Travel chain processing method and device, equipment and storage medium

Abstract

The application provides a travel chain processing method and device, equipment and storage medium, and relates to the technical field of big data. The travel chain processing method comprises the following steps: determining a feature constraint condition of a target travel chain according to a first historical travel chain and a second historical travel chain which are close in time sequence; the feature constraint condition comprises a constraint condition that at least one target travel feature in the target travel chain needs to meet; determining a target travel feature set meeting the feature constraint condition from a historical travel feature set; each historical travel feature in the historical travel feature set is determined based on a historical travel chain set corresponding to the historical travel feature set, and the historical travel chain set comprises historical travel chains with similar travel features; and generating the target travel chain by using a plurality of historical travel features in the target travel feature set. According to the application, missing features of various types can be completed, and the continuity, integrity and accuracy of the travel chain completion can be improved.

Description

Technical Field

[0001] This application relates to the field of big data technology, and in particular to a method, apparatus, device and storage medium for processing travel chains. Background Technology

[0002] Obtaining accurate travel demand is crucial for transportation planning and design. Traditional travel survey-based methods suffer from long collection cycles, high costs, and incomplete information. With the development of information technology and the widespread use of tools such as mobile communication devices that record Global Positioning System (GPS) trajectory data, more detailed and abundant travel trajectory information has been recorded, providing a data foundation for travel chain reconstruction. However, during the travel chain reconstruction process, issues such as incomplete travel chains, inaccurate origins or destinations, and discontinuities in adjacent travel chains may arise, leading to biases in subsequent travel analysis. Therefore, it is necessary to supplement the missing features of the travel chains. Summary of the Invention

[0003] This application provides a method, apparatus, device, and storage medium for processing travel chains to complete missing features of the travel chain.

[0004] In a first aspect, embodiments of this application provide a method for processing travel chains, comprising: determining feature constraints of a target travel chain based on a first historical travel chain and a second historical travel chain that are adjacent in time sequence; wherein, the target travel chain is a travel chain after missing features are added to the travel chain to be completed; the feature constraints include constraints that at least one target travel feature in the target travel chain must satisfy, and the target travel feature includes a missing feature; the travel chain to be completed is the first historical travel chain or the second historical travel chain or a travel chain located between the first historical travel chain and the second historical travel chain in time sequence; determining a target travel feature set that satisfies the feature constraints from a set of historical travel features; wherein each historical travel feature in the set of historical travel features is determined based on a set of historical travel chains corresponding to the set of historical travel features, and the set of historical travel chains includes historical travel chains with similar travel features; and generating a target travel chain using multiple historical travel features in the target travel feature set.

[0005] Secondly, embodiments of this application provide a method for processing travel chains, including: obtaining a target travel feature set; wherein each historical travel feature in the target travel feature set is determined based on a set of historical travel chains corresponding to the target travel feature set, and the set of historical travel chains includes multiple historical travel chains with similar travel features; inputting multiple historical travel features from the target travel feature set into a pre-trained travel chain generation model to obtain the target travel chain.

[0006] Thirdly, embodiments of this application provide a method for processing travel chains, including: obtaining a historical travel feature set; wherein each historical travel feature in the historical travel feature set is determined based on a set of historical travel chains corresponding to the historical travel feature set, and the set of historical travel chains includes multiple historical travel chains with similar travel features; using multiple historical travel features in the historical travel feature set, training an initial deep learning model to obtain a travel chain generation model for generating a target travel chain.

[0007] Fourthly, embodiments of this application provide a trip chain processing apparatus, including:

[0008] The feature constraint determination module is used to determine the feature constraints of a target travel chain based on a first historical travel chain and a second historical travel chain that are adjacent in time sequence. The target travel chain is the travel chain after the missing features of the travel chain to be completed are added. The feature constraints include constraints that at least one target travel feature in the target travel chain must satisfy, and the target travel feature includes the missing feature. The travel chain to be completed is either the first historical travel chain or the second historical travel chain, or a travel chain that is located between the first historical travel chain and the second historical travel chain in time sequence. The target travel feature set determination module is used to determine the target travel feature set that satisfies the feature constraints from the historical travel feature set. Each historical travel feature in the historical travel feature set is determined based on a set of historical travel chains corresponding to the historical travel feature set, and the set of historical travel chains includes historical travel chains with similar travel features. The target travel chain generation module is used to generate a target travel chain using multiple historical travel features from the target travel feature set.

[0009] Fifthly, embodiments of this application provide an electronic device, including a memory, a processor, and a computer program stored in the memory, wherein the processor, when executing the computer program, implements the method described in any one of the embodiments of this application.

[0010] Sixthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the method described in any one of the embodiments of this application.

[0011] According to the travel chain processing method provided in this application, feature constraints are constructed for the missing features of the travel chain to be completed. Then, based on these feature constraints, a corresponding target travel feature set is selected from the user's historical travel feature set. Finally, a target travel chain is generated using the target travel feature set, thus completing the missing features of the travel chain to be completed. Since the target travel feature set originates from the user's historical travel feature set and covers various travel scenarios, it possesses richer travel feature information. Therefore, the target travel chain generated based on the target travel feature set can complete various types of missing features, thereby improving the continuity, completeness, and accuracy of travel chain reconstruction or completion.

[0012] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application, it can be implemented according to the contents of the specification. In order to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0013] In the accompanying drawings, unless otherwise specified, the same reference numerals throughout the various drawings denote the same or similar parts or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings depict only some embodiments according to this application and should not be construed as limiting the scope of this application.

[0014] Figure 1 This is a schematic diagram of the system architecture for the application scenario of this application;

[0015] Figure 2 This is a flowchart of the trip chain processing method according to Embodiment 1 of this application;

[0016] Figure 3A , Figure 3B and Figure 3C This is a schematic diagram illustrating the three types of feature loss in Embodiment 1 of this application;

[0017] Figure 4A This is a schematic diagram of a travel chain without missing features in Embodiment 1 of this application;

[0018] Figure 4B This is a schematic diagram of a travel chain with missing features in Embodiment 1 of this application;

[0019] Figure 5 This is a flowchart illustrating an application example of the travel chain processing method according to Embodiment 1 of this application;

[0020] Figure 6 This is a flowchart of the travel chain processing method according to Embodiment 2 of this application;

[0021] Figure 7This is a flowchart of the travel chain processing method according to Embodiment 3 of this application;

[0022] Figure 8 This is a block diagram of the trip chain processing device according to Embodiment 4 of this application;

[0023] Figure 9 This is a block diagram of an electronic device used to implement embodiments of this application. Detailed Implementation

[0024] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the concept or scope of this application. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0025] To facilitate understanding of the technical solutions of the embodiments of this application, the relevant technologies of the embodiments of this application are described below. The following relevant technologies are optional solutions and can be combined with the technical solutions of the embodiments of this application in any way, and all of them fall within the protection scope of the embodiments of this application.

[0026] Application scenarios

[0027] A travel chain is a combination of travel characteristics between a travel origin and destination, including one or more intermediate activity points, and is a primary basis for traffic demand analysis. The main data sources for reconstructing travel chains include mobile phone signaling data, data collected by volunteers wearing GPS devices, and location-based services (LBS) data.

[0028] Compared to LBS data, mobile signaling data suffers from disadvantages such as lower positioning accuracy and unstable sampling frequency. While LBS data can achieve meter-level positioning accuracy, mobile signaling data typically achieves only 200-500 meters in densely populated urban areas, and only 800-1000 meters in sparsely populated suburban areas. Furthermore, when a user connects to multiple base stations within one second, multiple signaling records with the same timestamp but different locations are generated, easily leading to drift data and ping-pong data—noisy data that deviates significantly from the user's actual location. Secondly, the generation frequency of mobile signaling data depends on the frequency of the user's network usage. When the user uses the network frequently, the data generation interval can reach a high frequency of several minutes to several seconds, while conversely, it can generate low-frequency collections lasting several hours.

[0029] Compared to LBS data, volunteer-based data collection suffers from higher costs, smaller data volumes, and insufficient sample representativeness. Volunteer-based collection requires recruiting volunteers to participate in a survey over a specific period, significantly limiting the number of participants, the time frame, and their spatial distribution. The types of samples collected are also limited to the recruited volunteers. In contrast, LBS data can cover nearly 90% of mobile phone users, encompassing cities nationwide over extended periods, resulting in a much larger data volume and higher sample representativeness than volunteer-based collection.

[0030] Location-Based Services (LBS) utilizes various positioning technologies to determine the current location of devices and provides information resources and basic services to them via the mobile internet. First, users can use positioning technology to determine their geographical location, and then access location-related resources and information via the mobile internet. LBS data includes the user's geographic location coordinates recorded when services are provided to users using wireless communication networks (or satellite positioning systems) on mobile terminals. Compared to traditional mobile signaling data, LBS data contains rich spatiotemporal feature information and has advantages such as high accuracy, wide coverage, and abundant information. Through LBS data analysis, the travel trajectory information of mobile objects can be mined, thus providing crucial evidence for reconstructing travel chains.

[0031] However, LBS data also suffers from uneven data sampling, signal loss, and data collection interruptions caused by users disabling location services. Therefore, when reconstructing travel chains based on LBS, there may be issues such as incomplete travel chains, inaccurate origin and destination points, and discontinuous nearby travels, which can lead to deviations in subsequent travel analysis and research.

[0032] In related technologies, the main methods for travel chain completion are to complete the travel path by addressing the problem of sparse trajectory sampling during the travel process. For example, methods such as linear interpolation and Dijkstra's algorithm to search for feasible paths can solve the problem of missing parts of the trajectory during the travel process. However, these methods cannot solve the problem of missing key travel information (travel features) in the travel chain, such as the missing features of the first segment of the travel chain, the missing features of the last segment of the travel chain, or the missing features of the entire travel chain.

[0033] Figure 1 This diagram illustrates a system architecture for an application scenario according to an embodiment of this application. The historical travel feature database stores the historical travel information of one or more users. Each user's historical travel information includes multiple historical travel feature sets and multiple historical travel chain sets. Furthermore, each user's multiple historical travel feature sets corresponds one-to-one with that user's historical travel chain set.

[0034] Each historical travel chain set includes one or more historical travel chains with similar travel characteristics. For example, travel chain identification algorithms can be used to identify travel chains from multiple users' travel data, resulting in multiple historical travel chains and their travel characteristics. Travel data includes mobile signaling data or LBS data, and travel chain identification algorithms include, but are not limited to, algorithms using rule models, probabilistic models, or clustering. For instance, a travel chain identification algorithm can generate travel information for all users (travelers), including their departure and destination points (OD), mode of transport, purpose of travel, and route. This yields a set of stops for each user within a day. Trajectory points between adjacent stops for the same user are then linked chronologically to form a travel chain for that traveler, and all travel information is aggregated into the travel characteristics of the complete travel chain.

[0035] For example, the travel characteristics of each historical trip may include one or more of the following: user ID, travel date, start time, end time, start location (e.g., start latitude and longitude), end location (e.g., end latitude and longitude), travel mode, travel purpose, travel route, travel duration, and travel distance.

[0036] Furthermore, the travel chains of the same user over multiple days are sorted according to their start time to obtain multiple historical travel chains of the same user over multiple days, such as trip1, ... trip1. i trip i+1 Then, a clustering algorithm is used to aggregate historical travel chains with similar travel characteristics into the same historical travel chain set, thereby obtaining one or more historical travel chain sets trip_c = {trip_c1, ... trip_c...} n Each historical travel chain set trip_c can include one or more historical travel chains trip1, ...} with similar travel characteristics. n trip n+1 ...; For each historical travel chain set, travel features are extracted to obtain the corresponding historical travel feature set C = {C1, ..., C...} n Each set of historical travel features C includes multiple historical travel features, which include at least one of the following: origin location, destination location, start time, end time, travel duration, travel distance, mode of travel, travel route, and travel purpose.

[0037] For example, the travel features of each historical travel chain can be constructed into a travel feature matrix. Multiple historical travel chain travel feature matrices are clustered according to their similarity, thus grouping historical travel chains whose similarity to the travel feature matrices meets a threshold into the same historical travel chain set. The clustered travel feature matrix is ​​then the historical travel feature set corresponding to that historical travel chain set. The clustering algorithm can be a hierarchical clustering algorithm, which analyzes data at different levels based on the similarity between clusters (such as travel feature matrices), thereby forming a tree-like clustering structure.

[0038] In one application example, the trip chain processing device is used to execute the trip chain processing method provided in Embodiment 1 of this application, including constructing feature constraints for the missing features of the trip chain to be completed, and then generating a corresponding target trip chain based on the feature constraints. The target trip chain is the trip chain after the missing features of the trip chain to be completed are filled. Specifically, it may include: determining the feature constraints of the target trip chain based on the first and second historical trip chains that are adjacent in time sequence. The feature constraints include the constraints that the target travel features in the target trip chain to be generated must satisfy, including the constraints that the missing features must satisfy; determining the corresponding target travel feature set from multiple historical travel feature sets based on the constructed feature constraints; and then generating the target trip chain based on the target travel feature set. The target travel feature set determined based on the feature constraints of the missing features comes from the historical travel feature set, that is, the user's historical travel information, which covers a variety of travel scenarios. Therefore, it has richer travel feature information. The target trip chain generated based on the target travel feature set can fill in various types of missing features, thereby improving the continuity, completeness, and accuracy of trip chain restoration or trip chain completion.

[0039] In another application example, the trip chain processing device is also used to execute the trip chain processing method provided in Embodiment 2 of this application, including: inputting multiple historical travel features from the target trip feature set into a pre-trained trip chain generation model to obtain a target trip chain after completing the missing features of the trip chain to be completed. Each historical travel feature in the target trip feature set is determined based on a set of historical trip chains corresponding to the target trip feature set, and the set of historical trip chains includes multiple historical trip chains with similar travel features. In other words, the target trip feature set originates from the user's historical travel feature information, covering various travel scenarios, and therefore has richer travel feature information. Consequently, the target trip chain generated based on the target trip feature set can complete various types of missing features, thereby improving the continuity, completeness, and accuracy of trip chain restoration or completion.

[0040] In another application example, the trip chain processing device is also used to execute the trip chain processing method provided in Embodiment 3 of this application, including: training an initial deep learning model using multiple historical travel features in the historical travel feature set to obtain a trip chain generation model for generating a target trip chain. Each historical travel feature in the historical travel feature set is determined based on a set of historical trip chains corresponding to the historical travel feature set, and the set of historical trip chains includes multiple historical trip chains with similar travel features. That is, the historical travel feature set originates from the user's historical travel information, covering various travel scenarios, and therefore has richer travel feature information. Consequently, the trip chain generation model trained based on the historical travel feature set can complete various types of missing features to generate the target trip chain, thereby improving the continuity, completeness, and accuracy of trip chain restoration or completion.

[0041] For example, the historical travel feature database can be a database, such as a local database or a cloud database; the historical travel feature database can also be deployed on the same entity as the travel chain processing device, which is not limited in this embodiment. In addition, the travel chain processing device can be deployed on the same entity, or deployed on different entities based on different internal functions. The aforementioned entity can be a hardware device with data processing functions (such as a server, terminal device) or a hardware chip. The hardware chip can be a CPU, a graphics processing unit (GPU), a field-programmable gate array (FPGA), a neural-network processing unit (NPU), an artificial intelligence (AI) accelerator card, or a data processing unit (DPU), etc.; the travel chain processing device can also be an application, service, instance, software functional module, virtual machine (VM), container, or cloud server, etc.; or it can be a combination of hardware and software.

[0042] It should be noted that the above-mentioned application scenarios or application examples of the travel chain processing method provided in the embodiments of this application are for ease of understanding, and the embodiments of this application do not specifically limit the application of the travel chain processing method.

[0043] Furthermore, 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, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. The collection, use and processing of the relevant data shall comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation portals shall be provided for users to select, edit or refuse authorization.

[0044] The technical solution of this application and how it solves the aforementioned technical problems are described in detail below with specific embodiments. The listed specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0045] Example 1

[0046] Figure 2 A flowchart illustrating a trip chain processing method according to an embodiment of this application is provided. This trip chain processing method can be applied to a trip chain processing apparatus. Figure 2 As shown, the processing method for this travel chain includes:

[0047] Step S201: Determine the feature constraints of the target travel chain based on the first and second historical travel chains that are adjacent in time sequence; wherein, the target travel chain is the travel chain after the missing features of the travel chain to be completed are filled; the feature constraints include the constraints that at least one target travel feature in the target travel chain must satisfy, and the target travel feature includes the missing feature; the travel chain to be completed is the first historical travel chain or the second historical travel chain or the travel chain located between the first historical travel chain and the second historical travel chain in time sequence.

[0048] For ease of explanation, the first historical travel chain is represented as trip. i The second historical travel chain is represented as trip. i+1 The historical travel chains of the same user over multiple days are sorted by start time, which can yield multiple historical travel chains for that user over multiple days, trip1, ... trip1. i trip i+1 Each historical travel chain includes multiple travel characteristics, which may include at least one of the following: origin location, destination location, start time, end time, travel duration, travel distance, mode of transport, travel route, and travel purpose.

[0049] In this context, missing features refer to the missing travel characteristics, and historical travel chains with missing features are the travel chains to be completed. Therefore, it is necessary to complete the missing features of the travel chains to be completed to generate the target travel chain. For example, the missing feature type of the travel chain to be completed can be the entire segment missing feature type, the tail segment missing feature type, or the first segment missing feature type.

[0050] like Figure 3A As shown, the origin travel characteristics (including origin location and origin time) and destination travel characteristics (including destination location and end time) of the trip chain to be completed are both missing. Therefore, the trip chain to be completed is located in the first historical trip chain "trip" in chronological order. i Second historical travel chain trip i+1 The travel chain between the segments, i.e., the travel chain to be completed, is of the type of missing features for the entire segment. The generated target travel chain will be located at the first historical travel chain in time sequence. i Second historical travel chain trip i+1 In this process, it is necessary to complete the travel features between point A1 and point B1 to generate the target travel chain.

[0051] like Figure 3B As shown, the endpoint travel feature of the trip to be completed is missing. Therefore, the feature loss type of the trip to be completed is the tail segment feature loss type, that is, the trip to be completed is the first historical trip. i It is necessary to examine the first historical travel chain trip. i The tail segment features (travel features between point A2 and point B2) are used to complete the target travel chain.

[0052] like Figure 3C As shown, the starting point feature of the trip chain to be completed is missing. Therefore, the feature missing type of the trip chain to be completed is the first segment feature missing type, that is, the trip chain to be completed is the second historical trip chain. i+1 It is necessary to examine the second historical travel chain trip. i+1 The first segment of features (travel features between point A3 and point B3) is completed to generate the target travel chain.

[0053] Therefore, the target travel chain, the first historical travel chain, and the second historical travel chain are related in time. The target travel chain can be the first historical travel chain after the missing features are filled, the second historical travel chain after the missing features are filled, or a travel chain that is located between the first historical travel chain and the second historical travel chain in time sequence.

[0054] For example, if the trip chain to be completed is the first historical trip chain (trip) i Or the second historical travel chain trip i+1Then, the first and second historical travel chains that are adjacent in time can be understood as being adjacent in time; if the travel chain to be completed is located at the first historical travel chain (trip) in time... i Second historical travel chain trip i+1 The travel chains between them, the first historical travel chain and the second historical travel chain that are adjacent in time can be understood as the first historical travel chain and the second historical travel chain being separated by one travel chain.

[0055] The travel features in the target travel chain are called target travel features, which may include at least one of the following: origin location, destination location, start time, end time, travel duration, travel distance, mode of transport, travel route, and travel purpose. When constructing feature constraints, one or more target travel features can be selected, and constraints must be set for each of them. The combination of constraints for each target travel feature constitutes the feature constraints of the target travel chain. These feature constraints must include constraints for any missing features.

[0056] Step S202: Determine the target travel feature set that meets the feature constraints from the historical travel feature set; wherein each historical travel feature in the historical travel feature set is determined based on the historical travel chain set corresponding to the historical travel feature set, and the historical travel chain set includes historical travel chains with similar travel features.

[0057] For example, one or more historical travel feature sets of a user can be obtained from a historical travel feature database, and a target travel feature set that meets the feature constraints can be determined from them. It can be understood that determining the target travel feature set is equivalent to determining a target travel chain set that meets the feature constraints from multiple historical travel chain sets. The target travel chain set corresponds to the target travel feature set; that is, each historical travel feature in the target travel feature set is extracted from the target travel chain set. The method for constructing the historical travel feature set can be found in the application scenarios section and will not be elaborated upon here.

[0058] Step S203: Generate the target travel chain by utilizing multiple historical travel features in the target travel feature set.

[0059] For example, one can find historical travel features corresponding to missing features from the target travel feature set, and then generate the corresponding target travel features in the target travel chain based on these historical travel features. In other words, one can use the historical travel features corresponding to missing features to complete the missing features and thus generate the target travel chain.

[0060] For example, multiple historical travel features from the target travel feature set can be input into a pre-trained travel chain generation model to obtain the target travel chain after the missing features of the travel chain to be completed. Specifically, the travel chain generation model can be obtained by training an initial deep learning model based on each historical travel feature set in the historical travel feature library. The initial deep learning model can be a Generative Adversarial Network (GAN) or other unsupervised or supervised learning networks; this embodiment of the application does not limit this.

[0061] According to the technical solution of this application embodiment, feature constraints are constructed based on the missing features of the travel chain to be completed. Then, based on these feature constraints, a corresponding target travel feature set is selected from the user's historical travel feature set. Finally, a target travel chain is generated using the target travel feature set, thus completing the missing features of the travel chain to be completed. Since the target travel feature set originates from the user's historical travel feature set and covers various travel scenarios, it possesses richer travel feature information. Therefore, the target travel chain generated based on the target travel feature set can complete various types of missing features, thereby improving the continuity, completeness, and accuracy of travel chain reconstruction or completion.

[0062] In one implementation, in step S201, the characteristic constraints of the target travel chain are determined based on the first historical travel chain and the second historical travel chain that are adjacent in time sequence. This includes: in response to the connection distance between the first historical travel chain and the second historical travel chain being greater than a preset distance threshold, constructing at least one characteristic constraint corresponding to a preset characteristic missing type based on the travel characteristics of the first historical travel chain and the second historical travel chain; wherein the target travel characteristics include the starting position, start time, ending position, and end time, and the preset characteristic missing type includes any one of the following: whole segment characteristic missing type, tail segment characteristic missing type, and first segment characteristic missing type.

[0063] Based on the first historical travel chain trip i Second historical travel chain trip i+1 The continuity between them can determine whether there are any missing travel chains in the user's historical travel chain database, that is, whether there are any travel chains that need to be completed.

[0064] For example, calculating the first historical trip chain. i Second historical travel chain trip i+1 The connection distance: i.e., the first historical travel chain trip i End position end_lnglat i With the second historical travel chain trip i+1 The starting position of start_lnglati+1 Distance between i，i+1 :dis i，i+1 =f_d(end_lnglat i ,start_lnglat i+1 ), where f_d represents an algorithm for calculating the distance between two points, such as the Floyd algorithm.

[0065] If dis i，i+1 Less than or equal to the preset distance threshold thres dis Then it can be determined that the first historical travel chain trip i The characteristics of the destination and the second historical travel chain trip i+1 They share the same starting point characteristics, have continuous travel chains, and no missing features, such as... Figure 4A As shown; if dis i，i+1 Greater than the preset distance threshold thres ais Then it is believed that the first historical travel chain trip i The characteristics of the destination and the second historical travel chain trip i+1 The starting point characteristics differ significantly, the travel chains are discontinuous, and there are travel chains that need to be completed, such as... Figure 4B As shown.

[0066] In cases where there are travel chains that need to be completed, the first historical travel chain (trip) can be utilized. i Trip characteristics and second historical travel chain characteristics i+1 For at least one preset missing type, at least one corresponding feature constraint is established. Since all three preset missing types are possible, the next step will be to construct feature constraints based on at least one of the three preset missing types, and then search the historical travel feature database for candidate travel feature sets that meet any of the feature constraints. For explanations of the whole-segment feature missing type, the tail-segment feature missing type, and the first-segment feature missing type, please refer to [link to relevant documentation]. Figure 3A , Figure 3B and Figure 3C The above description will not be repeated here. For example, the feature constraint corresponding to the entire segment of missing features can be defined as the first feature constraint, the feature constraint corresponding to the tail segment of missing features can be defined as the second feature constraint, and the feature constraint corresponding to the first segment of missing features can be defined as the third feature constraint. Accordingly, the feature constraint can be one or more of the first, second, and third feature constraints.

[0067] The target travel characteristics include origin travel and destination travel characteristics. Origin travel characteristics include origin location and start time, while destination travel characteristics include destination location and end time.

[0068] For example, if the entire feature segment is missing, it is necessary to base it on the first historical trip chain. i The destination travel characteristics are set to the starting travel characteristics of the target travel chain, based on the second historical travel chain trip. i+1 The originating travel characteristics are used to set the destination travel characteristics of the target travel chain. If the tail segment features are missing, then it is necessary to base it on the second historical travel chain (trip). i+1 The originating travel characteristics are used to set the destination travel characteristics of the target travel chain. If the first segment feature is missing, it needs to be based on the first historical travel chain (trip). i The destination travel characteristics are set to the starting travel characteristics of the target travel chain.

[0069] Based on this, feature constraints for the target travel chain can be constructed for three different travel incompleteness scenarios, including the constraints that the target travel features such as origin location, start time, destination location, and end time must satisfy. In other words, feature constraints can be understood as a set of spatiotemporal constraints. For the three different travel incompleteness scenarios, spatiotemporal constraint sets can be constructed to achieve travel chain completion in different travel scenarios.

[0070] In one implementation, the preset feature missing type includes a tail segment feature missing type. Based on the target travel features of the first historical travel chain and the target travel features of the second historical travel chain, feature constraints corresponding to the preset feature missing type are constructed, including: constructing a first feature constraint, which includes: the distance between the starting position of the target travel chain and the ending position of the first historical travel chain is less than or equal to a preset distance threshold; the start time and end time of the target travel chain are both between the end time of the first historical travel chain and the start time of the second historical travel chain; and the distance between the ending position of the target travel chain and the starting position of the second historical travel chain is less than or equal to a preset distance threshold.

[0071] In other words, if the entire segment of features is missing, it is necessary to base it on the first historical trip chain. i The destination travel characteristics are set to the origin travel characteristics (including origin location and start time) of the target travel chain, based on the second historical travel chain trip. i+1 The origin travel characteristics are set to the destination travel chain, which includes the destination location and the end travel characteristics (including the end time).

[0072] For example, if the entire feature segment is missing, the first feature constraint constructed includes:

[0073] (1) The starting position of the target trip chain and the first historical trip chain i The distance between the endpoints i，i_i+1Less than or equal to the preset distance threshold thres dis :

[0074] dis i，i_i+1 ≤thres dis ;

[0075] (2) Start time of the target travel chain i_i+1 Trip is in the first historical travel chain i end_time i With the second historical travel chain trip i+1 start_time i+1 between:

[0076] start_time i_i+1 ∈[end_time i start_time i+1 ];

[0077] (3) The endpoint of the target trip chain and the second historical trip chain i+1 The distance between the starting positions of dis i+1，i_i+1 Less than or equal to the preset distance threshold thres dis :

[0078] dis i+1，i_i+1 ≤thres dis ;

[0079] (4) End time of the target travel chain i_i+1 Trip is in the first historical travel chain i end_time i With the second historical travel chain trip i+1 start_time i+1 between:

[0080] end_time i_i+1 ∈[end_time i start_time i+1 ].

[0081] Therefore, a set of spatiotemporal constraints (first feature constraints) that conform to actual travel scenarios can be constructed for the incomplete travel chain of the whole feature missing type.

[0082] In one implementation, the preset feature missing types include a tail segment feature missing type. Based on the target travel features of the first historical travel chain and the target travel features of the second historical travel chain, feature constraints corresponding to the preset feature missing types are constructed, including: constructing second feature constraints, which include: the starting position of the target travel chain is the starting position of the first historical travel chain; the start time of the target travel chain is the start time of the first historical travel chain; the distance between the ending position of the target travel chain and the starting position of the second historical travel chain is less than or equal to a preset distance threshold; and the ending time of the target travel chain is between the ending time of the first historical travel chain and the start time of the second historical travel chain.

[0083] In other words, if the type is missing tail features, then the trip chain to be completed is the first historical trip chain. i It needs to be based on the second historical travel chain trip i+1 The starting point travel characteristics are set to determine the destination travel characteristics of the travel chain.

[0084] For example, if the tail feature is missing, the constructed second feature constraint includes:

[0085] (1) The starting position of the target travel chain: start_lnglat i_new For the first historical travel chain trip i The starting position of start_lnglat i :

[0086] start_lnglat i_new =start_lnglat i ;

[0087] (2) Start time of the target travel chain i_new For the first historical travel chain trip i start_time i :

[0088] start_time i_new =start_time i ;

[0089] (3) End position of the target travel chain end_lnglat i_new With the second historical travel chain trip i+1 The starting position of start_lnglat i+1 The distance between them is less than or equal to a preset distance threshold. dis :

[0090] f_d(end_lnglati_new ,start_lnglat i+1 )≤thres dis ;

[0091] (4) End time of the target travel chain i_new Trip is in the first historical travel chain i end_time i With the second historical travel chain trip i+1 start_time i+1 between:

[0092] end_time i_new ∈[end_time i start_time i+1 ].

[0093] Therefore, a set of spatiotemporal constraints (second feature constraints) that conforms to actual travel scenarios can be constructed for travel chains that are missing tail features.

[0094] In one implementation, the preset feature missing types include a first-segment feature missing type. Based on the target travel features of the first historical travel chain and the target travel features of the second historical travel chain, feature constraints corresponding to the preset feature missing types are constructed, including: constructing a third feature constraint, which includes: the distance between the starting position of the target travel chain and the ending position of the first historical travel chain is less than or equal to a preset distance threshold; the start time of the target travel chain is between the end time of the first historical travel chain and the start time of the second historical travel chain; the ending position of the target travel chain is the ending position of the second historical travel chain; and the end time of the target travel chain is the end time of the second historical travel chain.

[0095] In other words, if the first segment of the travel chain is missing, then the travel chain to be completed is the second historical travel chain (trip). i+1 It needs to be based on the first historical travel chain trip i+1 The destination travel characteristics are set to the starting travel characteristics of the target travel chain.

[0096] For example, if the first segment feature is missing, the constructed third feature constraint includes:

[0097] (1) The starting position of the target travel chain: start_lnglat i_new With the first historical travel chain trip i End position end_lnglat i The distance between them is less than or equal to a preset distance threshold. dis :

[0098] f_d(start_lnglat i_new end_lnglat i )≤thres dis ;

[0099] (2) Start time of the target travel chain i_new Trip is in the first historical travel chain i end_time i With the second historical travel chain trip i+1 start_time i+1 between:

[0100] start_time i_new ∈[end_time i start_time i+1 ];

[0101] (3) End position of the target travel chain end_lnglat i_new For the second historical travel chain trip i+1 End position end_lnglat i+1 :

[0102] end_lnglat i_new =end_lnglat i+1 ;

[0103] (4) End time of the target travel chain i_new For the second historical travel chain trip i+1 end_time i+1 between:

[0104] end_time i_new =end_time i+1 .

[0105] Therefore, a set of spatiotemporal constraints (second feature constraints) that conforms to actual travel scenarios can be constructed for the incomplete travel chain of the first segment with missing features.

[0106] In one implementation, in step S202, determining the target travel feature set that meets the feature constraints from the historical travel feature set may include: determining multiple candidate travel feature sets that meet the feature constraints from multiple historical travel feature sets; and selecting the candidate travel feature set with the highest similarity to the target travel feature set as the target travel feature set.

[0107] For example, the feature constraints include constraints on multiple target travel features, thereby identifying a set of historical travel features that meet the feature constraints, which serves as the candidate travel feature set. Each historical travel feature in the candidate travel feature set is then used to construct a candidate travel feature matrix, thus each candidate travel feature set corresponds to a candidate travel feature matrix. The feature constraints include constraints on multiple target travel features, forming a spatiotemporal constraint feature matrix. The similarity between each candidate travel feature matrix and the spatiotemporal constraint feature matrix can be calculated, and then the candidate travel feature set corresponding to the candidate travel feature matrix with the highest similarity can be selected as the target travel feature set. The similarity calculation can utilize the K-Nearest Neighbor (KNN) classification algorithm.

[0108] In one implementation, in step S203, generating a target travel chain using multiple historical travel features from the target travel feature set may include: inputting multiple historical travel features from the target travel feature set into a pre-trained travel chain generation model to obtain the target travel chain; wherein the travel chain generation model is obtained by training a generative adversarial network based on multiple historical travel feature sets.

[0109] Historical travel feature sets, derived from users' historical travel information, cover various travel scenarios and therefore possess rich travel characteristic information. Thus, these historical travel feature sets can be used to train an initial deep learning model to obtain a travel chain generation model. Then, multiple historical travel features from the target travel feature set are input into the travel chain generation model to obtain the target travel chain output by the model.

[0110] For example, the initial deep learning model can be a supervised learning network or an unsupervised learning network, such as a GAN network. The training method can be one familiar to those skilled in the art, and will not be described in detail here.

[0111] The travel chain generation model generates travel chain information that conforms to the user's actual travel patterns, making the generated travel features more consistent with the actual travel feature distribution. It can also complete multiple travel features such as origin location, destination location, start time, end time, travel duration, travel distance, travel mode, travel route, and travel purpose, rather than being limited to the target travel features in the feature constraints.

[0112] In one embodiment, after generating the target travel chain in step S203, the method of this application embodiment may further include: adding the target travel chain to the corresponding historical travel chain set in response to the feature missing type of the travel chain to be completed being the whole segment feature missing type; or, updating the historical travel chain set where the travel chain to be completed is located using the target travel chain in response to the feature missing type of the travel chain to be completed being the tail segment feature missing type or the first segment feature missing type.

[0113] Understandably, once the target travel chain is generated, the actual type of missing features in the travel chain to be completed can be determined. For example, the type of missing features can be determined by comparing the travel features of the target travel chain with those of the first or second historical travel chain, or by determining the type of missing features based on the similarity between the target travel chain and the feature constraints.

[0114] For example, if the missing feature type of the travel chain to be completed is the whole segment missing feature type, a clustering algorithm can be used to cluster the target travel chain as a new historical travel chain into the corresponding historical travel chain set; if the missing feature type of the travel chain to be completed is the tail segment missing feature type or the first segment missing feature type, the target travel chain can be used to update the travel chain to be completed, and a clustering algorithm can be used on the updated travel chain to redetermine the historical travel chain set to which it belongs.

[0115] Based on this, the historical travel feature set can be continuously updated and enriched. Each updated historical travel feature set can be used to train the travel chain generation model, thereby improving the accuracy of the travel chain generation model.

[0116] In one embodiment, the method of this application embodiment may include: before step S201, identifying historical location service data to obtain multiple historical travel chains and travel characteristics of each historical travel chain; wherein, the travel characteristics of the historical travel chain include at least one of the following: origin location, destination location, start time, end time, travel duration, travel distance, travel mode, travel route, and travel purpose; clustering each historical travel chain based on the similarity of the travel characteristics of each historical travel chain to generate multiple historical travel chain sets and multiple historical travel characteristic sets.

[0117] Historical location service data can be historical LBS data. Based on a travel chain identification algorithm, travel chains can be identified from the historical LBS data of multiple users, generating travel information such as origin-destination (OD), mode of transport, purpose of travel, and route for all users. This yields a set of stop points for each traveler (user) within a day. Trajectory points between nearby stop points of the same user are then linked in chronological order to form a travel chain for that traveler, and all travel information is aggregated into the travel characteristics of the complete travel chain.

[0118] The travel chains of the same user over multiple days are sorted according to the start time of the travel chain to obtain multiple historical travel chains of the same user over multiple days. Then, the historical travel chains with similar travel characteristics are aggregated into the same historical travel chain set by the clustering algorithm, thereby obtaining one or more historical travel chain sets. The travel characteristics are extracted for each historical travel chain set to obtain the corresponding historical travel feature set. Each historical travel feature set C includes multiple historical travel features.

[0119] For example, the travel features of each historical travel chain can be constructed into a travel feature matrix. Multiple historical travel chain travel feature matrices are clustered according to their similarity, thus grouping historical travel chains whose similarity to the travel feature matrices meets a threshold into the same historical travel chain set. The clustered travel feature matrix is ​​then the historical travel feature set corresponding to that historical travel chain set. The clustering algorithm can be a hierarchical clustering algorithm, which analyzes the data at different levels based on the similarity between travel feature matrices, thereby forming a tree-like clustering structure.

[0120] The following is combined Figure 5 An application example of an embodiment of this application is introduced. For example... Figure 5 As shown, this application example mainly includes:

[0121] Step S501: Travel data preprocessing, including: identification of travel characteristics of historical travel chains; sorting the travel chains of the same user over multiple days according to the start time of the travel chain to obtain multiple historical travel chains of the same user over multiple days.

[0122] Step S502: Construction of historical travel feature database, including: obtaining multiple historical travel chains of the same user over multiple days, aggregating historical travel chains with similar travel characteristics into the same historical travel chain set based on hierarchical clustering algorithm; extracting travel features from each historical travel chain set to obtain the corresponding historical travel feature set.

[0123] Step S503: Determining missing travel features, including: calculating the connection distance of adjacent travel chains, such as calculating the first historical travel chain (trip). i Second historical travel chain trip i+1 If the connection distance is greater than the preset distance threshold, then there is a travel chain that needs to be completed.

[0124] Step S504: Spatiotemporal constraint set construction, including: determining the feature missing type of the travel chain to be completed; feature constraints of the target travel chain can be constructed for three different feature missing types, that is, spatiotemporal constraint sets can be constructed for three different travel missing situations.

[0125] Step S505: Complete missing features and generate target travel chain, including: based on the KNN algorithm, determine the target travel feature set that best matches the feature constraints (i.e. has the highest similarity to the target travel features) from the historical travel feature database; input multiple historical travel features from the target travel feature set into the pre-trained travel chain generation model to obtain the target travel chain, thereby completing the missing features.

[0126] Example 2

[0127] This application provides a method for processing travel chains, wherein at least some of the technical solutions and corresponding technical effects of Embodiment 1 can be applied to this application. Figure 6 A flowchart illustrating a trip chain processing method according to an embodiment of this application is provided. This trip chain processing method can be applied to a trip chain processing apparatus. Figure 6 As shown, the processing method for this travel chain includes:

[0128] Step S601: Obtain the target travel feature set; wherein, each historical travel feature in the target travel feature set is determined based on the historical travel chain set corresponding to the target travel feature set, and the historical travel chain set includes multiple historical travel chains with similar travel features;

[0129] Step S602: Input multiple historical travel features from the target travel feature set into the pre-trained travel chain generation model to obtain the target travel chain.

[0130] Historical travel feature sets, derived from users' historical travel information, cover various travel scenarios and therefore possess rich travel characteristic information. Thus, these historical travel feature sets can be used to train an initial deep learning model to obtain a travel chain generation model. Then, multiple historical travel features from the target travel feature set are input into the travel chain generation model to obtain the target travel chain output by the model.

[0131] According to the technical solution of this application embodiment, a travel chain generation model is used to generate travel chain information that conforms to the actual travel pattern distribution of the user, so that the generated travel features are more consistent with the actual travel feature distribution, and can complete multiple travel features such as starting position, ending position, start time, end time, travel duration, travel distance, travel mode, travel route, and travel purpose, and is not limited to the target travel features in the feature constraints.

[0132] Example 3

[0133] This application provides a method for processing travel chains, wherein at least some of the technical solutions and corresponding technical effects of embodiments one and two can be referenced in this application. Figure 7A flowchart illustrating a trip chain processing method according to an embodiment of this application is provided. This trip chain processing method can be applied to a trip chain processing apparatus. Figure 7 As shown, the processing method for this travel chain includes:

[0134] Step S701: Obtain the historical travel feature set; wherein, each historical travel feature in the historical travel feature set is determined based on the historical travel chain set corresponding to the historical travel feature set, and the historical travel chain set includes multiple historical travel chains with similar travel features;

[0135] Step S702: Using multiple historical travel features from the historical travel feature set, train an initial deep learning model to obtain a travel chain generation model for generating the target travel chain.

[0136] The initial deep learning model can be a supervised learning network or an unsupervised learning network, such as a GAN network. The training method can be one familiar to those skilled in the art, and will not be elaborated here.

[0137] Historical travel feature sets originate from users' historical travel information and cover a variety of travel scenarios. Therefore, they have richer travel feature information. Consequently, travel chain generation models trained based on historical travel feature sets can complete various types of missing features and generate target travel chains, thereby improving the continuity, completeness, and accuracy of travel chain reconstruction or completion.

[0138] Example 4

[0139] Corresponding to the application scenarios provided in the embodiments of this application and the method in Embodiment 1, the embodiments of this application also provide a trip chain processing device. For example... Figure 8 As shown, the trip chain processing device may include: a feature constraint determination module 801, used to determine the feature constraints of a target trip chain based on a first historical trip chain and a second historical trip chain that are adjacent in time sequence; wherein, the target trip chain is the trip chain after the missing features of the trip chain to be completed are filled; the feature constraints include constraints that at least one target trip feature in the target trip chain must satisfy, and the target trip feature includes a missing feature; the trip chain to be completed is the first historical trip chain or the second historical trip chain or a trip chain located between the first historical trip chain and the second historical trip chain in time sequence; a target trip feature set determination module 802, used to determine a target trip feature set that satisfies the feature constraints from a set of historical trip features; wherein, each historical trip feature in the set of historical trip features is determined based on a set of historical trip chains corresponding to the set of historical trip features, and the set of historical trip chains includes historical trip chains with similar trip features; and a target trip chain generation module 803, used to generate a target trip chain using multiple historical trip features in the target trip feature set.

[0140] In one implementation, the feature constraint determination module 801 is specifically used to: in response to the connection distance between the first historical travel chain and the second historical travel chain being greater than a preset distance threshold, construct feature constraints corresponding to preset feature missing types based on the travel characteristics of the first historical travel chain and the travel characteristics of the second historical travel chain; wherein, the target travel characteristics include the starting position, start time, ending position and end time, and the preset feature missing types include any one of the following: whole segment feature missing type, tail segment feature missing type and first segment feature missing type.

[0141] In one implementation, the preset feature missing type includes the whole segment feature missing type, and the feature constraint determination module 801 is specifically used to: construct a first feature constraint, the first feature constraint including: the distance between the starting position of the target travel chain and the ending position of the first historical travel chain is less than or equal to a preset distance threshold; the start time and end time of the target travel chain are both between the end time of the first historical travel chain and the start time of the second historical travel chain; and the distance between the ending position of the target travel chain and the starting position of the second historical travel chain is less than or equal to a preset distance threshold.

[0142] In one implementation, the preset feature missing type includes the tail segment feature missing type. The feature constraint determination module 801 is specifically used to: construct a second feature constraint, which includes: the starting position of the target travel chain is the starting position of the first historical travel chain; the start time of the target travel chain is the start time of the first historical travel chain; the distance between the ending position of the target travel chain and the starting position of the second historical travel chain is less than or equal to a preset distance threshold; and the ending time of the target travel chain is between the ending time of the first historical travel chain and the start time of the second historical travel chain.

[0143] In one implementation, the preset feature missing type includes the first segment feature missing type. The feature constraint determination module 801 is specifically used to: construct a third feature constraint, which includes: the distance between the starting position of the target travel chain and the ending position of the first historical travel chain is less than or equal to a preset distance threshold; the start time of the target travel chain is between the end time of the first historical travel chain and the start time of the second historical travel chain; the ending position of the target travel chain is the ending position of the second historical travel chain; and the end time of the target travel chain is the end time of the second historical travel chain.

[0144] In one implementation, the target travel feature set determination module 802 is specifically used to: determine multiple candidate travel feature sets that meet the feature constraints from multiple historical travel feature sets; and select the candidate travel feature set with the highest similarity to the target travel feature set as the target travel feature set.

[0145] In one implementation, the target travel chain generation module 803 is specifically used to: input multiple historical travel features from the target travel feature set into a pre-trained travel chain generation model to obtain the target travel chain; wherein, the travel chain generation model is obtained by training a generative adversarial network based on multiple historical travel feature sets.

[0146] In one embodiment, the trip chain processing device may further include a historical trip chain set update module, which, after generating the target trip chain, adds the target trip chain to the corresponding historical trip chain set in response to the feature missing type of the trip chain to be completed being the whole segment feature missing type; or updates the historical trip chain set where the trip chain to be completed is located using the target trip chain in response to the feature missing type of the trip chain to be completed being the tail segment feature missing type or the first segment feature missing type.

[0147] In one embodiment, the processing device for the travel chain may further include: an identification module, used to identify historical location service data to obtain multiple historical travel chains and travel characteristics of each historical travel chain; wherein the travel characteristics of the historical travel chain include at least one of the following: origin location, destination location, start time, end time, travel duration, travel distance, travel mode, travel route, and travel purpose; and a clustering module, used to cluster each historical travel chain based on the similarity of its travel characteristics to generate multiple sets of historical travel chains and multiple sets of historical travel characteristics.

[0148] The functions of each module in each device in the embodiments of this application can be found in the corresponding description in the above method, and they have corresponding beneficial effects, which will not be repeated here.

[0149] Example 5

[0150] Corresponding to the application scenarios provided in the embodiments of this application and the method in Embodiment 2, the embodiments of this application also provide a travel chain processing device. This travel chain processing device may include: a target travel feature set acquisition module, used to acquire a target travel feature set; wherein each historical travel feature in the target travel feature set is determined based on a set of historical travel chains corresponding to the target travel feature set, and the set of historical travel chains includes multiple historical travel chains with similar travel features; and an input module, used to input multiple historical travel features from the target travel feature set into a pre-trained travel chain generation model to obtain the target travel chain.

[0151] The functions of each module in each device in the embodiments of this application can be found in the corresponding description in the above method, and they have corresponding beneficial effects, which will not be repeated here.

[0152] Example 6

[0153] Corresponding to the application scenarios provided in the embodiments of this application and the method in Embodiment 3, the embodiments of this application also provide a travel chain processing device. This travel chain processing device may include: a historical travel feature set acquisition module, used to acquire a historical travel feature set; wherein each historical travel feature in the historical travel feature set is determined based on a set of historical travel chains corresponding to the historical travel feature set, and the set of historical travel chains includes multiple historical travel chains with similar travel features; and a training module, used to train an initial deep learning model using multiple historical travel features in the historical travel feature set to obtain a travel chain generation model for generating a target travel chain.

[0154] The functions of each module in each device in the embodiments of this application can be found in the corresponding description in the above method, and they have corresponding beneficial effects, which will not be repeated here.

[0155] Example 7

[0156] Figure 9 This is a block diagram of an electronic device used to implement embodiments of this application. Figure 9 As shown, the electronic device includes a memory 901 and a processor 902. The memory 901 stores a computer program that can run on the processor 902. When the processor 902 executes the computer program, it implements the method described in the above embodiments. The number of memories 901 and processors 902 can be one or more.

[0157] The electronic device also includes:

[0158] The communication interface 903 is used to communicate with external devices and exchange and transmit data.

[0159] If the memory 901, processor 902, and communication interface 903 are implemented independently, they can be interconnected via a bus to communicate with each other. This bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 9 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0160] Optionally, in a specific implementation, if the memory 901, processor 902, and communication interface 903 are integrated on a single chip, then the memory 901, processor 902, and communication interface 903 can communicate with each other through an internal interface.

[0161] This application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method provided in this application.

[0162] This application also provides a chip, which includes a processor for calling and executing instructions stored in a memory, causing a communication device with the chip installed to perform the method provided in this application.

[0163] This application also provides a chip, including: an input interface, an output interface, a processor, and a memory. The input interface, output interface, processor, and memory are connected through an internal connection path. The processor is used to execute code in the memory. When the code is executed, the processor is used to execute the method provided in the application embodiment.

[0164] It should be understood that the aforementioned processor can be a CPU, or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), FPGAs, or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. General-purpose processors can be microprocessors or any conventional processor. It is worth noting that the processor can be a processor supporting Advanced Reduced Instruction Set Machines (ARM) architecture.

[0165] Further, optionally, the aforementioned memory may include read-only memory and random access memory. The memory may be volatile memory or non-volatile memory, or may include both. Non-volatile memory may include read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory may include random access memory (RAM), which serves as an external cache. By way of example, but not limitation, many forms of RAM are available. Examples include Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), and Direct Rambus RAM (DR RAM).

[0166] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions according to this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another.

[0167] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0168] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "a plurality of" means two or more, unless otherwise explicitly specified.

[0169] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process. Furthermore, the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functionality involved.

[0170] The logic and / or steps described in the flowchart or otherwise herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus or device (such as a computer-based system, a processor-included system or other system that can fetch and execute instructions from, an instruction execution system, apparatus or device).

[0171] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. All or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware, the program being stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiments.

[0172] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. This storage medium can be a read-only memory, a disk, or an optical disk, etc.

[0173] The above description is merely an exemplary embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope described in this application, and these should all be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for processing travel chains, comprising: Based on the first and second historical travel chains that are adjacent in time sequence, the feature constraints of the target travel chain are determined; wherein, the target travel chain is the travel chain after the missing features of the travel chain to be completed are filled; the feature constraints include constraints that at least one target travel feature in the target travel chain must satisfy, and the target travel feature includes the missing feature; the travel chain to be completed is the first historical travel chain or the second historical travel chain or a travel chain that is located between the first historical travel chain and the second historical travel chain in time sequence; From the set of historical travel features, a target travel feature set that satisfies the feature constraints is determined; wherein each historical travel feature in the set of historical travel features is determined based on a set of historical travel chains corresponding to the set of historical travel features, and the set of historical travel chains includes historical travel chains with similar travel features; The target travel chain is generated by utilizing multiple historical travel features from the target travel feature set.

2. The processing method according to claim 1, wherein, Based on the first and second historical travel chains that are close in time sequence, the characteristic constraints of the target travel chain are determined, including: In response to the connection distance between the first historical travel chain and the second historical travel chain being greater than a preset distance threshold, at least one feature constraint corresponding to a preset feature missing type is constructed based on the travel characteristics of the first historical travel chain and the travel characteristics of the second historical travel chain; wherein, the target travel characteristics include the starting position, start time, ending position and end time, and the preset feature missing type includes at least one of the following: whole segment feature missing type, tail segment feature missing type and first segment feature missing type.

3. The processing method according to claim 2, wherein, The preset feature missing types include whole-segment feature missing types. Based on the target travel features of the first historical travel chain and the target travel features of the second historical travel chain, feature constraints corresponding to the preset feature missing types are constructed, including: A first feature constraint is constructed, which includes: the distance between the starting position of the target travel chain and the ending position of the first historical travel chain is less than or equal to the preset distance threshold; the start time and end time of the target travel chain are both between the end time of the first historical travel chain and the start time of the second historical travel chain; and the distance between the ending position of the target travel chain and the starting position of the second historical travel chain is less than or equal to the preset distance threshold.

4. The processing method according to claim 2, wherein, The preset feature missing types include tail segment feature missing types. Based on the target travel features of the first historical travel chain and the target travel features of the second historical travel chain, feature constraints corresponding to the preset feature missing types are constructed, including: A second feature constraint is constructed, which includes: the starting position of the target travel chain is the starting position of the first historical travel chain; the start time of the target travel chain is the start time of the first historical travel chain; the distance between the ending position of the target travel chain and the starting position of the second historical travel chain is less than or equal to the preset distance threshold; and the ending time of the target travel chain is between the ending time of the first historical travel chain and the start time of the second historical travel chain.

5. The processing method according to claim 2, wherein, The preset feature missing types include the first segment feature missing type. Based on the target travel features of the first historical travel chain and the target travel features of the second historical travel chain, feature constraints corresponding to the preset feature missing types are constructed, including: A third feature constraint is constructed, which includes: the distance between the starting position of the target travel chain and the ending position of the first historical travel chain is less than or equal to the preset distance threshold; the start time of the target travel chain is between the end time of the first historical travel chain and the start time of the second historical travel chain; the ending position of the target travel chain is the ending position of the second historical travel chain; and the end time of the target travel chain is the end time of the second historical travel chain.

6. The processing method according to claim 1, wherein, From the historical travel feature set, a target travel feature set that satisfies the aforementioned feature constraints is determined, including: From multiple historical travel feature sets, determine multiple candidate travel feature sets that satisfy the aforementioned feature constraints; Select the set of candidate travel features that has the highest similarity to the target travel features as the target travel feature set.

7. The processing method according to any one of claims 1 to 6, wherein, The target travel chain is generated using multiple historical travel features from the target travel feature set, including: The target travel chain is obtained by inputting multiple historical travel features from the target travel feature set into a pre-trained travel chain generation model; wherein the travel chain generation model is obtained by training a generative adversarial network based on multiple historical travel feature sets.

8. The processing method according to any one of claims 1 to 6, further comprising, after generating the target travel chain: In response to the fact that the feature missing type of the travel chain to be completed is the whole segment feature missing type, the target travel chain is added to the corresponding historical travel chain set; or... In response to the missing feature type of the travel chain to be completed being either the tail segment missing feature type or the first segment missing feature type, the historical travel chain set to which the travel chain to be completed belongs is updated using the target travel chain.

9. The method according to any one of claims 1 to 6, further comprising: Identify historical location service data to obtain multiple historical travel chains and the travel characteristics of each historical travel chain; wherein, the travel characteristics of the historical travel chain include at least one of the following: origin location, destination location, start time, end time, travel duration, travel distance, travel mode, travel route, and travel purpose; Based on the similarity of the travel characteristics of each historical travel chain, the historical travel chains are clustered to generate multiple sets of historical travel chains and multiple sets of historical travel characteristics.

10. A processing device for a travel chain, comprising: The feature constraint determination module is used to determine the feature constraints of a target travel chain based on a first historical travel chain and a second historical travel chain that are adjacent in time sequence; wherein, the target travel chain is the travel chain after the missing features of the travel chain to be completed are filled; the feature constraints include constraints that at least one target travel feature in the target travel chain must satisfy, and the target travel feature includes the missing feature; the travel chain to be completed is the first historical travel chain or the second historical travel chain, or a travel chain that is located between the first historical travel chain and the second historical travel chain in time sequence; The target travel feature set determination module is used to determine a target travel feature set that satisfies the feature constraints from a historical travel feature set; wherein, each historical travel feature in the historical travel feature set is determined based on a set of historical travel chains corresponding to the historical travel feature set, and the set of historical travel chains includes historical travel chains with similar travel features; The target travel chain generation module is used to generate the target travel chain by utilizing multiple historical travel features in the target travel feature set.

11. An electronic device comprising a memory, a processor, and a computer program stored in the memory, wherein the processor, when executing the computer program, implements the method of any one of claims 1-9.

12. A computer-readable storage medium storing a computer program that, when executed by a processor, implements the method of any one of claims 1-9.

Images