Path planning method, device and equipment of vehicle and storage medium

Abstract

This invention relates to the field of vehicle technology and discloses a method, apparatus, device, and storage medium for vehicle route planning. The method includes: acquiring target multidimensional semantic features from a user-issued natural language command and real-time scene information about the vehicle's environment; determining the target's historical route in a trip knowledge base based on the target multidimensional semantic features, wherein the trip knowledge base records the correspondence between different multidimensional semantic features and different historical routes; and performing route planning for the vehicle based on the target's historical route and real-time scene information. Applying the technical solution of this invention can solve the technical problems of low travel convenience and operational safety in existing technologies.

Description

Technical Field

[0001] This invention relates to the field of vehicle technology, specifically to a method, apparatus, device, and storage medium for vehicle path planning. Background Technology

[0002] With the rapid growth in demand for smart travel, navigation systems have evolved from simple route planning tools into key assistants in users' daily lives, especially in scenarios such as commuting, social activities, travel, and shopping, where users frequently need to navigate to previously visited locations.

[0003] In current technology, users still need to manually select locations from their historically visited sites to fit the navigation route, which presents technical problems of low travel convenience and low operational security. Summary of the Invention

[0004] In view of the above problems, embodiments of the present invention provide a method, apparatus, device and storage medium for vehicle route planning, which solves the technical problems of low travel convenience and low operational safety in the prior art.

[0005] According to one aspect of the present invention, a vehicle path planning method is provided, the method comprising:

[0006] Acquire the target multidimensional semantic features and real-time scene information of the vehicle's environment from the natural language commands issued by the user.

[0007] Based on the target's multidimensional semantic features, the target's historical itinerary is determined in the itinerary knowledge base, which records the correspondence between different multidimensional semantic features and different historical itineraries;

[0008] Based on the target's historical travel history and the real-time scene information, the vehicle's route is planned.

[0009] According to another aspect of the present invention, a vehicle path planning device is provided, comprising:

[0010] The acquisition module is used to acquire the target multidimensional semantic features and real-time scene information of the vehicle's environment from the natural language commands issued by the user.

[0011] The determination module is used to determine the target historical itinerary in the itinerary knowledge base based on the target multidimensional semantic features. The itinerary knowledge base records the correspondence between different multidimensional semantic features and different historical itineraries.

[0012] The processing module is used to perform route planning for the vehicle based on the target's historical travel history and the real-time scene information.

[0013] According to another aspect of the present invention, an electronic device is provided, including: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other through the communication bus;

[0014] The memory is used to store at least one executable instruction that causes the processor to perform operations such as the vehicle path planning method described above.

[0015] According to another aspect of the present invention, a computer-readable storage medium is provided:

[0016] The storage medium stores at least one executable instruction that causes the vehicle's path planning device / electronic device to perform the operation of the vehicle path planning method described above.

[0017] According to another aspect of the present invention, a computer program product is provided, including a computer program that, when executed by a processor, causes a vehicle's path planning device / electronic device to perform the operation of the above-described method.

[0018] This invention provides an embodiment of the invention that obtains the target multidimensional semantic features from the natural language command to be processed issued by the user and the real-time scene information of the vehicle's environment; determines the target historical journey in the journey knowledge base based on the target multidimensional semantic features, the journey knowledge base recording the correspondence between different multidimensional semantic features and different historical journeys; and performs route planning for the vehicle based on the target historical journey and the real-time scene information. This technical solution acquires the target's multidimensional semantic features from the user's natural language commands, enabling the vehicle to accurately parse key elements in the user's ambiguous expressions. This allows the vehicle to understand the user's intent without requiring precise address input, significantly reducing the barrier to human-vehicle interaction and operational complexity. By acquiring real-time scene information about the vehicle's environment, the vehicle can perceive the current dynamic environment, providing an environmental basis for subsequent personalized adaptation. Based on the target's multidimensional semantic features, the system determines the target's historical itinerary in the trip knowledge base, allowing the vehicle to accurately recall past locations based on the user's historical behavioral habits, achieving an intelligent upgrade from address navigation to memory navigation. Based on the target's historical itinerary and real-time scene information, the system performs route planning, enabling navigation routes and parking solutions to simultaneously integrate the user's personalized preferences with the actual needs of the current environment, thereby significantly improving driving safety, convenience, and the overall user experience.

[0019] The above description is merely an overview of the technical solutions of the embodiments of the present invention. In order to better understand the technical means of the embodiments of the present invention and to implement them in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the embodiments of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0020] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0021] Figure 1 A schematic diagram of the vehicle path planning system provided by the present invention is shown;

[0022] Figure 2 A flowchart of a first embodiment of the vehicle path planning method provided by the present invention is shown;

[0023] Figure 3 A flowchart of a second embodiment of the vehicle path planning method provided by the present invention is shown;

[0024] Figure 4 A flowchart of a third embodiment of the vehicle path planning method provided by the present invention is shown;

[0025] Figure 5 A flowchart of a fourth embodiment of the vehicle path planning method provided by the present invention is shown;

[0026] Figure 6 A flowchart of a fifth embodiment of the vehicle path planning method provided by the present invention is shown;

[0027] Figure 7 A flowchart of a sixth embodiment of the vehicle path planning method provided by the present invention is shown;

[0028] Figure 8 A schematic diagram of an embodiment of the vehicle path planning device provided by the present invention is shown;

[0029] Figure 9 A schematic diagram of an embodiment of the electronic device provided by the present invention is shown. Detailed Implementation

[0030] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0031] With the rapid growth in demand for smart travel, navigation systems have evolved from simple route planning tools into key assistants in users' daily lives. In scenarios such as commuting, social activities, travel, and shopping, users frequently need to navigate to previously visited locations.

[0032] However, existing navigation systems still rely on precise address input to process user commands, requiring users to manually search or look up the target location in an address book.

[0033] In addition, user travel scenarios are highly dynamic: for example, indoor parking spaces should be prioritized on rainy days, congested roads should be avoided on holidays, and parking spaces near elevators should be chosen when traveling with elderly people.

[0034] However, the above approach has the following technical problems:

[0035] 1) Historical location retrieval relies on precise addresses: Users need to enter precise addresses to navigate to previously visited locations (e.g., the badminton hall visited last month), which cannot be directly retrieved using vague natural language commands, making the operation cumbersome;

[0036] 2) Disconnect between scenario and navigation / parking recommendations: The existing navigation only plans routes based on real-time traffic conditions and does not recommend suitable parking solutions (such as indoor parking spaces, parking spaces close to the elevator) based on user scenarios (such as rainy days, holidays, or elderly companions).

[0037] 3) High interaction costs in multiple scenarios: Adjusting routes or searching for locations requires multiple manual screen interactions. Even with voice assistants, the understanding and execution capabilities for complex intentions are still insufficient, and they cannot automatically adapt to user habits, thus failing to achieve true intelligence and a seamless experience.

[0038] Based on the above-mentioned technical problems, the technical concept of this invention is as follows: To solve the problem that users cannot recall historical locations through fuzzy commands, the inventors conceived of constructing a user trip knowledge base, upgrading traditional single address data into a contextualized memory that includes multi-dimensional semantic features such as time, companions, and behavior. The current natural language command is semantically matched with the historical records in the trip knowledge base, thereby accurately locating the target historical trip and achieving intelligent understanding of fuzzy commands. Simultaneously, to address the pain point of navigation being disconnected from the scene, real-time scene information is further introduced. Combining real-time scene information with the target historical trip generates a highly personalized and dynamically adaptable route planning scheme, thereby solving the aforementioned problems.

[0039] Based on the above technical concept, Figure 1 A schematic diagram of the vehicle path planning system provided by the present invention is shown, as follows: Figure 1 As shown, the system includes a perception layer, a processing layer, an output layer, and a storage layer.

[0040] In the perception layer, the user terminal outputs natural language commands, which are collected by the command acquisition unit; the external data interface sends real-time scene data (weather, etc.) to the scene data acquisition unit; the parking lot data interface sends real-time parking lot data to the parking data acquisition unit.

[0041] In the processing layer, the natural language parsing module obtains natural language instructions and retrieves structured data from the user trip knowledge base (hereinafter referred to as the trip knowledge base) in the storage layer. After parsing the speech, it inputs the data into the scene matching unit for parking space matching and historical scene matching, so that the parking recommendation module and navigation planning module can make recommendations and plans.

[0042] In the output layer, the parking solution output unit outputs accurate parking lot recommendations to the user terminal, and the navigation result output unit outputs scene-matched navigation routes to the user terminal.

[0043] In the storage layer, user terminal selection operations can be optimized in the knowledge base optimization unit based on newly added data after user feedback, thus improving the real-time user itinerary database.

[0044] The above content provides a brief description of the present invention. The following detailed description, through specific embodiments, further illustrates the technical solution of the present invention. The executing entity of the present invention is an electronic device, specifically a vehicle, a central control device within a vehicle, or a user's terminal device. It should be noted that the following 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] Figure 2 A flowchart of a first embodiment of the vehicle path planning method provided by the present invention is shown, as follows: Figure 2 As shown, the method includes the following steps:

[0046] Step 21: Obtain the target multidimensional semantic features and real-time scene information of the vehicle's environment from the natural language command issued by the user.

[0047] In this step, vague commands issued by the user in natural language form are received through the vehicle's interactive interface and recorded as natural language commands to be processed.

[0048] At the same time, the vehicle's current environment is acquired in real time through onboard sensors and network connections, and recorded as real-time scene information, including weather data, holiday information, real-time traffic flow, etc.

[0049] Furthermore, features are extracted from the natural language instructions to be processed to obtain multidimensional semantic features, denoted as target multidimensional semantic features.

[0050] Optionally, the vehicle's camera sensors can be used to collect information about fellow passengers inside the vehicle and update the target's multidimensional semantic features.

[0051] Step 22: Based on the multidimensional semantic features of the target, determine the target's historical itinerary in the itinerary knowledge base;

[0052] Among them, the itinerary knowledge base records the correspondence between different multidimensional semantic features and different historical itineraries;

[0053] In this step, the target multidimensional semantic features from step 21 are used as retrieval conditions. A matching query is performed in the user trip knowledge base that is pre-built and stored in the vehicle or cloud to obtain historical trip records that meet all or a preset number of semantic features. The historical trip record is then identified as the target historical trip.

[0054] For example, each historical trip record in the trip knowledge base is associated with a corresponding multidimensional semantic feature. The target multidimensional semantic feature is compared with the multidimensional semantic features corresponding to the historical trips stored in the knowledge base.

[0055] It should be understood that historical itinerary data in the itinerary knowledge base needs to be saved and processed after obtaining the user's authorization.

[0056] Optionally, updates to the itinerary knowledge base can include the following implementations:

[0057] Step 1: Collect location data, scene tags, and behavior data of the vehicle's current journey;

[0058] In this implementation, with the user's authorization, multi-dimensional raw data is continuously collected during the current trip, including data on the location the vehicle arrives at, scene tags when the trip occurs (time, companions, weather, parking preferences), and user behavior data during the trip (duration of stay, consumption records (eating, shopping, etc.)).

[0059] Step 2: Construct the first multi-dimensional semantic feature based on the location data, scene tags, and behavioral data in the current itinerary;

[0060] In this implementation, location data, scene tags, and behavioral data in the current trip are integrated and processed to construct the first multi-dimensional semantic feature representing the current trip.

[0061] Step 3: Update the itinerary knowledge base based on the current itinerary and the first multidimensional semantic features.

[0062] In this implementation, the first multidimensional semantic feature is associated with the corresponding current trip to obtain a new multidimensional semantic feature and historical trip, which are then stored in the user trip knowledge base to complete the dynamic expansion and updating of the knowledge base.

[0063] Step 23: Based on the target's historical travel history and real-time scene information, perform route planning for the vehicle.

[0064] In this step, since the real-time scene information reflects the path to the target location in the historical journey and the environmental conditions in the parking area, the vehicle is route-planned based on this real-time scene information and the target historical journey so that the user can safely reach the required location.

[0065] Optionally, the target historical itinerary includes at least one location; accordingly, step 23 may include the following implementation:

[0066] Step 1: In response to the user's selection, determine the target location from at least one location;

[0067] In this implementation, multiple candidate historical locations can be output, i.e., at least one location. At this time, the at least one location is presented to the user through a human-computer interaction interface or voice broadcast, and the user's selection command is awaited.

[0068] Furthermore, upon receiving the user's selection of a location, the selected location is designated as the target location.

[0069] Step 2: Based on the target location and real-time scene information, plan the route for the vehicle.

[0070] In this implementation, the target location is used as the navigation endpoint. Combined with real-time scene information, path planning calculations are performed and the navigation route to the target location and the parking route to the target parking space at the target location are output.

[0071] The vehicle path planning method provided in this embodiment of the invention obtains the target multidimensional semantic features in the natural language command to be processed issued by the user and the real-time scene information of the vehicle's environment; determines the target historical journey in the journey knowledge base based on the target multidimensional semantic features, the journey knowledge base records the correspondence between different multidimensional semantic features and different historical journeys; and performs path planning for the vehicle based on the target historical journey and the real-time scene information. This technical solution acquires the target's multidimensional semantic features from the user's natural language commands, enabling the vehicle to accurately parse key elements in the user's ambiguous expressions. This allows the vehicle to understand the user's intent without requiring precise address input, significantly reducing the barrier to human-vehicle interaction and operational complexity. By acquiring real-time scene information about the vehicle's environment, the vehicle can perceive the current dynamic environment, providing an environmental basis for subsequent personalized adaptation. Based on the target's multidimensional semantic features, the system determines the target's historical itinerary in the trip knowledge base, allowing the vehicle to accurately recall past locations based on the user's historical behavioral habits, achieving an intelligent upgrade from address navigation to memory navigation. Based on the target's historical itinerary and real-time scene information, the system performs route planning, enabling navigation routes and parking solutions to simultaneously integrate the user's personalized preferences with the actual needs of the current environment, thereby significantly improving driving safety, convenience, and the overall user experience.

[0072] Based on the above embodiments, Figure 3 A flowchart of a second embodiment of the vehicle path planning method provided by the present invention is shown, as follows: Figure 3 As shown, obtaining the target multidimensional semantic features in the natural language command to be processed issued by the user in step 21 above may include the following steps:

[0073] Step 31: Obtain the natural language instructions to be processed;

[0074] In this step, the user's raw voice signal is collected and converted into a text-based natural language instruction to be processed.

[0075] The natural language instruction to be processed is the user's travel intention expressed in a colloquial manner, and the content may involve descriptive recollections of past trips.

[0076] For example, the natural language instruction to be processed could be: Go to the mall where you and your family went last week.

[0077] Step 32: Input the natural language instruction to be processed into the large language model to obtain the target multidimensional semantic features;

[0078] Among them, the large language model is trained based on multiple natural language instructions and multidimensional semantic features in different natural language instructions.

[0079] In this step, the natural language instructions to be processed obtained in step 31 are input into the pre-trained large language model, and the structured target multidimensional semantic features are output, which may include dimensions such as time features, peer features, and behavioral features.

[0080] Correspondingly, the large language model is pre-trained based on a large number of natural language instruction samples and their corresponding labeled multidimensional semantic features, extracting structured information from fuzzy spoken expressions and performing semantic parsing on input instructions.

[0081] For example, the target multidimensional semantic features corresponding to the shopping mall visited with family last week could be: time: last week, companions: family, behavior: shopping mall.

[0082] The vehicle route planning method provided in this invention acquires natural language commands to be processed; inputs these commands into a large language model to obtain target multidimensional semantic features. The large language model is trained based on multiple natural language commands and multidimensional semantic features from different natural language commands. In this solution, by acquiring the natural language commands to be processed, the vehicle can receive the user's travel intentions in the most natural and convenient way, completely eliminating the traditional navigation's mandatory requirement for precise address input. By inputting the natural language commands into the large language model, its powerful semantic understanding capabilities enable it to accurately extract target multidimensional semantic features such as time, fellow passengers, and behavior from ambiguous colloquial expressions, solving the problem that current keyword matching methods cannot understand complex intentions. Because the large language model is trained based on multiple natural language commands and multidimensional semantic features from different natural language commands, it possesses broad generalization capabilities and deep semantic parsing capabilities, adapting to the diverse expression habits of different users, thereby significantly improving the accuracy of command parsing.

[0083] Based on the above embodiments, Figure 4 A flowchart of a third embodiment of the vehicle path planning method provided by the present invention is shown, as follows: Figure 4 As shown, step 23 above may include the following steps:

[0084] Step 41: Determine the navigation route for the vehicle to reach the target location based on the target location in the target's historical travel history and real-time scene information;

[0085] In this step, the coordinates of the target location in the target's historical journey are used as the navigation endpoint, and the vehicle's current location is used as the starting point. Combining environmental parameters such as road condition data and holiday data in real-time scene information, one or more drivable routes from the starting point to the endpoint can be calculated through path planning algorithms, which serve as navigation routes.

[0086] If multiple routes exist, information such as the advantages and disadvantages of each route can be provided (e.g., avoiding congested areas around the shopping mall), allowing the user to select the route as the navigation route to the destination.

[0087] Step 42: When arriving at the target location based on the navigation route, determine the target parking space and the parking route to the target parking space based on the scene tags in the target's historical itinerary and the weather data in the real-time scene information.

[0088] In this step, when the vehicle travels to the vicinity of the target location according to the navigation route generated in step 41, the scene tags associated with the target location in the target's historical travel are retrieved. At the same time, the current weather data in the real-time scene information is combined with the structured data of the parking lot at the target location to filter out parking spaces that meet the scene tags and weather conditions as target parking spaces, and the driving route from the vehicle's current location to the target parking space is planned and recorded as the parking route.

[0089] For example, scene tags could be parking preferences, companion information (such as parking near the elevator for elderly people); parking lot structured data could be: parking space type, location distribution.

[0090] In one possible implementation, at least one parking space can be identified based on scene tags in the target's historical itinerary and weather data in the real-time scene information, and then the parking space selected by the user can be used as the target parking space.

[0091] Optionally, the vehicle's path planning method may also include:

[0092] Step 1: Respond to the user's parking space adjustment command and determine the target parking space after adjustment;

[0093] In this implementation, if a user issues a parking space adjustment command (e.g., select another parking space) through an interactive interface or voice command, the system receives the parking space adjustment command and identifies the user's final selected target parking space after adjustment.

[0094] For example, the target parking space is an indoor parking space on the B2 floor of XX shopping mall (30 meters from the elevator), and the user has selected a new parking space.

[0095] Step 2: Update the scene tags in the target historical trip based on the adjusted target parking space.

[0096] In this implementation, the feature information (such as location attributes) of the adjusted target parking space is then extracted and updated to the field of the scene label stored in association with the current trip and target location.

[0097] Specifically, it could be the parking preference field in the following scenario tags.

[0098] For example, if a user chooses to change to a parking space closer to the restaurant, the scene label for that target parking space will be updated to "parking closer to the restaurant".

[0099] Optionally, scene tags may also include: time information, behavioral data, and weather information.

[0100] The vehicle route planning method provided in this invention determines the navigation route to the target location based on the target location in the target's historical journey and real-time scene information. Upon arrival at the target location based on the navigation route, the method determines the target parking space and the parking route to that space based on scene tags in the target's historical journey and weather data in the real-time scene information. This solution, by determining the navigation route based on the target location in the target's historical journey and real-time scene information, enables the vehicle to dynamically plan a driving route that conforms to both user preferences and real-time traffic conditions, combining the user's past personalized habits with the current environment, thus improving the accuracy and practicality of route planning. Furthermore, by determining the target parking space and the parking route to that space based on scene tags in the target's historical journey and weather data in the real-time scene information upon arrival at the target location, the method allows the vehicle to proactively recommend the parking solution that best suits the user's current needs before arrival, greatly improving the convenience and comfort of travel in complex scenarios.

[0101] Based on the above embodiments, Figure 5 A flowchart of a fourth embodiment of the vehicle path planning method provided by the present invention is shown, as follows: Figure 5 As shown, one possible implementation of this solution may include the following steps:

[0102] Step 51, Begin;

[0103] Step 52: Storing user's historical itinerary with scene tags; collecting location / scene tags / behavioral data to construct user itinerary data;

[0104] Step 53: Multi-feature matching of fuzzy natural language commands; parsing semantic features, matching historical low points, and outputting a list of candidate locations;

[0105] Step 54: Dynamic generation of scenario-based navigation and parking solutions; obtain the current scenario, adjust the navigation route, and recommend candidate parking lots;

[0106] Step 55: Adaptive optimization based on user feedback; record user selections and update knowledge base scene tags;

[0107] Step 56: End (Recommendation for subsequent loop optimization).

[0108] Based on the above embodiments, Figure 6 A flowchart of a fifth embodiment of the vehicle path planning method provided by the present invention is shown, as follows: Figure 6 As shown, the process for historical scene matching and destination retrieval logic can include:

[0109] Step 61: User's natural language command; Example: Go to the mall I went to with my family last week;

[0110] Step 62: Semantic feature parsing;

[0111] Step 63: Extract core features (Time = last week, Companions = family members; Location = shopping mall);

[0112] Step 64, User Itinerary Knowledge Base (Structured Storage); Historical Data: Location = xx Shopping Mall;

[0113] Step 65: Feature logic matching; Time feature matching;

[0114] Step 66: Output the candidate location list; prioritize by highest matching degree.

[0115] Step 67: Destination Confirmation.

[0116] Based on the above embodiments, Figure 7 A flowchart of a sixth embodiment of the vehicle path planning method provided by the present invention is shown, as follows: Figure 7 As shown, strategies for recommending dynamic parking spaces may include:

[0117] Step 71: Collect current scene information; Example: Holiday + Rainy day + Companion = Elderly person;

[0118] Step 72, Scene Adaptation Rule Engine; Rainy days: Prioritize indoor parking spaces; Elderly: Prioritize spaces near elevators;

[0119] Step 73: Link real-time parking lot data;

[0120] Step 74: Parking space filtering and sorting; sort by matching degree from high to low;

[0121] Step 75, Recommendation Results; User Feedback: Changed to a parking space closer to xx restaurant;

[0122] Step 76: Update scene labels; proceed to step 72.

[0123] The vehicle path planning method provided in this embodiment of the invention has the following beneficial effects:

[0124] 1) Reduced operating costs: Supports fuzzy natural language commands to recall historical locations without the need to input precise addresses. This reduces user operation steps by at least half, allowing users to express their intentions in the most natural way. It greatly reduces the interaction burden and cognitive load during driving and improves safety.

[0125] 2) Enhanced Contextual Adaptability: By deeply integrating refined user historical contextual tags with real-time multi-dimensional contextual information into navigation and parking recommendations, the system gains decision-making capabilities similar to human drivers. It can provide personalized and dynamically optimized solutions based on contexts such as weather, holidays, and personal habits, which can significantly improve user satisfaction.

[0126] 3) Personalized accuracy optimization: The knowledge base is continuously updated based on user feedback, which can continuously improve the accuracy of recommendations, thereby enhancing users' travel efficiency and enjoyment.

[0127] Figure 8 A schematic diagram of an embodiment of the vehicle path planning device provided by the present invention is shown. Figure 8 As shown, the device includes:

[0128] The acquisition module 81 is used to acquire the target multidimensional semantic features and the real-time scene information of the vehicle's environment in the natural language command to be processed issued by the user.

[0129] The determination module 82 is used to determine the target's historical itinerary in the itinerary knowledge base based on the target's multidimensional semantic features. The itinerary knowledge base records the correspondence between different multidimensional semantic features and different historical itineraries.

[0130] The processing module 83 is used to plan the route for the vehicle based on the target's historical travel history and real-time scene information.

[0131] In one or more embodiments, the acquisition module 81 acquires the target multidimensional semantic features in the natural language command to be processed issued by the user, specifically for:

[0132] Obtain the natural language instructions to be processed;

[0133] The natural language instruction to be processed is input into the large language model to obtain the target multidimensional semantic features. The large language model is trained based on multiple natural language instructions and multidimensional semantic features from different natural language instructions.

[0134] In one or more embodiments, the processing module 83 performs route planning for the vehicle based on the target historical itinerary and real-time scene information, specifically for:

[0135] Based on the target location in the target's historical travel history and real-time scene information, determine the navigation route for the vehicle to reach the target location;

[0136] When arriving at the target location based on the navigation route, the target parking space and the parking route to the target parking space are determined based on the scene tags in the target's historical journey and the weather data in the real-time scene information.

[0137] In one or more embodiments, the processing module 83 is further configured to:

[0138] In response to a user's parking space adjustment command, determine the target parking space after adjustment;

[0139] Update the scene tags in the target historical trip based on the adjusted target parking space.

[0140] In one or more embodiments, the target historical itinerary includes at least one location;

[0141] Accordingly, based on the target's historical travel history and real-time scenario information, route planning is performed for the vehicle, including:

[0142] In response to the user's selection action, the target location is determined from at least one location;

[0143] Based on the target location and real-time scene information, the vehicle's route is planned.

[0144] In one or more embodiments, the processing module 83 is further configured to:

[0145] Collect location data, scene tags, and behavioral data of the vehicle during its current journey;

[0146] Based on location data, scene tags, and behavioral data in the current itinerary, construct the first multi-dimensional semantic feature;

[0147] Update the itinerary knowledge base based on the current itinerary and the first multidimensional semantic features.

[0148] In one or more embodiments, the scene tags include: parking preferences, companion information, time information, and weather information.

[0149] It should be noted that the division of the various modules in the above device is merely a logical functional division. In actual implementation, they can be fully or partially integrated into a single physical element, or they can be physically separated. Furthermore, these modules can be implemented entirely in software through processing element calls, or entirely in hardware. Alternatively, some modules can be implemented through processing element calls in software, while others can be implemented in hardware. Moreover, these modules can be integrated together or implemented independently. The processing element here can be an integrated circuit with signal processing capabilities. During implementation, each step of the above method or each of the above modules can be completed through the integrated logic circuits in the hardware of the processor element or through software instructions.

[0150] As can be seen from the above, the vehicle path planning device provided in this embodiment of the invention can accurately parse the key elements in the user's fuzzy expression by acquiring the target multidimensional semantic features in the natural language command to be processed issued by the user. Thus, the user can understand the intent without having to input an exact address, greatly reducing the threshold and complexity of human-vehicle interaction. By acquiring real-time scene information of the vehicle's environment, the vehicle can perceive the current dynamic environment, providing an environmental basis for subsequent personalized adaptation. Based on the target multidimensional semantic features, the target's historical journey is determined in the journey knowledge base, enabling the vehicle to accurately recall past locations based on the user's historical behavioral habits, realizing an intelligent upgrade from address navigation to memory navigation. Based on the target's historical journey and real-time scene information, the vehicle's path planning allows the navigation route and parking plan to simultaneously integrate the user's personalized preferences and the actual needs of the current environment, thereby significantly improving driving safety, convenience, and overall user experience.

[0151] Figure 9 A schematic diagram of an embodiment of the electronic device provided by the present invention is shown, such as... Figure 9 As shown, the electronic device may include: a processor 92, a communications interface 94, a memory 96, and a communications bus 98.

[0152] The processor 92, communication interface 94, and memory 96 communicate with each other via communication bus 98. Communication interface 94 is used to communicate with other network elements such as clients or other servers. The processor 92 executes program 90, specifically performing the relevant steps in the above method embodiments.

[0153] Specifically, program 90 may include program code, which includes computer-executable instructions.

[0154] Processor 92 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The electronic device includes one or more processors, which may be processors of the same type, such as one or more CPUs, or processors of different types, such as one or more CPUs and one or more ASICs.

[0155] Memory 96 is used to store program 90. Memory 96 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0156] Specifically, program 90 can be called by processor 92 to cause the electronic device to perform the following operations:

[0157] Acquire the target multidimensional semantic features and real-time scene information of the vehicle's environment from the natural language commands issued by the user.

[0158] Based on the target's multidimensional semantic features, the target's historical itinerary is determined in the itinerary knowledge base, which records the correspondence between different multidimensional semantic features and different historical itineraries;

[0159] Based on the target's historical travel history and real-time scene information, the vehicle's route is planned.

[0160] In one or more embodiments, obtaining the target multidimensional semantic features in the natural language command to be processed issued by the user includes:

[0161] Obtain the natural language instructions to be processed;

[0162] The natural language instruction to be processed is input into the large language model to obtain the target multidimensional semantic features. The large language model is trained based on multiple natural language instructions and multidimensional semantic features from different natural language instructions.

[0163] In one or more embodiments, route planning for the vehicle is performed based on the target historical journey and real-time scene information, including:

[0164] Based on the target location in the target's historical travel history and real-time scene information, determine the navigation route for the vehicle to reach the target location;

[0165] When arriving at the target location based on the navigation route, the target parking space and the parking route to the target parking space are determined based on the scene tags in the target's historical journey and the weather data in the real-time scene information.

[0166] In one or more embodiments, the following is also performed:

[0167] In response to a user's parking space adjustment command, determine the target parking space after adjustment;

[0168] Update the scene tags in the target historical trip based on the adjusted target parking space.

[0169] In one or more embodiments, the target historical itinerary includes at least one location;

[0170] Accordingly, based on the target's historical travel history and real-time scenario information, route planning is performed for the vehicle, including:

[0171] In response to the user's selection action, the target location is determined from at least one location;

[0172] Based on the target location and real-time scene information, the vehicle's route is planned.

[0173] In one or more embodiments, the following is also performed:

[0174] Collect location data, scene tags, and behavioral data of the vehicle during its current journey;

[0175] Based on location data, scene tags, and behavioral data in the current itinerary, construct the first multi-dimensional semantic feature;

[0176] Update the itinerary knowledge base based on the current itinerary and the first multidimensional semantic features.

[0177] In one or more embodiments, the scene tags include: parking preferences, companion information, time information, and weather information.

[0178] As can be seen from the above, the electronic device provided in this embodiment of the invention can accurately parse the key elements in the user's fuzzy expression by acquiring the target multidimensional semantic features in the natural language command to be processed issued by the user. Thus, the user can understand the intent without having to input an exact address, greatly reducing the threshold and complexity of human-vehicle interaction. By acquiring real-time scene information of the vehicle's environment, the vehicle can perceive the current dynamic environment, providing an environmental basis for subsequent personalized adaptation. Based on the target multidimensional semantic features, the target historical route is determined in the trip knowledge base, enabling the vehicle to accurately recall past locations based on the user's historical behavior habits, realizing an intelligent upgrade from address navigation to memory navigation. Based on the target historical route and real-time scene information, the vehicle performs route planning, enabling the navigation route and parking plan to simultaneously integrate the user's personalized preferences and the actual needs of the current environment, thereby significantly improving driving safety, convenience, and overall user experience.

[0179] This invention provides a computer-readable storage medium storing at least one executable instruction that, when executed on a vehicle audio adjustment device / electronic device, causes the vehicle audio adjustment device / electronic device to perform the vehicle audio adjustment method in any of the above method embodiments.

[0180] Specifically, the executable instructions can be used to cause the vehicle's path planning device / vehicle to perform the following operations:

[0181] Acquire the target multidimensional semantic features and real-time scene information of the vehicle's environment from the natural language commands issued by the user.

[0182] Based on the target's multidimensional semantic features, the target's historical itinerary is determined in the itinerary knowledge base, which records the correspondence between different multidimensional semantic features and different historical itineraries;

[0183] Based on the target's historical travel history and real-time scene information, the vehicle's route is planned.

[0184] In one or more embodiments, obtaining the target multidimensional semantic features in the natural language command to be processed issued by the user includes:

[0185] Obtain the natural language instructions to be processed;

[0186] The natural language instruction to be processed is input into the large language model to obtain the target multidimensional semantic features. The large language model is trained based on multiple natural language instructions and multidimensional semantic features from different natural language instructions.

[0187] In one or more embodiments, route planning for the vehicle is performed based on the target historical journey and real-time scene information, including:

[0188] Based on the target location in the target's historical travel history and real-time scene information, determine the navigation route for the vehicle to reach the target location;

[0189] When arriving at the target location based on the navigation route, the target parking space and the parking route to the target parking space are determined based on the scene tags in the target's historical journey and the weather data in the real-time scene information.

[0190] In one or more embodiments, the following is also performed:

[0191] In response to a user's parking space adjustment command, determine the target parking space after adjustment;

[0192] Update the scene tags in the target historical trip based on the adjusted target parking space.

[0193] In one or more embodiments, the target historical itinerary includes at least one location;

[0194] Accordingly, based on the target's historical travel history and real-time scenario information, route planning is performed for the vehicle, including:

[0195] In response to the user's selection action, the target location is determined from at least one location;

[0196] Based on the target location and real-time scene information, the vehicle's route is planned.

[0197] In one or more embodiments, the following is also performed:

[0198] Collect location data, scene tags, and behavioral data of the vehicle during its current journey;

[0199] Based on location data, scene tags, and behavioral data in the current itinerary, construct the first multi-dimensional semantic feature;

[0200] Update the itinerary knowledge base based on the current itinerary and the first multidimensional semantic features.

[0201] In one or more embodiments, the scene tags include: parking preferences, companion information, time information, and weather information.

[0202] As can be seen from the above, the vehicle path planning device / electronic device provided in this embodiment of the invention can accurately parse the key elements in the user's fuzzy expression by acquiring the target multidimensional semantic features in the natural language command to be processed issued by the user. Thus, the user can understand the intent without inputting an exact address, greatly reducing the threshold and complexity of human-vehicle interaction. By acquiring the real-time scene information of the vehicle's environment, the vehicle can perceive the current dynamic environment, providing an environmental basis for subsequent personalized adaptation. Based on the target multidimensional semantic features, the target historical route is determined in the trip knowledge base, enabling the vehicle to accurately recall past locations based on the user's historical behavior habits, realizing an intelligent upgrade from address navigation to memory navigation. Based on the target historical route and real-time scene information, the vehicle performs path planning, enabling the navigation route and parking plan to simultaneously integrate the user's personalized preferences and the actual needs of the current environment, thereby significantly improving driving safety, convenience, and overall user experience.

[0203] This invention provides a computer program product, including a computer program that, when executed by a processor, implements the above-described vehicle path planning method.

[0204] Its implementation principle and technical effects are as disclosed above.

[0205] The description of the various embodiments above tends to emphasize the differences between the various embodiments. The similarities or similarities between them can be referred to, and for the sake of brevity, they will not be repeated here.

[0206] The methods disclosed in the various method embodiments provided by this invention can be arbitrarily combined without conflict to obtain new method embodiments.

[0207] The features disclosed in the various product embodiments provided by this invention can be arbitrarily combined without conflict to obtain new product embodiments.

[0208] The features disclosed in the various method or device embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new method or device embodiments.

[0209] It should be noted that the aforementioned computer-readable storage media can be ROM, Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Ferromagnetic Random Access Memory (FRAM), Flash Memory, Magnetic Surface Memory, Optical Disc, or Compact Disc Read-Only Memory (CD-ROM), etc. It can also be various vehicles that include one or any combination of the above-mentioned storage media.

[0210] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0211] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0212] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware nodes. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, vehicle terminal, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0213] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatuses, devices, and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0214] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0215] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The algorithms or displays provided herein for the functions specified in the boxes or boxes are not inherently related to any particular computer, virtual system, or other device. Furthermore, the embodiments of this invention are not directed to any particular programming language.

[0216] It should be noted that the above embodiments are illustrative of the invention and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names. The steps in the above embodiments, unless otherwise specified, should not be construed as limiting the order of execution.

Claims

1. A method of path planning for a vehicle, characterized by, The method includes: Acquire the target multidimensional semantic features and real-time scene information of the vehicle's environment from the natural language commands issued by the user. Based on the target's multidimensional semantic features, the target's historical itinerary is determined in the itinerary knowledge base, which records the correspondence between different multidimensional semantic features and different historical itineraries; Based on the target's historical travel history and the real-time scene information, the vehicle's route is planned.

2. The method of claim 1, wherein, The acquisition of target multidimensional semantic features from the user-issued natural language command to be processed includes: Obtain the natural language instruction to be processed; The natural language instruction to be processed is input into a large language model to obtain the target multidimensional semantic features. The large language model is trained based on multiple natural language instructions and multidimensional semantic features from different natural language instructions.

3. The method according to claim 1 or 2, characterized in that, The step of planning a route for the vehicle based on the target historical travel history and the real-time scene information includes: Based on the target locations in the target's historical travel history and the real-time scene information, determine the navigation route for the vehicle to reach the target location; When arriving at the target location based on the navigation route, the target parking space and the parking route to the target parking space are determined according to the scene tags in the target's historical journey and the weather data in the real-time scene information.

4. The method according to claim 3, characterized in that, The method further includes: In response to a user's parking space adjustment command, determine the target parking space after adjustment; Update the scene labels in the target historical trip based on the adjusted target parking space.

5. The method according to claim 1 or 2, characterized in that, The target historical itinerary includes at least one location; Accordingly, the step of planning a route for the vehicle based on the target historical itinerary and the real-time scene information includes: In response to the user's selection action, the target location is determined among the at least one location; Based on the target location and the real-time scene information, the vehicle performs route planning.

6. The method according to claim 1 or 2, characterized in that, The method further includes: Collect location data, scene tags, and behavior data of the vehicle during its current journey; Based on the location data, scene tags, and behavior data in the current itinerary, a first multi-dimensional semantic feature is constructed; The itinerary knowledge base is updated based on the current itinerary and the first multidimensional semantic features.

7. The method according to claim 3, characterized in that, The scene tags include: parking preferences, companion information, time information, and weather information.

8. A path planning device for a vehicle, characterized in that, The device includes: The acquisition module is used to acquire the target multidimensional semantic features and real-time scene information of the vehicle's environment from the natural language commands issued by the user. The determination module is used to determine the target historical itinerary in the itinerary knowledge base based on the target multidimensional semantic features. The itinerary knowledge base records the correspondence between different multidimensional semantic features and different historical itineraries. The processing module is used to perform route planning for the vehicle based on the target's historical travel history and the real-time scene information.

9. An electronic device, characterized in that, include: The processor, memory, communication interface, and communication bus are provided, wherein the processor, memory, and communication interface communicate with each other via the communication bus. The memory is used to store at least one executable instruction that causes the processor to perform the operation of the vehicle path planning method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The storage medium stores at least one executable instruction; When the executable instructions are executed on the vehicle / vehicle path planning device, the electronic device / vehicle path planning device causes the vehicle path planning method as described in any one of claims 1-7 to perform the operation.

Images