Positioning method and apparatus for trajectory problem

Abstract

The present disclosure provides a positioning method for trajectory problems, relates to the technical field of high-precision map, and particularly relates to the technical field of open platforms and trajectory mining. The method comprises: obtaining a trajectory problem, and calling a large language model to plan and decompose the trajectory problem to obtain one or more trajectory sub-problems and task information of each trajectory sub-problem; for each trajectory sub-problem, calling a target tool for solving the trajectory sub-problem based on the task information, and analyzing the trajectory sub-problem based on the target tool to obtain a reply result of the trajectory sub-problem; and determining a positioning result for the trajectory problem according to the reply result of each trajectory sub-problem. The present disclosure integrates technologies such as artificial intelligence, machine learning and data analysis, can avoid waste of time resources, improve the efficiency and accuracy of trajectory problem positioning, improve the flexibility of trajectory mining, and more quickly and intelligently solve complex system problems.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of high-precision maps, and in particular to the technical field of open platforms and trajectory mining. BACKGROUND

[0002] In related technologies, when positioning a trajectory problem, since the trajectory problem may involve multiple system modules or interactive components, each part needs to be checked step by step, which wastes time and reduces the positioning accuracy of the trajectory problem. In addition, since the trajectory problem may be a complex of multiple sub-problems, when positioning the trajectory problem, multiple interrelated problems need to be solved, which increases the difficulty of trajectory problem positioning and repair and reduces the efficiency of trajectory problem positioning.

[0003] Therefore, how to avoid wasting time resources, improve the efficiency and accuracy of trajectory problem positioning, improve the flexibility of trajectory mining, and more quickly and intelligently solve complex system problems has become one of the important research directions. SUMMARY

[0004] The present disclosure provides a positioning method for a trajectory problem.

[0005] According to an aspect of the present disclosure, a positioning method for a trajectory problem is provided, which is executed by an agent device, and the method comprises:

[0006] Obtaining a trajectory problem, and calling a large language model to plan and decompose the trajectory problem to obtain one or more trajectory sub-problems and task information of each trajectory sub-problem;

[0007] For each trajectory sub-problem, calling a target tool for solving the trajectory sub-problem based on the task information, and analyzing the trajectory sub-problem based on the target tool to obtain a reply result of the trajectory sub-problem;

[0008] According to the reply result of each trajectory sub-problem, a positioning result for the trajectory problem is determined.

[0009] In the present disclosure, the large language model is called to plan and decompose the trajectory problem to obtain one or more trajectory sub-problems and task information of each trajectory sub-problem, which can solve multiple interrelated problems at the same time, thereby avoiding wasting time resources. For each trajectory sub-problem, a target tool for solving the trajectory sub-problem is called based on the task information, which realizes the integration of technologies such as artificial intelligence, machine learning and data analysis, analyzes the trajectory sub-problem based on tool resources in different fields, and obtains a reply result of the trajectory sub-problem, which can avoid errors caused by manual operation and improve the efficiency and accuracy of trajectory problem positioning.

[0010] According to another aspect of the present disclosure, a positioning method for a trajectory problem is provided, executed by a target tool, comprising:

[0011] receiving a request instruction sent by the agent device;

[0012] parsing the request instruction, in response to the request instruction being a first request instruction containing a trajectory sub-problem, analyzing the trajectory sub-problem, obtaining a reply result of the trajectory sub-problem, and sending the reply result of the trajectory sub-problem to the agent device;

[0013] in response to the request instruction being a preset second request instruction, sending a first trajectory query vector in a trajectory query database to the agent device.

[0014] In the embodiments of the present disclosure, in the case where the problem involves multiple fields, the trajectory problem can be analyzed by different tools to call knowledge and resources in different fields to solve the trajectory problem, realize the cooperation of multiple fields, and more quickly and intelligently solve complex trajectory problems, thereby improving the flexibility of trajectory mining.

[0015] According to another aspect of the present disclosure, a positioning device for a trajectory problem is provided, comprising:

[0016] an obtaining module configured to obtain a trajectory problem and call a large language model to plan and decompose the trajectory problem to obtain one or more trajectory sub-problems and task information of each trajectory sub-problem;

[0017] an analysis module configured to, for each trajectory sub-problem, call a target tool for solving the trajectory sub-problem based on the task information, and analyze the trajectory sub-problem based on the target tool to obtain a reply result of the trajectory sub-problem;

[0018] a determination module configured to determine a positioning result for the trajectory problem according to the reply result of each trajectory sub-problem.

[0019] According to another aspect of the present disclosure, a positioning device for a trajectory problem is provided, comprising:

[0020] a receiving module configured to receive a request instruction sent by an agent device;

[0021] a first processing module configured to parse the request instruction, in response to the request instruction being a first request instruction containing a trajectory sub-problem, analyze the trajectory sub-problem, obtain a reply result of the trajectory sub-problem, and send the reply result of the trajectory sub-problem to the agent device;

[0022] a second processing module configured to, in response to the request instruction being a preset second request instruction, send a first trajectory query vector in a trajectory query database to the agent device.

[0023] According to another aspect of the present disclosure, an electronic device is provided, comprising at least one processor, and

[0024] a memory connected with the at least one processor in communication; wherein,

[0025] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the positioning method for trajectory problems according to the first aspect of the present disclosure, or perform the positioning method for trajectory problems according to the second aspect of the present disclosure.

[0026] According to another aspect of the present disclosure, a non-transitory computer readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable a computer to perform the positioning method for trajectory problems according to the first aspect of the present disclosure, or perform the positioning method for trajectory problems according to the second aspect of the present disclosure.

[0027] According to another aspect of the present disclosure, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the steps of the positioning method for trajectory problems according to the first aspect of the present disclosure, or implements the steps of the positioning method for trajectory problems according to the second aspect of the present disclosure.

[0028] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present disclosure, nor to limit the scope of the present disclosure. Other features of the present disclosure will become apparent through the following description. BRIEF DESCRIPTION OF DRAWINGS

[0029] The accompanying drawings are used to better understand the present scheme, and do not limit the present disclosure. Among them:

[0030] Figure 1 is a structure diagram of a trajectory mining system according to an embodiment of the present disclosure;

[0031] Figure 2 is a flowchart of a positioning method for trajectory problems according to an embodiment of the present disclosure;

[0032] Figure 3 is a flowchart of a positioning method for trajectory problems according to an embodiment of the present disclosure;

[0033] Figure 4 is a flowchart of a positioning method for trajectory problems according to an embodiment of the present disclosure;

[0034] Figure 5 is a flowchart of a positioning method for trajectory problems according to an embodiment of the present disclosure;

[0035] Figure 6 is a schematic diagram of a positioning method for a trajectory problem according to an embodiment of the present disclosure;

[0036] Figure 7 is a flowchart of a positioning method for a trajectory problem according to an embodiment of the present disclosure;

[0037] Figure 8 is a structural diagram of a positioning device for a trajectory problem according to an embodiment of the present disclosure;

[0038] Figure 9 is a structural diagram of a positioning device for a trajectory problem according to an embodiment of the present disclosure;

[0039] Figure 10 is a block diagram of an electronic device for implementing the method according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0040] Exemplary embodiments of the present disclosure are described herein with reference to the accompanying drawings, which are included to provide a thorough understanding of embodiments of the present disclosure by a person of ordinary skill in the art, and should be considered in connection with the following detailed description, and should not be considered as a limitation to the scope of the present disclosure. Thus, it will be appreciated that those skilled in the art will be able to devise various arrangements that, although not explicitly described or shown herein, embody the principles of the present disclosure and, as such, have structural and / or functional equivalents that are within the scope of the present disclosure. Similarly, it will be appreciated that, where possible, the description and drawings have been simplified and generalizations are included in the description, like those common to the skills in the art. Thus, those skilled in the art will recognize that the application described herein can be embodied in other specific forms without departing from the spirit or essential characteristics of the present disclosure.

[0041] Embodiments of the present disclosure relate to the technical field of high-precision map, and in particular to the technical field of open platform, trajectory mining, etc.

[0042] The principle of high-precision map technology is mainly to use various sensor technologies, such as satellite navigation, laser radar, camera, millimeter wave radar, inertial measurement unit, etc., to collect and process roads, buildings, traffic facilities, etc. with high precision and multi-angle, form digital map data, and fuse the map data with real-time positioning data to provide map information with precision and real-time.

[0043] Open platform (Open Platform) in the software industry and network, open platform refers to the software system through the public application programming interface (API) or function (function) to make external programs can increase the function of the software system or use the resources of the software system, without changing the source code of the software system. In the Internet era, the services of the website are packaged into a series of computer-identifiable data interfaces and opened to the public for third-party developers to use. This behavior is called Open API, and the platform that provides open API is called open platform.

[0044] Trajectory mining system, a continuous sequence of coordinate points of a vehicle is called a trajectory. With trajectory data, a trajectory mining system can help analyze and mine trajectory information from geographic location data, providing valuable insights and decision support for various fields, including:

[0045] Traffic management and planning, trajectory mining systems can analyze trajectory data of vehicles, pedestrians, public transportation, etc., helping traffic management departments better understand traffic flow, congestion, and travel patterns. This helps optimize road planning, traffic signal control, and improve traffic efficiency and safety.

[0046] Navigation and route planning, based on trajectory data analysis, trajectory mining systems can provide more intelligent navigation and route planning services. It can consider real-time traffic conditions, travel time, optimal routes, etc., to provide accurate navigation recommendations for users.

[0047] Urban planning and intelligence, trajectory mining systems can help urban planners understand residents' travel habits, activity ranges, and needs, optimizing urban planning and facility layout. This plays an important role in the intelligent development of cities.

[0048] The positioning method for trajectory problems and its device of the present disclosure will be described below in combination with the reference drawings.

[0049] First, the trajectory mining system in the embodiments of the present disclosure is introduced, Figure 1 is a structural diagram of the trajectory mining system according to an embodiment of the present disclosure, as Figure 1 shown, the trajectory mining system 100 includes an agent device 110, a tool system 120, a task planning system 130, and an execution system 140, wherein the three subsystems are controlled and information is exchanged through the agent device.

[0050] The task planning system 130 can include a subgoal decomposition expert and a question and answer expert.

[0051] The tool system 120 can include a plurality of tools that can analyze trajectories. Optionally, the tools can be a pre-deployed open platform, such as a trajectory query system, a trajectory content extraction system, a trajectory analysis system, etc., or a pre-deployed database, such as a trajectory query database.

[0052] Figure 2 is a flowchart of the positioning method for trajectory problems according to an embodiment of the present disclosure, as Figure 2 shown, the method includes the following steps:

[0053] S201, obtain a trajectory problem, and call a large language model to plan and decompose the trajectory problem to obtain one or more trajectory sub-problems and task information of each trajectory sub-problem.

[0054] In the embodiments of the present disclosure, the execution subject can be an agent device. The agent device receives a request message such as an http request sent by a terminal device, and obtains a trajectory problem to be located after analyzing the request message.

[0055] As shown in Figure 1 The task planning system 130 includes a large language model. The agent device calls a target decomposition expert to perform task analysis and planning on the trajectory problem through the large language model (Large Language Model, LLM) to obtain one or more trajectory sub-problems and task information of each trajectory sub-problem.

[0056] In the embodiments of the present disclosure, the task information can include tool identification and attribute parameters of a target tool to be called, etc.

[0057] For example, the trajectory problem can be "the trajectory on road ID123 is abnormal, the road has been blocked, and there should be no car trajectory passing through", and the trajectory sub-problems can include three, which are "target 1: query the trajectory of road ID123", "target 2: analyze the queried trajectory and extract the probability of vehicle passing", and "target 3: based on the extracted probability of vehicle passing, statistically analyze and judge the problem reason".

[0058] In some embodiments, the task information further includes a task identification, which can indicate the order of analyzing the trajectory sub-problems based on the target tool, i.e., the analysis order of the trajectory sub-problems. For example, in the embodiments of the present disclosure, the order of the trajectory sub-problems is "target 1: query the trajectory of road ID123", "target 2: analyze the queried trajectory and extract the probability of vehicle passing", and "target 3: based on the extracted probability of vehicle passing, statistically analyze and judge the problem reason".

[0059] S202, for each trajectory sub-problem, call a target tool for solving the trajectory sub-problem based on the task information, and analyze the trajectory sub-problem based on the target tool to obtain a reply result of the trajectory sub-problem.

[0060] As shown in Figure 1As shown, the agent device 110 can call a target tool for solving the trajectory sub-problem based on the task information. If the target tool is a trajectory query database, the agent device obtains attribute parameters of the trajectory query database based on the task information, and combines the attribute parameters and the trajectory sub-problem. In some implementations, the agent device matches the combined trajectory sub-problem with multiple pieces of data in the trajectory query database, and generates a reply result according to one or more pieces of data with a higher matching degree. In some implementations, a question and answer expert is called by the agent device, and the question and answer expert analyzes the trajectory sub-problem based on the trajectory query database to obtain a reply result of the trajectory sub-problem.

[0061] If the target tool is one of a trajectory query system (Traj Fetch System), a trajectory content extraction system (Traj Schema System), and a trajectory analysis system (Traj Stastic System), the target tool is called based on the task information to analyze the trajectory sub-problem, and a reply result of the trajectory sub-problem is obtained.

[0062] S203, according to the reply result of each trajectory sub-problem, a positioning result for the trajectory problem is determined.

[0063] In some embodiments, according to the reply result, an answer to each trajectory sub-problem is obtained, and the answers to each trajectory sub-problem are combined to obtain a positioning result for the trajectory problem.

[0064] In some embodiments, according to the reply result, an answer to each trajectory sub-problem is obtained, and the answers to each trajectory sub-problem are spliced according to the analysis order of each trajectory sub-problem to obtain a positioning result for the trajectory problem.

[0065] Optionally, after obtaining the positioning result for the trajectory problem, the positioning result can be sent to the terminal device.

[0066] In the embodiments of the present disclosure, a large language model is called to plan and decompose the trajectory problem to obtain one or more trajectory sub-problems and task information of the trajectory sub-problems, multiple interrelated problems can be solved at the same time, thereby avoiding waste of time resources, for each trajectory sub-problem, a target tool for solving the trajectory sub-problem is called based on the task information, the fusion of technologies such as artificial intelligence, machine learning, and data analysis is realized, different domain tool resources are used to analyze the trajectory sub-problem, and a reply result of the trajectory sub-problem is obtained, which can avoid errors caused by manual operation and improve the efficiency and accuracy of trajectory problem positioning.

[0067] Figure 3 is a flowchart of a positioning method for a trajectory problem according to an embodiment of the present disclosure, as shown, the method comprises the following steps: Figure 3 ​

[0068] S301, obtain a trajectory problem, and call a large language model to plan and decompose the trajectory problem to obtain one or more trajectory sub-problems and task information of each trajectory sub-problem.

[0069] For the introduction of step S301, please refer to the related content in the above embodiments, which will not be repeated here.

[0070] S302, for each trajectory sub-problem, determining the tool identification of the target tool based on the task information, and calling the target tool for solving the trajectory sub-problem based on the tool identification of the target tool.

[0071] In some embodiments, the task information of the trajectory sub-problem contains the tool identification of the target tool used to solve the trajectory sub-problem.

[0072] In the implementation of the present disclosure, the tool identification of the target tool can be a number, a word or a letter, or any combination of numbers, words or letters, which is not limited in the embodiments of the present disclosure.

[0073] For example, in the embodiments of the present disclosure, the trajectory sub-problem "target 1: query the trajectory of road ID 123" corresponds to the trajectory query system, the trajectory sub-problem "target 2: analyze the queried trajectory and extract the vehicle running probability" corresponds to the trajectory content extraction system, and the trajectory sub-problem "target 3: based on the extracted vehicle running probability, statistically analyze and judge the problem reason" corresponds to the trajectory analysis system.

[0074] S303, analyzing the trajectory sub-problem based on the target tool to obtain a reply result of the trajectory sub-problem.

[0075] For the introduction of step S303, please refer to the related content in the above embodiments, which will not be repeated here.

[0076] S304, determining whether each trajectory sub-problem has an answer according to the reply result, and obtaining a trajectory sub-problem without an answer from the trajectory sub-problems.

[0077] In some embodiments, the reply result of the trajectory sub-problem is analyzed, and if a first identification and text indicating that the trajectory sub-problem has been solved are obtained, it is determined that the trajectory sub-problem has an answer.

[0078] In some embodiments, the reply result of the trajectory sub-problem is analyzed, and if a second identification indicating that the trajectory sub-problem has not been solved is obtained, it is determined that the trajectory sub-problem does not have an answer.

[0079] In some embodiments, if the agent device does not listen to the reply result of the trajectory sub-problem of the target tool within a preset time, it is determined that the trajectory sub-problem does not have an answer.

[0080] S305, continue to call the large language model to plan and decompose the trajectory sub-problems without answers to update the trajectory sub-problems and task information.

[0081] In some embodiments, a first trajectory sub-problem without an answer in the trajectory sub-problems is determined, and an answer to a second trajectory sub-problem other than the first trajectory sub-problem is obtained, the large language model is called to plan and decompose the first trajectory sub-problem and the answer to the second trajectory sub-problem, to obtain a third trajectory sub-problem and task information of the third trajectory sub-problem, the third trajectory sub-problem and the task information of the third trajectory sub-problem are updated as new trajectory sub-problems and task information, and subsequent steps of solving the third trajectory sub-problem are executed until each trajectory sub-problem has a corresponding answer.

[0082] In some embodiments, a first trajectory sub-problem without an answer in the trajectory sub-problems is determined, the large language model is called to plan and decompose the first trajectory sub-problem, to obtain a fourth trajectory sub-problem and task information of the fourth trajectory sub-problem, the first trajectory sub-problem and the task information of the first trajectory sub-problem are replaced by the fourth trajectory sub-problem and the task information of the fourth trajectory sub-problem to update the trajectory sub-problems and the task information, and subsequent steps of solving the fourth trajectory sub-problem are executed until each trajectory sub-problem has a corresponding answer.

[0083] That is, if there is a trajectory sub-problem without an answer, in order to locate the trajectory problem, the first trajectory sub-problem can be re-input into the large language model for planning and decomposition to update the trajectory sub-problems. On the other hand, the answers to the solved trajectory sub-problems can also affect the planning and decomposition process of the large language model. In order to improve accuracy, the answers to the first trajectory sub-problem and the second trajectory sub-problem can also be jointly input into the large language model for planning and decomposition.

[0084] As shown in FIG. 1, Figure 1 In some embodiments, the agent device 110 calls the execution system 140, which in the form of a chain (Chain) calls the task planning system 130 and the tool system 120 in a loop to process the trajectory sub-problems until each trajectory sub-problem has a corresponding answer, wherein the target decomposition expert is used to call the large language model to plan and decompose the trajectory sub-problems, and the tool system is used to call the target tool to analyze the trajectory sub-problems.

[0085] In other implementations, the agent device 110 can also directly call the task planning system 130 and the tool system 120 in a chain (Chain) to process the trajectory sub-problems in a loop until each trajectory sub-problem has a corresponding answer. The present disclosure does not limit this.

[0086] S306, determine the positioning result for the trajectory problem according to the reply result of each trajectory sub-problem.

[0087] The introduction of step S306 can refer to the related content in the above embodiments, which will not be repeated here.

[0088] In the embodiments of the present disclosure, according to the reply result, it is judged whether each trajectory sub-problem has an answer, and the trajectory sub-problem without an answer is obtained from the trajectory sub-problems, and the trajectory sub-problem without an answer is continuously called to plan and decompose by the large language model to update the trajectory sub-problem and the task information. Through the cycle solution of the trajectory sub-problem, the complex trajectory problem can be solved more quickly and more intelligently, and the flexibility of trajectory mining is improved.

[0089] Figure 4 is a flowchart of a positioning method for a trajectory problem according to an embodiment of the present disclosure, as shown in Figure 4 The method comprises the following steps:

[0090] S401, obtain a trajectory problem, and call a large language model to plan and decompose the trajectory problem to obtain one or more trajectory sub-problems and task information of each trajectory sub-problem.

[0091] S402, for each trajectory sub-problem, call a target tool for solving the trajectory sub-problem based on the task information.

[0092] The introduction of steps S401-S402 can refer to the related content in the above embodiments, which will not be repeated here.

[0093] S403, if the target tool is one of the trajectory query system, the trajectory content extraction system and the trajectory analysis system, the attribute parameters of the target tool are obtained based on the task information, and the attribute parameters and the trajectory sub-problem are combined.

[0094] Optionally, if the target tool is the trajectory query system, the attribute parameters of the target tool at least include the specified time information, the specified location range and the information of the terminal device. If the target tool is the trajectory content extraction system or the trajectory analysis system, the attribute parameters of the target tool at least include the vehicle information, the coordinate information of the vehicle and the timestamp information.

[0095] S404, generate a first request instruction according to the combined trajectory sub-problem, and send the first request instruction to the target tool.

[0096] The target tool receives and analyzes the first request instruction, obtains the combined trajectory sub-problem, and the reply result is obtained by the target tool analyzing the combined trajectory sub-problem.

[0097] In some embodiments, if the target tool is a trajectory query system, the trajectory sub-problems combined according to the specified time range, province, terminal device information, etc. are input into the target tool, and then the target tool queries the specified trajectory information to obtain the reply result of the trajectory sub-problems. That is, the combined trajectory sub-problems are sent to the target tool, and the reply result of the trajectory sub-problems sent by the target tool is received, which is obtained by the target tool analyzing the combined trajectory sub-problems.

[0098] In some embodiments, if the target tool is a trajectory content extraction system, the vehicle ID, coordinate information (latitude and longitude), timestamp, vehicle type, and trajectory sub-problems are combined and input into the target tool, and then the target tool extracts the key information of the trajectory, such as trajectory quality information and vehicle information corresponding to the trajectory (probability of being a car), to obtain the reply result of the trajectory sub-problems.

[0099] In some embodiments, if the target tool is a trajectory analysis system, the vehicle ID, coordinate information (latitude and longitude), timestamp, vehicle type, and trajectory sub-problems are combined and input into the target tool, and then the target tool performs statistics on the basic trajectory information and makes judgments based on the statistical information to obtain the reply result of the trajectory sub-problems.

[0100] S405, receiving the reply result of the trajectory sub-problems sent by the target tool.

[0101] The target tool sends the reply result of the trajectory sub-problems to the agent device, and the agent device receives the reply result of the trajectory sub-problems sent by the target tool.

[0102] S406, determining the positioning result of the trajectory problem according to the reply result of each trajectory sub-problem.

[0103] For details of step S406, please refer to the related content in the above embodiments, which will not be repeated here.

[0104] In the embodiments of the present disclosure, the attribute parameters of the target tool are obtained based on the task information, the attribute parameters and the trajectory sub-problems are combined, and the first request instruction is generated and sent to the target tool. The target tool analyzes the combined trajectory sub-problems. In the case of a problem involving multiple fields, different tools are used to analyze and solve the trajectory problem, multiple fields are cooperated, time and resource waste is avoided, and the efficiency and accuracy of trajectory problem positioning are improved.

[0105] Figure 5 is a flowchart of a positioning method for a trajectory problem according to an embodiment of the present disclosure, as shown in Figure 5 the method comprises the following steps:

[0106] S501, obtain a trajectory problem, and call a large language model to plan and decompose the trajectory problem to obtain one or more trajectory sub-problems and task information of each trajectory sub-problem.

[0107] S502, for each trajectory sub-problem, call a target tool for solving the trajectory sub-problem based on the task information.

[0108] The introduction of steps S501-S502 can refer to the related content in the above embodiments, which will not be described here again.

[0109] S503, if the target tool is a pre-deployed trajectory query database, a preset second request instruction is sent to the trajectory query database, and a first trajectory query vector sent by the trajectory query database is received.

[0110] S504, obtain attribute parameters of the target tool based on the task information, and combine the attribute parameters and the trajectory sub-problem.

[0111] In some embodiments, if the target tool is a pre-deployed trajectory query database, the attribute parameters of the target tool based on the task information can be terminal device information, time information, etc.

[0112] S505, embedding mapping is performed on the combined trajectory sub-problem to obtain a second trajectory query vector.

[0113] In some embodiments, embedding is performed on the combined trajectory sub-problem to obtain a second trajectory query vector, so as to subsequently determine a reply result of the trajectory sub-problem from the trajectory query database according to the second trajectory query vector.

[0114] S506, matching is performed on the second trajectory query vector and the first trajectory query vector to obtain a matching degree, and the reply result of the trajectory sub-problem is determined according to the matching degree.

[0115] In the embodiments of the present disclosure, matching is performed on the second trajectory query vector and the first trajectory query vector to obtain a matching degree. Optionally, if there is no third trajectory query vector in the first trajectory query vector that has a matching degree greater than a preset matching threshold with the second trajectory query vector, it is determined that the reply result of the trajectory sub-problem does not have an answer to the trajectory sub-problem. If there is a third trajectory query vector in the first trajectory query vector that has a matching degree greater than the preset matching threshold with the second trajectory query vector, the third trajectory query vector is decoded, and the obtained answer to the trajectory sub-problem is taken as the reply result of the trajectory sub-problem.

[0116] S507, determine a positioning result for the trajectory problem according to the reply result of each trajectory sub-problem.

[0117] The description of step S507 can refer to the related content in the above embodiments, which will not be repeated here.

[0118] In the embodiments of the present disclosure, if the target tool is a pre-deployed trajectory query database, the attribute parameters of the target tool are obtained based on the task information, the attribute parameters and the trajectory sub-problems are combined, the combined trajectory sub-problems are embedded and mapped to obtain second trajectory query vectors, the second trajectory query vectors and the first trajectory query vectors are matched to obtain matching degrees, and the reply results of the trajectory sub-problems are determined according to the matching degrees. According to the step-by-step investigation and analysis of the trajectory query database on the trajectory sub-problems, the present disclosure can avoid missing the details of the answers, improve the efficiency and accuracy of the trajectory problem positioning, and improve the flexibility of trajectory mining.

[0119] It should be noted that in the embodiments of the present disclosure, among the plurality of trajectory sub-problems obtained by decomposing the trajectory problem, there can be trajectory sub-problems that need to call the trajectory query system, the trajectory content extraction system or the trajectory analysis system, and there can also be trajectory sub-problems that need to call the trajectory query database. The method for obtaining the reply results can refer to the description in the above embodiments, and the embodiments of the present disclosure do not limit this.

[0120] Figure 6 is a schematic diagram of a positioning method for a trajectory problem according to an embodiment of the present disclosure, as Figure 6 Optionally, in the embodiments of the present disclosure, the process of establishing the trajectory query database can include: obtaining a plurality of historical trajectory question and answer pairs, and embedding and mapping the plurality of historical trajectory question and answer pairs to obtain a plurality of first trajectory query vectors. Optionally, the historical trajectory question and answer pair includes a historical trajectory problem and a historical answer corresponding to the historical trajectory problem. In some embodiments, since the historical trajectory question and answer pair can be a larger file, it needs to be divided into a plurality of smaller text blocks. The file can be cut according to a fixed size, that is, the plurality of historical trajectory question and answer pairs are parsed and divided to obtain a plurality of trajectory information text blocks, the plurality of trajectory information text blocks are embedded and mapped to obtain a plurality of first trajectory query vectors, and then the trajectory query database is generated according to the plurality of first trajectory query vectors.

[0121] In some embodiments, if the target tool is a pre-deployed trajectory query database, the attribute parameters of the trajectory query database and the trajectory sub-problems are combined, the combined trajectory sub-problems are embedded and mapped to obtain second trajectory query vectors, the second trajectory query vectors are input into the trajectory query database, the trajectory query database matches the second trajectory query vectors with the first trajectory query vectors to obtain matching degrees, third trajectory query vectors with matching degrees greater than a preset matching threshold are obtained, and the third trajectory query vectors are input into a large language model for natural language processing to obtain the reply results of the trajectory sub-problems.

[0122] The present disclosure constructs a trajectory query database according to historical trajectory question and answer pairs, so as to facilitate subsequent step-by-step investigation and analysis of trajectory sub-problems, avoid missing answer details, improve the efficiency and accuracy of trajectory problem positioning, and improve the flexibility of trajectory mining.

[0123] Figure 7 is a flowchart of a positioning method for a trajectory problem according to an embodiment of the present disclosure, as shown in Figure 7 The method is executed by a target tool and includes the following steps:

[0124] S701, receiving a request instruction sent by an agent device.

[0125] In the embodiment of the present disclosure, the target tool can be a pre-deployed open platform, such as one of a trajectory query system, a trajectory content extraction system, and a trajectory analysis system, or a trajectory query database.

[0126] Optionally, the request instruction sent by the agent device is a first request instruction containing a trajectory sub-problem, or a preset second request instruction, wherein the trajectory sub-problem is a trajectory sub-problem after attribute parameters are combined. The process of obtaining the request instruction by the agent device can be referred to the description in the above embodiment, which will not be described here.

[0127] S702, analyzing the request instruction, in response to the request instruction being the first request instruction containing the trajectory sub-problem, analyzing the trajectory sub-problem, obtaining a reply result of the trajectory sub-problem, and sending the reply result of the trajectory sub-problem to the agent device.

[0128] In the embodiment of the present disclosure, when the target tool is one of the trajectory query system, the trajectory content extraction system, and the trajectory analysis system, after receiving the first request instruction, analyzing the first request instruction, and obtaining the trajectory sub-problem, the trajectory sub-problem is analyzed based on the function and resource of the target tool. Optionally, if the target tool can solve the trajectory sub-problem, that is, the answer to the trajectory sub-problem is obtained, the reply result is generated according to the first identifier indicating that the trajectory sub-problem is solved and the answer; if the target tool does not solve the trajectory sub-problem, that is, the answer to the trajectory sub-problem is not obtained, the reply result is generated according to the second identifier indicating that the trajectory sub-problem is not solved.

[0129] In the embodiment of the present disclosure, the answer can be in a text format, and in other implementations, the answer text can also be in other formats, which are not limited in the embodiment of the present disclosure.

[0130] S703, in response to the request instruction being the preset second request instruction, sending a first trajectory query vector in the trajectory query database to the agent device.

[0131] In the embodiments of the present disclosure, when the target tool is a trajectory query database, a second request instruction is received, and a first trajectory query vector in the trajectory query database is sent to the proxy device.

[0132] In the embodiments of the present disclosure, in the case that the problem involves multiple fields, the trajectory problem is analyzed by different tools, the knowledge and resources of different fields are called to solve the trajectory problem, the cooperation of multiple fields is realized, the complex trajectory problem can be solved more quickly and more intelligently, and the flexibility of trajectory mining is improved.

[0133] Figure 8 FIG. 1 is a structural diagram of a positioning device for a trajectory problem according to an embodiment of the present disclosure, as shown in FIG. 1, the positioning device for the trajectory problem 800 includes: Figure 8

[0134] The acquisition module 810 is configured to acquire the trajectory problem, and call a large language model to plan and decompose the trajectory problem, to acquire one or more trajectory sub-problems and task information of each trajectory sub-problem.

[0135] The analysis module 820 is configured to, for each trajectory sub-problem, call a target tool for solving the trajectory sub-problem based on the task information, and analyze the trajectory sub-problem based on the target tool, to acquire a reply result of the trajectory sub-problem.

[0136] The determination module 830 is configured to determine a positioning result for the trajectory problem according to the reply result of each trajectory sub-problem.

[0137] In some embodiments, the analysis module 820 is further configured to:

[0138] determine, according to the reply result, whether each trajectory sub-problem has an answer, and acquire, from the trajectory sub-problems, a trajectory sub-problem that does not have an answer;

[0139] continue to call the large language model to plan and decompose the trajectory sub-problem that does not have an answer, to update the trajectory sub-problem and the task information.

[0140] In some embodiments, the analysis module 820 is further configured to:

[0141] determine a first trajectory sub-problem that does not have an answer in the trajectory sub-problems, and acquire an answer of a second trajectory sub-problem other than the first trajectory sub-problem;

[0142] call the large language model to plan and decompose the answer of the first trajectory sub-problem and the second trajectory sub-problem, to obtain a third trajectory sub-problem and task information of the third trajectory problem;

[0143] update the third trajectory sub-problem and the task information of the third trajectory sub-problem as new trajectory sub-problems and task information.​

[0144] In some embodiments, the analysis module 820 is further configured to:

[0145] determine a tool identity of the target tool based on the task information, and invoke the target tool for solving the trajectory sub-problem based on the tool identity of the target tool.

[0146] In some embodiments, the analysis module 820 is further configured to:

[0147] if the target tool is one of the pre-deployed trajectory query system, the trajectory content extraction system, and the trajectory analysis system, obtain attribute parameters of the target tool based on the task information, and combine the attribute parameters with the trajectory sub-problem;

[0148] generate a first request instruction based on the combined trajectory sub-problem, and send the first request instruction to the target tool;

[0149] receive a reply result for the trajectory sub-problem sent by the target tool, the reply result being obtained by the target tool analyzing the combined trajectory sub-problem.

[0150] In some embodiments, the analysis module 820 is further configured to:

[0151] if the target tool is the trajectory query system, the attribute parameters of the target tool at least include specified time information, specified location range, and information of a terminal device; or

[0152] if the target tool is the trajectory content extraction system or the trajectory analysis system, the attribute parameters of the target tool at least include vehicle information, coordinate information of the vehicle, and timestamp information.

[0153] In some embodiments, the analysis module 820 is further configured to:

[0154] if the target tool is the pre-deployed trajectory query database, send a preset second request instruction to the trajectory query database, and receive a first trajectory query vector sent by the trajectory query database;

[0155] obtain attribute parameters of the target tool based on the task information, and combine the attribute parameters with the trajectory sub-problem;

[0156] perform embedding mapping on the combined trajectory sub-problem to obtain a second trajectory query vector;

[0157] match the second trajectory query vector with the first trajectory query vector to obtain a matching degree, and determine a reply result of the trajectory sub-problem according to the matching degree.

[0158] In some embodiments, the analysis module 820 is further configured to:

[0159] If there is no third trajectory query vector in the first trajectory query vector that has a matching degree greater than the preset matching threshold, it is determined that the reply result of the trajectory sub-problem has no answer to the trajectory sub-problem.

[0160] If there is a third trajectory query vector in the first trajectory query vector that has a matching degree greater than the preset matching threshold, the third trajectory query vector is decoded to obtain an answer to the trajectory sub-problem.

[0161] In some embodiments, the positioning apparatus 800 for the trajectory problem further includes a database establishing module 840 configured to:

[0162] obtain a plurality of historical trajectory question and answer pairs, the historical trajectory question and answer pair including a historical trajectory question and a historical answer corresponding to the historical trajectory question;

[0163] perform embedding mapping on the plurality of historical trajectory question and answer pairs to obtain a plurality of first trajectory query vectors, and generate a trajectory query database according to the plurality of first trajectory query vectors.

[0164] In some embodiments, the database establishing module 840 is further configured to:

[0165] perform parsing and segmentation on the plurality of historical trajectory question and answer pairs to obtain a plurality of trajectory information text blocks;

[0166] perform embedding mapping on the plurality of trajectory information text blocks to obtain the plurality of first trajectory query vectors.

[0167] In some embodiments, the determining module 830 is further configured to:

[0168] obtain an answer to each trajectory sub-problem according to the reply result;

[0169] combine the answer to each trajectory sub-problem to obtain a positioning result for the trajectory problem.

[0170] In some embodiments, the analyzing module 820 is further configured to:

[0171] for each trajectory sub-problem, determine a task identifier corresponding to the trajectory sub-problem according to task information of the trajectory sub-problem;

[0172] determine an analysis order of the trajectory sub-problem based on the task identifier corresponding to the trajectory sub-problem.

[0173] In some embodiments, the determining module 830 is further configured to:

[0174] obtain an answer to each trajectory sub-problem according to the reply result;

[0175] For each trajectory sub-problem, the answers to the trajectory sub-problems are combined according to the analysis sequence of the trajectory sub-problems to obtain a positioning result for the trajectory problem.

[0176] In the embodiments of the present disclosure, the large language model is called to plan and decompose the trajectory problem to obtain one or more trajectory sub-problems and task information of the trajectory sub-problems, multiple interrelated problems can be solved at the same time, thereby avoiding waste of time resources, for each trajectory sub-problem, a target tool for solving the trajectory sub-problem is called based on the task information, fusion of technologies such as artificial intelligence, machine learning and data analysis is realized, the trajectory sub-problems are analyzed based on tool resources in different fields, and a reply result of the trajectory sub-problem is obtained, errors caused by manual operation can be avoided, and the efficiency and accuracy of trajectory problem positioning are improved.

[0177] Figure 9 is a structural diagram of a positioning device for a trajectory problem according to an embodiment of the present disclosure, as Figure 9 shown, the positioning device for the trajectory problem 900 includes:

[0178] The receiving module 910 is configured to receive a request instruction sent by the agent device.

[0179] The first processing module 920 is configured to parse the request instruction, in response to the request instruction being a first request instruction containing a trajectory sub-problem, analyze the trajectory sub-problem, obtain a reply result of the trajectory sub-problem, and send the reply result of the trajectory sub-problem to the agent device.

[0180] The second processing module 930 is configured to, in response to the request instruction being a preset second request instruction, send a first trajectory query vector in the trajectory query database to the agent device.

[0181] In the embodiments of the present disclosure, the trajectory sub-problem is analyzed based on the target tool to obtain a reply result of the trajectory sub-problem, and the reply result of the trajectory sub-problem is sent to the agent device, in the case where the problem involves multiple fields, the trajectory problem is analyzed by different tools, knowledge and resources in different fields can be called to solve the trajectory problem, and multiple fields can be cooperated, the complex trajectory problem can be solved more quickly and more intelligently, and the flexibility of trajectory mining is improved.

[0182] According to the embodiments of the present disclosure, the present disclosure further provides an electronic device, a readable storage medium and a computer program product.

[0183] Figure 10is a block diagram of an electronic device for implementing embodiments of the present disclosure. The electronic device can implement the positioning method for a trajectory problem of the embodiments of the present disclosure, and the electronic device is intended to represent a variety of forms of digital computers, such as laptops, desktops, tablets, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular telephones, smartphones, wearable devices, and other similar computing devices. The components shown in the figure, their connections and relationships, and their functions, are meant to be examples only, and are not intended to limit the implementations of the present disclosure described and / or claimed in this document.

[0184] As shown in Figure 10 , the device 1000 includes a computing unit 1001 that can perform various appropriate actions and processes in accordance with a computer program stored in a read-only memory (ROM) 1002 or a computer program loaded from a storage unit 1008 into a random access memory (RAM) 1003. In the RAM 1003, various programs and data required for the operation of the device 1000 can also be stored. The computing unit 1001, the ROM 1002, and the RAM 1003 are connected to each other through a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.

[0185] A plurality of components in the device 1000 are connected to the I / O interface 1005, including: an input unit 1006, such as a keyboard, a mouse, etc.; an output unit 1007, such as various types of displays, speakers, etc.; a storage unit 1008, such as a magnetic disk, an optical disk, etc.; and a communication unit 1009, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1009 allows the device 1000 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunications networks.

[0186] The computing unit 1001 can be various general and / or special purpose processing components having processing and computing capabilities. Some examples of the computing unit 1001 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 1001 performs various methods and processes described above, such as the localization method for trajectory problems. For example, in some embodiments, the localization method for trajectory problems can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 1008. In some embodiments, part or all of the computer program can be loaded and / or installed onto the device 1000 via the ROM 1002 and / or the communication unit 1009. When the computer program is loaded onto the RAM 1003 and executed by the computing unit 1001, one or more steps of the localization method for trajectory problems described above can be performed. Alternatively, in other embodiments, the computing unit 1001 can be configured to perform the localization method for trajectory problems by any other suitable means, such as by means of firmware.

[0187] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a programmable logic device (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0188] Program code for carrying out methods of the present disclosure can be written in any combination of one or more programming languages. The program code can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the program code, when executed by the processor or controller, produces a means for implementing the functions / acts specified in the flowcharts and / or block diagrams. The program code can be executed entirely on a machine, partially on a machine, partially on a machine and partially on a remote machine or entirely on a remote machine or server.

[0189] In the context of this disclosure, a machine-readable medium can be a tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include but is not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0190] To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0191] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0192] The computer system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0193] It should be understood that the various forms of flow shown above can be used to reorder, add, or delete steps. For example, the steps described in the present disclosure can be performed in parallel, in series, or in a different order, as long as the desired results of the technology disclosed in the present disclosure can be achieved, which is not limited herein.

[0194] The above detailed description does not constitute a limitation on the protection scope of the present disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A localization method for a trajectory problem, wherein, Performed by the agent device, including: Obtain the trajectory problem and call the large language model to plan and decompose the trajectory problem to obtain one or more trajectory sub-problems and their respective task information; For each of the trajectory sub-problems, a target tool for solving the trajectory sub-problem is invoked based on the task information, and the trajectory sub-problem is analyzed based on the target tool to obtain the answer result of the trajectory sub-problem; Based on the response to each of the trajectory sub-questions, the localization result for the trajectory question is determined; The step of analyzing the trajectory sub-problem based on the target tool to obtain the answer to the trajectory sub-problem includes: If the target tool is a pre-deployed trajectory query database, a preset second request instruction is sent to the trajectory query database, and a first trajectory query vector is received from the trajectory query database; Based on the task information, the attribute parameters of the target tool are obtained, and the attribute parameters are combined with the trajectory sub-problem; The combined trajectory subproblems are embedded and mapped to obtain the second trajectory query vector; The second trajectory query vector is matched with the first trajectory query vector to obtain the matching degree, and the answer result of the trajectory sub-question is determined based on the matching degree.

2. The method of claim 1, wherein, After analyzing the trajectory sub-problem based on the target tool to obtain the answer to the trajectory sub-problem, the method further includes: Based on the response results, determine whether each trajectory sub-question has an answer, and obtain the trajectory sub-questions that do not have an answer from the trajectory sub-questions; The large language model is then invoked to plan and decompose the unanswered trajectory subproblems in order to update the trajectory subproblems and the task information.

3. The method according to claim 2, wherein, The step of continuing to call the large language model to plan and decompose the trajectory sub-problems that have no answers, in order to update the trajectory sub-problems and the task information, includes: Identify the first trajectory sub-problem that has no answer, and obtain the answer to the second trajectory sub-problem other than the first trajectory sub-problem; The large language model is invoked to perform planning decomposition on the answers to the first trajectory subproblem and the second trajectory subproblem to obtain the third trajectory subproblem and the task information of the third trajectory problem; The third trajectory subproblem and its task information are updated to the new trajectory subproblem and task information.

4. The method according to claim 2, wherein, The step of continuing to call the large language model to plan and decompose the trajectory sub-problems that have no answers, in order to update the trajectory sub-problems and the task information, includes: Identify the first trajectory subproblem that has no answer among the trajectory subproblems; The large language model is invoked to perform planning decomposition on the first trajectory subproblem, resulting in the fourth trajectory subproblem and the task information of the fourth trajectory problem; The first trajectory subproblem and its task information are replaced with the fourth trajectory subproblem and its task information to update the trajectory subproblem and its task information.

5. The method according to any one of claims 2-4, wherein, The step of invoking the target tool to solve the trajectory sub-problem based on the task information includes: Based on the task information, the tool identifier of the target tool is determined, and the target tool for solving the trajectory sub-problem is invoked based on the tool identifier of the target tool.

6. The method according to any one of claims 1-4, wherein, The step of analyzing the trajectory sub-problem based on the target tool to obtain the answer to the trajectory sub-problem includes: If the target tool is one of the tools in a pre-deployed trajectory query system, trajectory content extraction system, and trajectory analysis system, the attribute parameters of the target tool are obtained based on the task information, and the attribute parameters are combined with the trajectory sub-problem; A first request instruction is generated based on the combined trajectory sub-problems, and the first request instruction is sent to the target tool; The system receives a response from the target tool regarding the trajectory sub-problem, the response being obtained by the target tool through analysis of the combined trajectory sub-problem.

7. The method according to claim 6, wherein, include: If the target tool is the trajectory query system, then the attribute parameters of the target tool include specified time information, specified location range, and terminal device information; or If the target tool is the trajectory content extraction system or the trajectory analysis system, then the attribute parameters of the target tool include vehicle information, vehicle coordinate information, and timestamp information.

8. The method according to any one of claims 1-4, wherein, Determining the response result for the trajectory sub-question based on the matching degree includes: If there is no third trajectory query vector in the first trajectory query vector that has a matching degree greater than a preset matching threshold with the second trajectory query vector, the answer to the trajectory sub-question is determined to be that there is no answer to the trajectory sub-question; or If there exists a third trajectory query vector in the first trajectory query vector whose matching degree with the second trajectory query vector is greater than a preset matching threshold, the third trajectory query vector is decoded, and the answer to the obtained trajectory sub-question is used as the response result of the trajectory sub-question.

9. The method according to any one of claims 1-4, wherein, The process of establishing the trajectory query database includes: Obtain multiple historical trajectory question-and-answer pairs, wherein each historical trajectory question-and-answer pair includes a historical trajectory question and the corresponding historical answer to the historical trajectory question; Multiple historical trajectory question-and-answer pairs are embedded and mapped to obtain multiple first trajectory query vectors, and the trajectory query database is generated based on the multiple first trajectory query vectors.

10. The method according to claim 9, wherein, The embedding and mapping of multiple historical trajectory question-and-answer pairs to obtain multiple first trajectory query vectors includes: The historical trajectory question-and-answer pairs are parsed and segmented to obtain multiple trajectory information text blocks; Multiple trajectory information text blocks are embedded and mapped to obtain multiple first trajectory query vectors.

11. The method according to claim 1, wherein, The step of determining the localization result for the trajectory problem based on the response result of each trajectory sub-problem includes: Based on the response results, obtain the answer to each of the trajectory sub-questions; The answers to each of the trajectory sub-questions are combined to obtain the localization result for the trajectory question.

12. The method according to claim 1, wherein, Before analyzing the trajectory sub-problem based on the target tool, the method further includes: For each of the trajectory sub-problems, a task identifier corresponding to the trajectory sub-problem is determined based on the task information of the trajectory sub-problem; The analysis order of the trajectory sub-problems is determined based on the task identifiers corresponding to the trajectory sub-problems.

13. The method according to claim 12, wherein, The step of determining the localization result for the trajectory problem based on the response result of each trajectory sub-problem includes: Based on the response results, obtain the answer to each of the trajectory sub-questions; For each of the trajectory sub-problems, the answers to the trajectory sub-problems are combined according to the analysis order of the trajectory sub-problems to obtain the localization result for the trajectory problem.

14. A localization method for trajectory problems, wherein, Executed by the target tool, including: Receive request instructions sent by the agent device; The request instruction is parsed, and in response to the request instruction being a first request instruction containing a trajectory sub-question, the trajectory sub-question is analyzed, the response result of the trajectory sub-question is obtained, and the response result of the trajectory sub-question is sent to the agent device; In response to the request instruction being a preset second request instruction, a first trajectory query vector from the trajectory query database is sent to the agent device; The second request instruction is sent by the agent device when the target tool is the trajectory query database; The first trajectory query vector is used to enable the agent device to obtain the attribute parameters of the target tool based on the task information, combine the attribute parameters and the trajectory sub-question; perform embedding mapping on the combined trajectory sub-question to obtain the second trajectory query vector; match the second trajectory query vector with the first trajectory query vector to obtain the matching degree, and determine the answer result of the trajectory sub-question based on the matching degree.

15. A positioning device for trajectory problems, wherein, include: The acquisition module is used to acquire the trajectory problem and call the large language model to plan and decompose the trajectory problem to obtain one or more trajectory sub-problems and their respective task information; The analysis module is used to, for each of the trajectory sub-problems, invoke the target tool for solving the trajectory sub-problem based on the task information, and analyze the trajectory sub-problem based on the target tool to obtain the answer result of the trajectory sub-problem; The determining module is used to determine the positioning result for the trajectory problem based on the response result of each trajectory sub-problem; The analysis module is further configured to: If the target tool is a pre-deployed trajectory query database, a preset second request instruction is sent to the trajectory query database, and a first trajectory query vector is received from the trajectory query database; Based on the task information, the attribute parameters of the target tool are obtained, and the attribute parameters are combined with the trajectory sub-problem; The combined trajectory subproblems are embedded and mapped to obtain the second trajectory query vector; The second trajectory query vector is matched with the first trajectory query vector to obtain the matching degree, and the answer result of the trajectory sub-question is determined based on the matching degree.

16. The apparatus according to claim 15, wherein, The analysis module is also used for: Based on the response results, determine whether each trajectory sub-question has an answer, and obtain the trajectory sub-questions that do not have an answer from the trajectory sub-questions; The large language model is then invoked to plan and decompose the unanswered trajectory subproblems in order to update the trajectory subproblems and the task information.

17. The apparatus according to claim 16, wherein, The analysis module is also used for: Identify the first trajectory sub-problem that has no answer, and obtain the answer to the second trajectory sub-problem other than the first trajectory sub-problem; The large language model is invoked to perform planning decomposition on the answers to the first trajectory subproblem and the second trajectory subproblem to obtain the third trajectory subproblem and the task information of the third trajectory problem; The third trajectory subproblem and its task information are updated to the new trajectory subproblem and task information.

18. The apparatus according to claim 16, wherein, The analysis module is also used for: Identify the first trajectory subproblem that has no answer among the trajectory subproblems; The large language model is invoked to perform planning decomposition on the first trajectory subproblem, resulting in the fourth trajectory subproblem and the task information of the fourth trajectory problem; The first trajectory subproblem and its task information are replaced with the fourth trajectory subproblem and its task information to update the trajectory subproblem and its task information.

19. The apparatus according to any one of claims 16-18, wherein, The analysis module is also used for: Based on the task information, the tool identifier of the target tool is determined, and the target tool for solving the trajectory sub-problem is invoked based on the tool identifier of the target tool.

20. The apparatus according to any one of claims 15-18, wherein, The analysis module is also used for: If the target tool is one of the tools in a pre-deployed trajectory query system, trajectory content extraction system, and trajectory analysis system, the attribute parameters of the target tool are obtained based on the task information, and the attribute parameters are combined with the trajectory sub-problem; A first request instruction is generated based on the combined trajectory sub-problems, and the first request instruction is sent to the target tool; The system receives a response from the target tool regarding the trajectory sub-problem, the response being obtained by the target tool through analysis of the combined trajectory sub-problem.

21. The apparatus according to claim 20, wherein, The analysis module is also used for: If the target tool is the trajectory query system, then the attribute parameters of the target tool include specified time information, specified location range, and terminal device information; or If the target tool is the trajectory content extraction system or the trajectory analysis system, then the attribute parameters of the target tool include vehicle information, vehicle coordinate information, and timestamp information.

22. The apparatus according to any one of claims 15-18, wherein, The analysis module is also used for: If there is no third trajectory query vector in the first trajectory query vector that has a matching degree greater than a preset matching threshold with the second trajectory query vector, the answer result of the trajectory sub-question is determined to be that there is no answer to the trajectory sub-question; or If there exists a third trajectory query vector in the first trajectory query vector whose matching degree with the second trajectory query vector is greater than a preset matching threshold, the third trajectory query vector is decoded, and the answer to the trajectory sub-question is used as the response result of the trajectory sub-question.

23. The apparatus according to any one of claims 15-18, wherein, The device further includes a database creation module for: Obtain multiple historical trajectory question-and-answer pairs, wherein each historical trajectory question-and-answer pair includes a historical trajectory question and the corresponding historical answer to the historical trajectory question; Multiple historical trajectory question-and-answer pairs are embedded and mapped to obtain multiple first trajectory query vectors, and the trajectory query database is generated based on the multiple first trajectory query vectors.

24. The apparatus according to claim 23, wherein, The database creation module is also used for: The historical trajectory question-and-answer pairs are parsed and segmented to obtain multiple trajectory information text blocks; Multiple trajectory information text blocks are embedded and mapped to obtain multiple first trajectory query vectors.

25. The apparatus according to claim 15, wherein, The determining module is further configured to: Based on the response results, obtain the answer to each of the trajectory sub-questions; The answers to each of the trajectory sub-questions are combined to obtain the localization result for the trajectory question.

26. The apparatus according to claim 15, wherein, The analysis module is also used for: For each of the trajectory sub-problems, a task identifier corresponding to the trajectory sub-problem is determined based on the task information of the trajectory sub-problem; The analysis order of the trajectory sub-problems is determined based on the task identifiers corresponding to the trajectory sub-problems.

27. The apparatus according to claim 26, wherein, The determining module is further configured to: Based on the response results, obtain the answer to each of the trajectory sub-questions; For each of the trajectory sub-problems, the answers to the trajectory sub-problems are combined according to the analysis order of the trajectory sub-problems to obtain the localization result for the trajectory problem.

28. A positioning device for trajectory problems, wherein, include: The receiving module is used to receive request instructions sent by the agent device; The first processing module is configured to parse the request instruction, and in response to the request instruction being a first request instruction containing a trajectory sub-question, analyze the trajectory sub-question, obtain the response result of the trajectory sub-question, and send the response result of the trajectory sub-question to the agent device; The second processing module is used to send a first trajectory query vector from the trajectory query database to the agent device in response to the request instruction being a preset second request instruction. The second request instruction is sent by the agent device when the target tool is the trajectory query database; The first trajectory query vector is used to enable the agent device to obtain the attribute parameters of the target tool based on the task information, combine the attribute parameters and the trajectory sub-question; perform embedding mapping on the combined trajectory sub-question to obtain the second trajectory query vector; match the second trajectory query vector with the first trajectory query vector to obtain the matching degree, and determine the answer result of the trajectory sub-question based on the matching degree.

29. An electronic device comprising: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-13, or to perform the method of claim 14.

30. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-13, or to perform the method according to claim 14.

31. A computer program product comprising a computer program that, when executed by a processor, implements the steps of the method according to any one of claims 1-13, or performs the steps of the method according to claim 14.

Images