provided in this application embodiment executed by server 104 alone as an example, specifically, server 104 can obtain a joint dataset, which includes vehicle network information and road network information of multiple vehicles. Based on the joint dataset, it obtains vehicle operation information and vehicle routes of multiple vehicles. Then, based on the vehicle operation information, it segments the vehicle routes to obtain at least one vehicle sub-route; server 104 also extracts the trajectory key features of at least one vehicle sub-route; then, based on the trajectory key features of at least one vehicle sub-route, it obtains the trajectory clustering results of the vehicle routes of multiple vehicles; finally, based on the trajectory clustering results of the vehicle routes of multiple vehicles, server 104 performs distributed calculation of route similarity for the vehicle routes of different clusters to obtain route similarity calculation results; based on the route similarity calculation results, it obtains similar route groups. The terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can be smart speakers, smart TVs, smart air conditioners, smart vehicle devices, etc. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted devices. Server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0080] In some embodiments, the above-described terminology and implementation environment are combined, such as... Figure 2 As shown, a similar route identification method based on big data is provided. This method is applied to a server and includes the following steps:

[0081] Step S202: Obtain the joint dataset, which includes vehicle network information and road network information of multiple vehicles.

[0082] The joint dataset consists of two main types: first, vehicle network information transmitted via TBox, including vehicle VIN, GPS timestamp, instrument speed, GPS speed, engine speed, instantaneous fuel consumption rate, longitude, and latitude; and second, road network information, including longitude, latitude, province name, city name, and area code. The two types of data are linked and merged using latitude and longitude information to form a complete joint dataset. The joint dataset then undergoes preprocessing: first, null values ​​are padded and the data is sorted according to GPS timestamps; second, the format is converted from string to numeric type; and finally, the instrument speed, GPS speed, and engine speed are corrected.

[0083] Step S204: Based on the joint dataset, obtain the vehicle operation information and vehicle travel routes of multiple vehicles.

[0084] The vehicle operation information includes the vehicle's operating distance (e.g., the distance between the starting point A and the destination B) and intermediate stops (e.g., A1, A2, A3, B1, B2, and B3), and the vehicle's route includes the route from the starting point A to the destination B.

[0085] Step S206: Based on the vehicle operation information, the vehicle travel route is divided to obtain at least one vehicle travel sub-route.

[0086] In this embodiment of the application, the vehicle travel route is divided into multiple sub-routes based on intermediate stops, and each sub-routes includes multiple latitude and longitude points.

[0087] Step S208: Extract the key trajectory features of at least one vehicle driving sub-route.

[0088] The key features of the trajectory include trajectory density features, trajectory curvature features, trajectory velocity features, and trajectory dwell features.

[0089] Step S210: Based on the key trajectory features of at least one vehicle driving sub-route, obtain the trajectory clustering results of the vehicle driving routes of multiple vehicles.

[0090] The Quick Bundles algorithm uses typical latitude and longitude feature point stream data extracted from the routes as input. Each route is defined as an ordered set of points of fixed length. The algorithm determines whether to add the current route to an existing route cluster by calculating the minimum average direct flip distance. Quick Bundles does not perform reallocation or updates during the computation process, resulting in extremely high computational speed and efficiency. Once all routes have been clustered, a list is generated. Clustering transforms a group of multiple routes into multiple groups of fewer routes, effectively reducing the computational load for subsequent detailed similarity comparisons.

[0091] Step S212: Based on the trajectory clustering results of the vehicle travel routes of multiple vehicles, the route similarity is calculated in a distributed manner for the vehicle travel routes of different clusters to obtain the route similarity calculation results.

[0092] Based on the trajectory clustering results, pairwise similarity comparisons are performed on routes within the same category. If a category contains n routes, the number of pairwise comparisons is ((n)×(n-1)) / 2. Since routes between different categories are not compared, distributed computation can be performed, placing the similarity calculations for different categories in different executors to achieve efficient parallel processing to obtain the similarity between two routes.

[0093] Step S214: Based on the route similarity calculation results, obtain the similar route group.

[0094] Since similarity is calculated by comparing routes pairwise, similar routes will appear in pairs. However, there may be cases where one route is similar to multiple routes simultaneously, requiring further processing. Starting with the first pair of similar routes, a new group of similar routes is created and labeled. If any route in the next pair appears in a previously created group, the pair is added to that group. If not, new and labeled similar route groups are created continuously until all pairs of similar routes appear in the same group.

[0095] In summary, the above embodiments extract key driving features from vehicle network and road network information during vehicle operation, then cluster the trajectories using a clustering algorithm, and finally distribute the routes from different clusters to different actuators for computation to obtain similar route groups. By combining several algorithm simplification methods, the computational load is greatly reduced, computational efficiency is improved, and similar route identification based on big data is achieved.

[0096] In one embodiment, such as Figure 3 As shown, step S206 includes steps S302-S306.

[0097] Step S302: Based on the vehicle operation information, obtain the current vehicle type information and vehicle GPS timestamp information. The vehicle type information is divided into at least passenger car type and commercial vehicle type.

[0098] Step S304: Determine whether the current vehicle type belongs to the passenger car type.

[0099] Step S306: If the current vehicle type is a passenger car, then the vehicle travel route is divided according to the vehicle GPS timestamp information to obtain at least one vehicle travel sub-route.

[0100] If the current vehicle type is not a passenger car, then based on the vehicle operation information, the current vehicle's cargo information and POI information are analyzed; based on the vehicle's GPS timestamp information, cargo information, and POI information, the vehicle's driving route is segmented to obtain at least one vehicle driving sub-route.

[0101] Specifically, the vehicle GPS timestamp refers to the start and end times of the trip determined by the GPS system. Based on the vehicle GPS timestamp information, for example, if the time difference between the end timestamp of route (A1-A2) and the start timestamp of route (A2-A3) is greater than 5 minutes, it indicates that the vehicle has been at location A2 for at least 5 minutes. If the vehicle type is a passenger car, the end timestamp of route (A1-A2) and route (A2-A3) can be divided into two separate sub-routes; otherwise, they are not divided into two separate vehicle travel sub-routes. In the preferred embodiment, 5 minutes can be used as a time threshold as the basis for dividing two adjacent sub-routes. If the vehicle type is not a passenger car (e.g., a commercial vehicle), cargo information and POI information need to be considered, combined with the vehicle GPS timestamp, and a preset weight ratio (e.g., vehicle GPS timestamp information, cargo information, and POI information account for 0.6, 0.3, and 0.1 respectively) to determine two adjacent vehicle travel sub-routes.

[0102] In summary, this division method can save on the number of vehicle travel sub-routes and reduce the amount of calculations required later.

[0103] In one embodiment, such as Figure 4 As shown, step S208 includes steps S402-S408.

[0104] Step S402: Obtain the current vehicle's 1Hz positioning data based on the vehicle's operating information.

[0105] Step S404: Based on the current 1Hz positioning data of the vehicle, perform equidistant sampling on at least one vehicle driving sub-route to obtain equidistant sampling results.

[0106] Step S406: Based on the equidistant sampling results, the minimum description length algorithm is used to calculate the gain between sampling points, and the calculation result of the gain between sampling points is generated.

[0107] Step S408: Based on the calculation results of the gain between sampling points, extract the key trajectory features of at least one vehicle driving sub-route.

[0108] 1Hz positioning data, collected once per second, is typically used in low-speed motion applications. The process involves reading 1Hz positioning data from the vehicle's sub-route, including longitude, latitude, and timestamps. The distance between adjacent positioning points is then calculated, and the number of positioning points to be retained is determined based on the required equidistant sampling interval. Following this calculation, corresponding positioning points are selected along the vehicle's sub-route according to the equidistant sampling interval, yielding the equidistant sampling results. These results are then stored and processed, for example, saved as a CSV file or a table in a database, for subsequent vehicle trajectory analysis and visualization.

[0109] The gain between sampling points also reflects the vehicle's turning behavior between two adjacent sampling points. Therefore, the curvature change of the route can be calculated by the change in the gain between sampling points. The specific implementation process is as follows: For the gain between two adjacent sampling points, if their difference exceeds a threshold, it indicates that the vehicle has turned between these two sampling points, and this turning point can be used as the curvature change point of the route. By calculating the distance between adjacent curvature change points, the degree of curvature change of the route can be obtained, i.e., the curvature change of the route. The curvature change reflects different key trajectory features, as detailed below:

[0110] Trajectory density characteristics: The intervals where the gain between sampling points is 0 indicate that the vehicle's driving direction is relatively stable and the trajectory density is high. These intervals can be used to extract the main driving route of the vehicle.

[0111] Trajectory curvature characteristics: The interval with a large gain between sampling points indicates that the vehicle's driving direction changes greatly and the trajectory curvature is large. These intervals can be used to extract the curves and corners of the vehicle's driving.

[0112] Trajectory speed characteristics: The intervals with smaller gains between sampling points indicate that the vehicle speed is relatively stable. These intervals can be used to extract the road sections such as highways or urban expressways where the vehicle is traveling.

[0113] Trajectory dwell characteristics: If there is a long time interval in the interval where the gain is 0 between sampling points, it indicates that the vehicle is dwelling at that position. These intervals can be used to extract the dwell point information of the vehicle.

[0114] In summary, the calculated gain between sampling points can provide various key trajectory features, which is helpful for analyzing and processing vehicle trajectories. To simplify the calculation, the 1Hz data in the driving route is first sampled at equal intervals. Then, the gain between points is calculated using the minimum description length algorithm, which is widely used in information compression methods, ensuring the simplicity and accuracy of the extracted feature points. By transforming the data from dense to sparse through equal-interval sampling, and then calculating the gain between data points based on the minimum description length algorithm, the key feature points of the route are effectively extracted, reducing the sample size for subsequent calculations by hundreds of times.

[0115] In one embodiment, step S210 specifically includes: clustering the trajectory of each vehicle driving route according to the key features of the trajectory of at least one vehicle driving sub-route, and generating the trajectory clustering results of the vehicle driving routes of multiple vehicles in a list.

[0116] Specifically, key trajectory features of each vehicle's route are extracted, such as trajectory density, trajectory curvature, trajectory speed, and trajectory dwell characteristics. The trajectory data of each route is sampled at a certain resolution to obtain point cloud data for each route. The point cloud data of all routes is processed using the Quick Bundles algorithm, and then divided into several local regions. Registration is performed on each local region to obtain the rotation matrix and translation vector of the point cloud data within that region. Based on the rotation matrix and translation vector of all local regions, the final rotation matrix and translation vector of the entire point cloud data are calculated. The point cloud data of all routes are registered according to the calculated rotation matrix and translation vector to obtain the registration result of the entire point cloud data. Clustering of the registration results yields trajectory clustering results for multiple vehicle routes, with each cluster representing a different vehicle route pattern. Analysis and processing of the trajectory data in each cluster reveals the relevant characteristics and patterns of that type of vehicle route. Once all routes have been clustered, a list is generated. Clustering can transform a group of multiple routes into multiple groups of fewer routes, effectively reducing the computational burden of subsequent detailed similarity comparisons.

[0117] In summary, using the Quick Bundles algorithm to cluster vehicle routes can effectively analyze and process vehicle driving data, providing valuable reference information for fields such as traffic management, road network planning, and urban planning.

[0118] In one embodiment, step S212 specifically includes: based on the trajectory clustering results of the vehicle travel routes of multiple vehicles, using the LCSS algorithm to calculate the length of the longest common subsequence of any two vehicle travel routes in the same cluster, performing normalization processing, and obtaining the route similarity.

[0119] Specifically, based on the clustering results, pairwise similarity comparisons are performed on routes within the same category. If a category contains n routes, the number of pairwise comparisons is ((n)×(n-1)) / 2. Since routes between different categories are not compared, Spark can be used for distributed computation, placing the similarity calculations for different categories in different Executors to achieve efficient parallel processing. During the pairwise comparison process, the LCSS algorithm is used to find the longest common subsequence (LCS) between two routes. The length of the LCS is calculated and normalized to obtain the similarity between the two routes. Then, a custom similarity threshold is defined; two routes with a similarity exceeding this threshold (e.g., 90%) are identified.

[0120] In summary, using the LCSS algorithm to calculate the length of the longest common subsequence of any two vehicle routes in the same cluster and then normalizing it makes it more robust for trajectory similarity measurement.

[0121] In one embodiment, step S214 specifically includes: performing deduplication on similar route groups and then visualizing the results.

[0122] Since similarity is calculated pairwise, similar routes are identified in pairs. However, there are cases where a route is similar to multiple routes simultaneously, requiring further processing. Similarity is transitive; if route A is similar to both routes B and C, then route B is considered similar to route C. Based on this theory, starting with the first pair of similar routes, a new group of similar routes is created and labeled. If any route from the next pair appears in a previously created group, that pair is added to that group. If not, new and labeled similar route groups are created continuously until all pairs of similar routes appear in the same group. Because each group of similar routes is added in pairs, duplicate labels are removed from each group.

[0123] The similar route identification is visualized. Specifically, for the identified groups of similar vehicle routes, the results are visualized in tabular form. For each route within each group of similar routes, the output includes the similar route group number, the vehicle's VIN for each route, start time, end time, average speed, fuel consumption per 100 kilometers, and start and end locations. This completes the entire automatic similar route identification method.

[0124] In summary, after deduplication of similar route groups, the visualization technology can help users better understand and analyze vehicle driving data, thereby providing valuable reference information for fields such as traffic management, road network planning, and urban planning.

[0125] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0126] Based on the same inventive concept, this application also provides a big data-based similar route identification device for implementing the big data-based similar route identification method described above. The solution provided by this device is similar to the implementation described in the above method; therefore, the specific limitations in one or more big data-based similar route identification device embodiments provided below can be found in the limitations of the big data-based similar route identification method described above, and will not be repeated here.

[0127] In one embodiment, such as Figure 5 As shown, a similar route identification device based on big data is provided, including:

[0128] The information acquisition module 502 is used to acquire a joint dataset, which includes vehicle network information and road network information of multiple vehicles.

[0129] The information processing module 504 is used to preprocess the above joint dataset and obtain the vehicle operation information and vehicle travel routes of multiple vehicles based on the processed joint dataset.

[0130] The information processing module 504 is also used to divide the vehicle travel route according to the vehicle operation information to obtain at least one vehicle travel sub-route.

[0131] The information processing module 504 is also used to extract the key trajectory features of at least one of the vehicle's driving sub-routes.

[0132] The information processing module 504 is also used to obtain the trajectory clustering results of the vehicle travel routes of multiple vehicles based on the trajectory key features of at least one of the vehicle travel sub-routes.

[0133] The information processing module 504 is also used to perform distributed calculation of route similarity on the vehicle routes of different clusters according to the trajectory clustering results of the vehicle routes of multiple vehicles, and obtain route similarity calculation results.

[0134] The information processing module 504 is also used to obtain a group of similar routes based on the route similarity calculation results.

[0135] In other embodiments, the information acquisition module 502 is also used to acquire the vehicle type information and vehicle GPS timestamp information of the current vehicle based on the vehicle operation information. The vehicle type information is at least divided into passenger car type and commercial vehicle type.

[0136] The device also includes a determination module for determining whether the current vehicle is a passenger car.

[0137] The information processing module 504 is also used to, if the current vehicle type is a passenger car, divide the vehicle's driving route according to the vehicle's GPS timestamp information to obtain at least one vehicle driving sub-route.

[0138] If the current vehicle type is not a passenger car, then based on the vehicle operation information, the current vehicle's cargo information and POI information are analyzed; based on the vehicle's GPS timestamp information, cargo information, and POI information, the vehicle's driving route is segmented to obtain at least one vehicle driving sub-route.

[0139] In other embodiments, the information acquisition module 502 is also used to acquire the current vehicle's 1Hz positioning data based on the vehicle's operating information.

[0140] The information processing module 504 is also used to perform equidistant sampling on at least one vehicle driving sub-route based on the current vehicle's 1Hz positioning data to obtain equidistant sampling results.

[0141] The information processing module 504 is also used to calculate the gain between sampling points using the minimum description length algorithm based on the equidistant sampling results, and generate the calculation results of the gain between sampling points.

[0142] The information processing module 504 is also used to extract the key trajectory features of at least one vehicle driving sub-route based on the calculation results of the gain between sampling points.

[0143] In other embodiments, the information processing module 504 is further configured to cluster the trajectory of each vehicle driving route using the Quick bundles algorithm based on the key trajectory features of at least one vehicle driving sub-route, and generate a list of trajectory clustering results for multiple vehicle driving routes.

[0144] In other embodiments, the information processing module 504 is further configured to calculate the length of the longest common subsequence of any two vehicle routes in the same cluster based on the trajectory clustering results of the vehicle routes of multiple vehicles using the LCSS algorithm, and perform normalization processing to obtain the route similarity.

[0145] In other embodiments, the information processing module 504 is also used to perform deduplication processing on similar route groups and then display them visually.

[0146] The modules in the aforementioned big data-based similar route identification device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0147] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 6 As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores relevant data for a big data-based similar route identification method. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a big data-based similar route identification method.

[0148] In one embodiment, a computer-readable storage medium is provided. A computer program is stored thereon, which, when executed by a processor, performs the following steps:

[0149] Step S202: Obtain the joint dataset, which includes vehicle network information and road network information of multiple vehicles.

[0150] Step S204: Based on the joint dataset, obtain the vehicle operation information and vehicle travel routes of multiple vehicles.

[0151] Step S206: Based on the vehicle operation information, the vehicle travel route is divided to obtain at least one vehicle travel sub-route.

[0152] Step S208: Extract the key trajectory features of at least one vehicle driving sub-route.

[0153] Step S210: Based on the key trajectory features of at least one vehicle driving sub-route, obtain the trajectory clustering results of the vehicle driving routes of multiple vehicles.

[0154] Step S212: Based on the trajectory clustering results of the vehicle travel routes of multiple vehicles, the route similarity is calculated in a distributed manner for the vehicle travel routes of different clusters to obtain the route similarity calculation results.

[0155] Step S214: Based on the route similarity calculation results, obtain the similar route group.

[0156] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0157] Step S202: Obtain the joint dataset, which includes vehicle network information and road network information of multiple vehicles.

[0158] Step S204: Based on the joint dataset, obtain the vehicle operation information and vehicle travel routes of multiple vehicles.

[0159] Step S206: Based on the vehicle operation information, the vehicle travel route is divided to obtain at least one vehicle travel sub-route.

[0160] Step S208: Extract the key trajectory features of at least one vehicle driving sub-route.

[0161] Step S210: Based on the key trajectory features of at least one vehicle driving sub-route, obtain the trajectory clustering results of the vehicle driving routes of multiple vehicles.

[0162] Step S212: Based on the trajectory clustering results of the vehicle travel routes of multiple vehicles, the route similarity is calculated in a distributed manner for the vehicle travel routes of different clusters to obtain the route similarity calculation results.

[0163] Step S214: Based on the route similarity calculation results, obtain the similar route group.

[0164] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data shall comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0165] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0166] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0167] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for identifying similar routes based on big data, characterized in that, The method includes: Obtain a joint dataset, which includes vehicle network information and road network information for multiple vehicles; Based on the joint dataset, vehicle operation information and vehicle travel routes for multiple vehicles are obtained; Based on the vehicle operation information, the vehicle travel route is divided to obtain at least one vehicle travel sub-route; Extract the key trajectory features of at least one of the vehicle's driving sub-routes; Based on the key trajectory features of at least one of the vehicle travel sub-routes, obtain the trajectory clustering results of multiple vehicle travel routes, including: Based on the key trajectory features of at least one of the vehicle driving sub-routes, the Quick Bundles algorithm is used to cluster the trajectory of each vehicle driving route, generating a list of trajectory clustering results for multiple vehicle driving routes. Specifically, the trajectory data of each vehicle driving route is sampled at a certain resolution to obtain point cloud data for each route. The point cloud data of all routes is processed using the Quick Bundles algorithm, and then divided into several local regions. Registration is performed on each local region to obtain the rotation matrix and translation vector of the point cloud data within that region. Based on the rotation matrix and translation vector of all local regions, the final rotation matrix and translation vector of the entire point cloud data are calculated. The point cloud data of all routes are registered according to the calculated rotation matrix and translation vector to obtain the registration result of the entire point cloud data. The registration results are then clustered to obtain the trajectory clustering results for multiple vehicle driving routes, with each cluster representing a different vehicle driving route pattern. Based on the trajectory clustering results of the vehicle travel routes of multiple vehicles, the route similarity is calculated in a distributed manner for the vehicle travel routes of different clusters to obtain the route similarity calculation results. Based on the route similarity calculation results, a group of similar routes is obtained.

2. The method according to claim 1, characterized in that, The step of dividing the vehicle's travel route based on the vehicle's operating information to obtain at least one vehicle travel sub-route includes: Based on the vehicle operation information, obtain the current vehicle type information and vehicle GPS timestamp information. The vehicle type information is at least divided into passenger car type and commercial vehicle type. If the current vehicle is a passenger car, then the vehicle's route is segmented based on the vehicle's GPS timestamp information to obtain at least one vehicle route.

3. The method according to claim 2, characterized in that, After obtaining the vehicle type information and vehicle GPS timestamp information of the current vehicle based on the vehicle operation information, the method further includes: If the current vehicle type is not a passenger car, then the cargo information and POI information of the current vehicle are obtained by analyzing the vehicle operation information. Based on the vehicle's GPS timestamp information, cargo information, and POI information, the vehicle's driving route is segmented to obtain at least one vehicle driving sub-route.

4. The method according to claim 1, characterized in that, The extraction of key trajectory features from at least one of the vehicle's driving sub-routes includes: Based on the vehicle operation information, obtain the current vehicle's 1Hz positioning data; Based on the current vehicle's 1Hz positioning data, at least one of the vehicle's driving sub-routes is sampled at equal intervals to obtain equal interval sampling results; Based on the equidistant sampling results, the minimum description length algorithm is used to calculate the gain between sampling points, and the calculation results of the gain between sampling points are generated. Based on the calculation results of the gain between the sampling points, the key trajectory features of at least one of the vehicle driving sub-routes are extracted.

5. The method according to claim 1, characterized in that, The method also includes, Based on the trajectory clustering results of multiple vehicle routes, the LCSS algorithm is used to calculate the length of the longest common subsequence of any two vehicle routes in the same cluster, and then normalization is performed to obtain the route similarity.

6. A similar route identification device based on big data, characterized in that, The device includes: The information acquisition module is used to acquire a joint dataset, which includes vehicle network information and road network information of multiple vehicles; The information processing module is used to obtain vehicle operation information and vehicle travel routes of multiple vehicles based on the joint dataset. The information processing module is also used to divide the vehicle's driving route according to the vehicle's operating information to obtain at least one vehicle driving sub-route; The information processing module is also used to extract the key trajectory features of at least one of the vehicle's driving sub-routes; The information processing module is further configured to obtain trajectory clustering results of multiple vehicle routes based on the trajectory key features of at least one of the vehicle driving sub-routes. Specifically, it is configured to cluster the route trajectories of each vehicle route using the Quick Bundles algorithm based on the trajectory key features of at least one of the vehicle driving sub-routes, and generate trajectory clustering results of multiple vehicle routes in a list. Specifically, the trajectory data of each vehicle route is sampled at a certain resolution to obtain point cloud data for each route; the point cloud data of all routes is processed using the Quick Bundles algorithm, and then the point cloud data is divided into several local regions. Each local region is registered to obtain the rotation matrix and translation vector of the point cloud data within that region; the final rotation matrix and translation vector of the entire point cloud data are calculated based on the rotation matrix and translation vector of all local regions; the point cloud data of all routes are registered according to the calculated rotation matrix and translation vector to obtain the registration result of the entire point cloud data; and the registration result is clustered to obtain the trajectory clustering results of multiple vehicle routes, where each cluster represents a pattern of vehicle routes. The information processing module is also used to perform distributed calculation of route similarity for the vehicle routes of different clusters based on the trajectory clustering results of the vehicle routes of multiple vehicles, and obtain route similarity calculation results. The information processing module is also used to obtain a group of similar routes based on the route similarity calculation results.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

9. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images