Method and device for determining a parking position of a vehicle, vehicle

Abstract

The application provides a parking position determination method of a vehicle, the method comprising: determining a target area according to a destination position of the vehicle, the maximum distance between the target area and the destination position being less than a distance threshold; obtaining real-time traffic information of the target area at a current time, the real-time traffic information comprising at least one road element; and determining a parking position of the vehicle in the target area at the current time according to the real-time traffic information and a target machine learning model. The application can make the vehicle park at the determined parking position, and make the user get on and off the vehicle more conveniently and safely at the determined parking position.

Description

Technical Field

[0001] This application relates to the field of autonomous driving, and more particularly to a method and apparatus for determining the parking location of a vehicle, and a vehicle. Background Technology

[0002] With the development of computer technology, the application of autonomous driving technology is becoming increasingly widespread. In autonomous driving scenarios, vehicles can automatically stop to allow users to get on and off.

[0003] In existing solutions, when a user needs to get on or off the vehicle, they can send a driving instruction containing their destination location to the vehicle. The vehicle then determines the destination location based on the received instruction. The vehicle then travels to the destination location and stops there, allowing the user to get on or off at that location.

[0004] However, the destination location may have obstacles or poor road conditions (such as water accumulation, silt, or deep potholes), which may prevent the vehicle from parking or make it difficult for users to get in and out of the vehicle after parking. Summary of the Invention

[0005] This application provides a method and device for determining the parking location of a vehicle, as well as a vehicle, making it more convenient and safer for the vehicle to park at a determined parking location and for users to get on and off the vehicle at a determined parking location.

[0006] In a first aspect, this application provides a method for determining the parking location of a vehicle, the method comprising: determining a target area based on the vehicle's destination location, wherein the maximum distance between the target area and the destination location is less than a distance threshold; acquiring real-time traffic information of the target area at the current moment, wherein the real-time traffic information includes at least one road element; and determining the parking location of the vehicle in the target area at the current moment based on the real-time traffic information and a target machine learning model.

[0007] This method determines the vehicle's current parking position in the target area based on real-time traffic information and the target machine learning model. The method takes into account changes in traffic conditions during the parking position determination process, enabling the vehicle to park at the determined parking position and making it more convenient and safer for users to get on and off the vehicle at the determined parking position.

[0008] Furthermore, the target machine learning model is a model obtained by training on data and using algorithms to discover the patterns and essence of the data. Real-time traffic information typically includes a wide variety of road elements, making its information quite complex. The target machine learning model can process these diverse road elements, thereby accurately and efficiently determining the optimal stopping location from multiple candidate locations. Specifically, during training, the target machine learning model can adaptively determine the importance of various road elements and the redundancy among them based on the training labels. This allows the trained target machine learning model to accurately determine the importance and redundancy of various road elements corresponding to each candidate location, and thus determine and output the optimal stopping location from multiple candidate locations based on the importance and redundancy of these road elements.

[0009] The target machine learning model may include at least one of the following: VisionTransformers model, Convolutional Neural Networks (CNN) model, Graph Transformers model, or Graph Neural Networks (GNN) model.

[0010] The target region is the area near the destination location, which may or may not include the destination location. For example, the target region can be a circular or rectangular area with the destination location as its geometric center.

[0011] The destination location can be determined by the vehicle based on the received driving instructions. Users can send driving instructions through applications running on peripheral devices or terminal devices (such as mobile phones, computers, and wristbands).

[0012] In one possible implementation, the real-time traffic information includes at least one of the following road elements: drivable area, vehicles, roadblocks, pedestrians, guardrails, flooded areas, muddy areas, and potholes. The road elements included in the real-time traffic information are typically dynamic.

[0013] Real-time traffic information can be obtained by identifying target areas using sensor systems in vehicles. For example, the sensor system can acquire images of the target area using radar or cameras, and then a computer vision system can process and analyze these images to obtain real-time traffic information for the target area. The computer vision system can obtain real-time traffic information through at least one of the following processing and analysis methods: obstacle detection, lane detection, drivable area segmentation, water accumulation area segmentation, mud area segmentation, and pothole area segmentation.

[0014] Optionally, some road elements in real-time traffic information have the same meaning. For example, vehicles, roadblocks, pedestrians, or guardrails in real-time traffic information refer to areas outside the drivable area in the target region (also known as non-drivable areas). Therefore, real-time traffic information may not include vehicles, roadblocks, pedestrians, or guardrails, as long as the drivable area does not include vehicles, roadblocks, pedestrians, or guardrails.

[0015] In one possible implementation, the method further includes: acquiring map information of the target area at the current time, the map information including at least one road element; determining the stopping position of the vehicle in the target area at the current time based on the real-time traffic information and the target machine learning model, including: determining the stopping position based on the real-time traffic information, the map information, and the target machine learning model.

[0016] Optionally, map information can be obtained from a map, which can be a high-definition map. High-definition maps typically represent road elements using at least one map layer, with each map layer representing one type of road element. For example, map information may include at least one of the following road elements: lane lines, sidewalks, curbs, road gradient, speed limits, traffic flow, and drivable areas.

[0017] Map information can be pre-stored in vehicles or in the cloud.

[0018] In one possible implementation, the method further includes: updating the map information using the real-time traffic information to obtain updated map information, the updated map information including road elements in the real-time traffic information; determining the parking location based on the real-time traffic information, the map information, and the target machine learning model includes: determining the parking location based on the updated map information and the target machine learning model.

[0019] For any road element in the real-time traffic information, if the road element already exists in the map information, the road element in the map information will be updated to match the road element in the real-time traffic information. If the road element does not exist in the map information, the road element will be added to the map information.

[0020] Some road elements in real-time traffic information and map information have the same meaning. Vehicles can represent road elements with the same meaning using different map layers or the same map layer. This application does not limit this.

[0021] Utilizing maps to store real-time traffic information is advantageous, as map data structures are well-suited for this purpose. Furthermore, updated map information can be used for subsequent autonomous driving and automatic parking (e.g., route planning or determining the next stopping location), thus improving the utilization rate of real-time traffic information.

[0022] In one possible implementation, determining the parking location based on the real-time traffic information, the map information, and the target machine learning model includes: inputting the real-time traffic information and the map information into the target machine learning model to obtain the parking location output by the target machine learning model.

[0023] In this implementation, the target area includes multiple candidate locations. A target machine learning model is used to determine the stopping priority for each candidate location within the target area based on the input real-time traffic information, and outputs the candidate location with the highest stopping priority. This stopping priority can be represented by letters or Arabic numerals. Optionally, the multiple candidate locations can be all locations within the target area.

[0024] The target machine learning model can be obtained through a supervised learning training process. Optionally, a training region can be determined based on the training location, where the maximum distance between the training region and the training location is less than a distance threshold, and the training region includes multiple locations. The initial model is trained using real-time traffic information and map information of the training region (i.e., real-time traffic information and map information corresponding to multiple locations in the training region) and training labels corresponding to multiple locations in the training region, to obtain the target machine learning model.

[0025] During the training process of this model, the process of a vehicle automatically driving to the training location can be simulated to obtain real-time traffic information and map information of the training area.

[0026] In one possible implementation, the method further includes: obtaining label parameters corresponding to the plurality of locations through simulation, wherein the label parameters corresponding to the location include at least one of the following: the time taken to walk from the training location to the location, the delay of traffic flow at the location when the vehicle stops at the location, the parking status at the location, and the probability of an accident occurring at the location; and determining the training label corresponding to the location based on the label parameters corresponding to the location.

[0027] Different types of models have different input data formats. Therefore, when the format of real-time traffic information differs from the input data format of the target machine learning model, the real-time traffic information needs to be encoded to match the input format requirements of the target machine learning model. The encoded data is then input into the target machine learning model to obtain the parking location output by the model.

[0028] In one possible implementation, the input format of the target machine learning model is required to be a feature map, such as a Vision Transformers model or a CNN model. In this case, real-time traffic information data needs to be encoded into feature maps. Determining the stop location based on the real-time traffic information, the map information, and the target machine learning model includes: obtaining at least one feature map based on the real-time traffic information and the map information, the feature map indicating at least one road element in the real-time traffic information and the map information; and inputting the at least one feature map into the target machine learning model to obtain the stop location output by the target machine learning model.

[0029] For example, a raster map corresponding to the target area from a bird's-eye view can be determined. Then, based on at least one road element in the real-time traffic information, at least one road element corresponding to the position of the center point of each raster is determined, resulting in a feature map used to indicate the at least one road element.

[0030] In this implementation, the target region includes multiple candidate locations. The target machine learning model is used to determine the docking priority of each candidate location in the target region based on at least one feature map as input, and outputs the candidate location with the highest docking priority. The target machine learning model can be obtained through a supervised learning training process, which is described below.

[0031] Optionally, a training region can be determined based on the training location, and real-time traffic information of the training region can be obtained. Then, at least one feature map is obtained based on the real-time traffic information of the training region, and the initial model is trained using at least one feature map and training labels corresponding to multiple locations in the training region to obtain the target machine learning model.

[0032] In one possible implementation, the input format of the target machine learning model is required to be a dataset, such as a Graph Transformers model or a GNN model. In this case, the real-time traffic information data needs to be encoded into a dataset. The dataset can include non-Euclidean data formats, which refer to discretely distributed data formats. The step of determining the parking location based on the real-time traffic information, the map information, and the target machine learning model includes: determining initial candidate location sets corresponding to road elements in the real-time traffic information and the map information, wherein the initial candidate location sets include at least one candidate location from a plurality of candidate locations; merging the initial candidate location sets corresponding to all road elements in the real-time traffic information and the map information to obtain a candidate location set; obtaining data sets corresponding to the candidate locations in the candidate location set based on the road elements corresponding to the candidate locations in the candidate location set in the real-time traffic information and the map information, wherein the data set corresponding to any candidate location in the candidate location set is used to indicate the road element corresponding to any candidate location in the real-time traffic information and the map information; and inputting the data sets corresponding to the candidate locations in the candidate location set into the target machine learning model to obtain the parking location output by the target machine learning model.

[0033] For example, for each road element in the real-time traffic information and map information, its corresponding initial candidate location set can be the key locations in that road element.

[0034] In this implementation, the target area includes multiple candidate locations. The target machine learning model is used to determine the docking priority of each candidate location in the input candidate location set based on the corresponding datasets, and outputs the candidate location with the highest docking priority. The target machine learning model can be obtained through a supervised learning training process, which is described below.

[0035] Optionally, a training region can be determined based on the training location, and real-time traffic information of the training region can be obtained. Then, an initial location set corresponding to each road element in the real-time traffic information of the training region is determined. This initial location set includes at least one of multiple locations in the training region. The initial location sets corresponding to all road elements in the real-time traffic information of the training region are merged to obtain a location set. Based on the road elements corresponding to the locations in the location set, a data set corresponding to each location in the location set is obtained. The data set corresponding to any location in the location set is used to indicate the road element corresponding to any location in the real-time traffic information of the training region. Finally, the initial model is trained using the data set corresponding to each location in the location set and the training labels to obtain the target machine learning model.

[0036] In one possible implementation, the target area includes multiple candidate locations. Determining the parking location based on the real-time traffic information, the map information, and the target machine learning model includes: determining one or more feature data corresponding to each of the multiple candidate locations based on the real-time traffic information, the map information, and the destination location. The one or more feature data includes at least one of the following: parking availability feature, distance feature, traffic obstruction feature, and traffic flow feature; inputting the one or more feature data corresponding to each of the multiple candidate locations into the target machine learning model to obtain the parking location output by the target machine learning model; wherein, the parking availability feature represents the parking status of the candidate location, the distance feature represents the distance between the destination location and the candidate location, the traffic obstruction feature represents the degree of obstruction of traffic flow when the vehicle is parked at the candidate location, and the traffic flow feature represents the traffic flow at the candidate location.

[0037] The parking feature can be represented as F1, and the parking status can include at least one of the following: not parking, parking is possible, parking is possible but the road surface condition is poor (e.g., the candidate location is in a waterlogged area, a muddy area or a pothole area, etc.).

[0038] Distance features can be represented by two distances, F2 and F3. F2 represents the distance between the target location and the candidate location in the first direction, and F3 represents the distance between the target location and the candidate location in the second direction. For example, the first direction can be the direction in which the length of the vehicle is located, and the second direction can be the direction in which the width of the vehicle is located.

[0039] The traffic flow obstruction feature can be represented as F4. F4 can be obtained by the width of the road where the candidate location is located and the width of the vehicle. For example, F4 can be the ratio of the width of the vehicle to the width of the road where the candidate location is located.

[0040] Traffic flow characteristics can be represented as F5, which can be obtained from the coordinates of candidate locations and map information. For example, a vehicle can determine the traffic flow corresponding to the coordinates of a candidate location based on the traffic flow data in the map information, and then obtain the F5 value.

[0041] In this implementation, the target machine learning model is used to determine the docking priority of each of the multiple candidate locations based on one or more feature data corresponding to each of the input candidate locations, and outputs the candidate location with the highest docking priority. The target machine learning model can be obtained through a supervised learning training process, which is described below.

[0042] Optionally, a training area can be determined based on the training location, and real-time traffic information of the training area can be obtained. Then, based on the real-time traffic information of the training area and the destination location, one or more feature data points corresponding to multiple locations within the training area are determined. Finally, the initial model is trained using the one or more feature data points corresponding to the multiple locations and the training labels to obtain the target machine learning model.

[0043] In one possible implementation, the method further includes: repeatedly executing the parking location determination process until the vehicle stops moving; the parking location determination process includes: acquiring real-time traffic information of the target area at the current moment; and determining the parking location of the vehicle in the target area at the current moment based on the real-time traffic information and the target machine learning model.

[0044] Since real-time traffic information is constantly changing during the vehicle's autonomous driving process, cyclically executing the stopping location determination process allows for the determination of the optimal stopping location based on the changing real-time traffic information. For example, the vehicle can re-execute the aforementioned processes 201 to 204 every interval t, where t can be 0.25 seconds, 0.3 seconds, or 0.5 seconds, etc. This embodiment of the application does not limit the value of t.

[0045] Secondly, this application provides a vehicle parking location determination device, the device comprising: a first determination module, configured to determine a target area based on the vehicle's destination location, wherein the maximum distance between the target area and the destination location is less than a distance threshold; a first acquisition module, configured to acquire real-time traffic information of the target area at the current moment, wherein the real-time traffic information includes at least one road element; and a second determination module, configured to determine the parking location of the vehicle in the target area at the current moment based on the real-time traffic information and a target machine learning model.

[0046] In one possible implementation, the real-time traffic information includes at least one of the following road elements: drivable area, vehicles, roadblocks, pedestrians, guardrails, waterlogged areas, muddy areas, and potholes.

[0047] In one possible implementation, the device further includes: a second acquisition module, configured to acquire map information of the target area at the current time, the map information including at least one road element; and a second determination module, comprising: a determination unit, configured to determine the parking location based on the real-time traffic information, the map information, and the target machine learning model.

[0048] In one possible implementation, the device further includes: an update module, configured to update the map information using the real-time traffic information to obtain updated map information, the updated map information including road elements in the real-time traffic information; and a determination unit, specifically configured to determine the parking location based on the updated map information and the target machine learning model.

[0049] In one possible implementation, the determining unit is specifically used to input the real-time traffic information and the map information into the target machine learning model to obtain the parking location output by the target machine learning model.

[0050] In one possible implementation, the determining unit is specifically configured to obtain at least one feature map based on the real-time traffic information and the map information, the feature map indicating at least one road element in the real-time traffic information and the map information; and input the at least one feature map into the target machine learning model to obtain the parking location output by the target machine learning model.

[0051] In one possible implementation, the target area includes multiple candidate locations. The determining unit is specifically configured to determine initial candidate location sets corresponding to road elements in the real-time traffic information and the map information, respectively. The initial candidate location sets include at least one candidate location among the multiple candidate locations. The initial candidate location sets corresponding to all road elements in the real-time traffic information and the map information are merged to obtain a candidate location set. Based on the road elements corresponding to the candidate locations in the candidate location set in the real-time traffic information and the map information, a data set corresponding to each candidate location in the candidate location set is obtained. The data set corresponding to each candidate location in the candidate location set is input into the target machine learning model to obtain the parking location output by the target machine learning model.

[0052] In one possible implementation, the target area includes multiple candidate locations. The determining unit is specifically configured to determine one or more feature data corresponding to each of the multiple candidate locations based on the real-time traffic information, the map information, and the destination location. The one or more feature data includes at least one of the following feature data: parking availability feature, distance feature, traffic obstruction feature, and traffic flow feature. The one or more feature data corresponding to each of the multiple candidate locations are input into the target machine learning model to obtain the parking location output by the target machine learning model. Wherein, the parking availability feature represents the parking status of the candidate location, the distance feature represents the distance between the destination location and the candidate location, the traffic obstruction feature represents the degree of obstruction of traffic flow when the vehicle is parked at the candidate location, and the traffic flow feature represents the traffic flow of the candidate location.

[0053] In one possible implementation, the device further includes: a loop module for repeatedly executing the parking location determination process until the vehicle stops moving; the parking location determination process includes: acquiring real-time traffic information of the target area at the current moment; and determining the parking location of the vehicle in the target area at the current moment based on the real-time traffic information and a target machine learning model.

[0054] Thirdly, this application provides a vehicle parking location determination device, comprising: one or more processors; a memory for storing one or more computer programs or instructions; and when the one or more computer programs or instructions are executed by the one or more processors, causing the one or more processors to implement the method as described in any one of the first aspects.

[0055] Fourthly, this application provides a vehicle parking location determination device, comprising: a processor, the processor being configured to perform the method as described in any one of the first aspects.

[0056] Fifthly, this application provides a computer-readable storage medium including a computer program or instructions that, when executed on a computer, cause the computer to perform the method described in any one of the first aspects.

[0057] In a sixth aspect, this application provides a computer program product comprising computer program code, which, when run on a computer, causes the computer to perform the method described in any one of the first aspects.

[0058] In a seventh aspect, this application provides a vehicle, the vehicle including a control system and a vehicle parking position determining device as described in any one of the second to fourth aspects; the vehicle parking position determining device is used to determine the parking position of the vehicle in a target area; the control system is used to control the parking of the vehicle according to the determined parking position. Attached Figure Description

[0059] Figure 1 A schematic diagram illustrating the process of automatic driving and automatic parking of a vehicle, provided as an embodiment of this application;

[0060] Figure 2 A functional block diagram of a vehicle provided in an embodiment of this application;

[0061] Figure 3 A flowchart illustrating a method for determining the parking location of a vehicle, provided in an embodiment of this application;

[0062] Figure 4 This is a schematic diagram illustrating a method for determining a parking location based on real-time traffic information and a target machine learning model, provided in an embodiment of this application.

[0063] Figure 5 A schematic diagram illustrating another method for determining a parking location based on real-time traffic information and a target machine learning model, provided in an embodiment of this application.

[0064] Figure 6 A schematic diagram illustrating another method for determining a parking location based on real-time traffic information and a target machine learning model, provided in an embodiment of this application.

[0065] Figure 7 A schematic diagram illustrating another method for determining a parking location based on real-time traffic information and a target machine learning model, provided in an embodiment of this application.

[0066] Figure 8 This is a schematic diagram illustrating a scenario where a vehicle is parked at a candidate location, as provided in an embodiment of this application.

[0067] Figure 9 A flowchart illustrating another method for determining the parking location of a vehicle provided in an embodiment of this application;

[0068] Figure 10 A block diagram of a vehicle parking position determination device provided in an embodiment of this application;

[0069] Figure 11 A block diagram of another vehicle parking position determination device provided in an embodiment of this application;

[0070] Figure 12 A block diagram of a second determining module provided in an embodiment of this application;

[0071] Figure 13 A block diagram of another vehicle parking position determination device provided in an embodiment of this application;

[0072] Figure 14 A block diagram of another vehicle parking position determination device provided in an embodiment of this application;

[0073] Figure 15 A schematic diagram of a vehicle parking position determination device provided in an embodiment of this application;

[0074] Figure 16 This is a schematic diagram of another vehicle parking position determination device provided in an embodiment of this application. Detailed Implementation

[0075] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0076] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0077] The terms "first" and "second," etc., used in the specification and claims of this application are used to distinguish different objects, rather than to describe a specific order of objects. For example, "first range" and "second range," etc., are used to distinguish different ranges, rather than to describe a specific order of ranges.

[0078] In the embodiments of this application, the words "in one example," "examplely," or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "in one example," "examplely," or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the words "in one example," "examplely," or "for example" is intended to present the relevant concepts in a specific manner.

[0079] In the description of the embodiments in this application, unless otherwise stated, "at least one" means one or more, and "multiple" means two or more. For example, multiple processing units refer to two or more processing units; multiple systems refer to two or more systems.

[0080] In autonomous driving scenarios, vehicles can be configured to operate in fully or partially autonomous driving modes. For example, a vehicle can control itself while in autonomous driving mode, and the current state of the vehicle and its surrounding environment can be determined through human intervention. This includes identifying the possible behaviors of at least one other vehicle in the surrounding environment and determining the confidence level corresponding to the probability of that other vehicle performing its possible behavior. The vehicle can then be controlled based on this determined information. When in autonomous driving mode, the vehicle can be positioned to operate without human interaction.

[0081] For example, autonomous driving scenarios can include automatic parking, automatic recall, and automatic shuttle scenarios. In an automatic parking scenario, the vehicle can automatically park near the parking lot entrance, and automatically park itself after the user gets out. In an automatic recall scenario, the vehicle automatically drives from its current location to the user's designated location and automatically parks, allowing the user to get on and off the vehicle after it automatically stops. In an automatic shuttle scenario, the vehicle automatically stops temporarily during its autonomous driving process to allow users to get on and off.

[0082] The following explains the automatic driving and parking processes of vehicles in autonomous driving scenarios. Please refer to [link / reference]. Figure 1 , Figure 1 This is a schematic diagram illustrating the process of automatic driving and automatic parking of a vehicle, as provided in an embodiment of this application. When a user needs to get on or off the vehicle, they can send a driving command carrying the destination location to the vehicle. The vehicle receives the driving command carrying the destination location, then plans a driving route based on the destination location and starts driving from the current location towards the destination location. When the vehicle reaches the vicinity of the destination location (for example, tens of meters away from the destination location), it determines the parking position, parks itself according to the parking position, and then the user gets on or off the vehicle.

[0083] In related technologies, vehicles typically directly determine the destination location as the parking spot. However, the destination location may have obstacles, such as temporary construction, temporary fences, or being occupied by other vehicles, preventing the vehicle from parking automatically. In addition, the road conditions at the destination location may be poor, such as having standing water, silt, or deep potholes, which reduces the convenience for users to get in and out of the vehicle, thus affecting the user's experience.

[0084] This application provides a method for determining the parking position of a vehicle, in order to determine the parking position during the automatic driving and automatic parking processes of the vehicle. This method for determining the parking position of a vehicle can be applied to vehicles; please refer to [link / reference needed]. Figure 2 , Figure 2This application provides a functional block diagram of a vehicle 100, which may include various subsystems, such as a driving system 101, a sensor system 102, a control system 103, one or more peripheral devices 104, a power supply 105, a user interface 106, and a computer system 107. Optionally, the vehicle 100 may include more or fewer subsystems, and each subsystem may include multiple components. Furthermore, each subsystem and component of the vehicle 100 can be interconnected via wired or wireless means, for example, via an in-vehicle communication network, which may include a Controller Area Network Bus (CAN Bus) or an in-vehicle Ethernet network.

[0085] The propulsion system 101 may include components for providing power to the vehicle 100. In one embodiment, the propulsion system 101 may include an engine 1011, an energy source 1012, a transmission 1013, and wheels 1014.

[0086] Engine 1011 can be an internal combustion engine, an electric motor, an air compression engine, or a combination of one or more other types of engines. For example, a combination of multiple engines could include a hybrid engine consisting of a gasoline engine and an electric motor, or a hybrid engine consisting of an internal combustion engine and an air compression engine. Engine 1011 is used to convert energy source 1012 into mechanical energy.

[0087] Energy source 1012 may include gasoline, diesel, other petroleum-based fuels, propane, other compressed gas-based fuels, ethanol, solar panels, batteries, and other sources of electricity. Energy source 1012 may also provide energy to other systems of vehicle 100.

[0088] The transmission 1013 can transmit mechanical power from the engine 1011 to the wheels 1014. The transmission 1013 may include a gearbox, a differential, and a drive shaft. In one embodiment, the transmission 1013 may also include other components, such as a clutch. The drive shaft may include one or more axles that can be coupled to one or more wheels 1014.

[0089] Sensor system 102 may include several sensors that sense information about vehicle 100 itself and its surrounding environment. For example, sensor system 102 may include a positioning system 1021 (which may be a Global Positioning System (GPS), BeiDou Navigation Satellite System, or other positioning systems), an inertial measurement unit (IMU) 1022, radar 1023 (which may be lidar), a laser rangefinder 1024, a camera 1025, a computer vision system 1026, and a sensor fusion algorithm 1027. Sensor system 102 may also include sensors from the vehicle 100's internal systems (e.g., an in-vehicle air quality monitor, fuel gauge, and oil temperature gauge). Data from one or more of these sensors can be used to detect objects and their corresponding characteristics (position, shape, orientation, speed, etc.). This detection and identification is a key function for the safe operation of vehicle 100.

[0090] The positioning system 1021 can be used to estimate the geographical location of the vehicle 100. The IMU 1022 can be used to sense changes in the position and orientation of the vehicle 100 based on inertial acceleration. In one embodiment, the IMU 1022 can be a combination of an accelerometer and a gyroscope.

[0091] Radar 1023 can use radio signals to sense objects in the surrounding environment of vehicle 100. In some embodiments, in addition to sensing objects, radar 1023 can also be used to sense the speed and / or direction of travel of objects. Radar 1023 includes, but is not limited to, lidar and millimeter-wave radar.

[0092] The laser rangefinder 1024 can use lasers to sense objects in the environment in which the vehicle 100 is located. In some embodiments, the laser rangefinder 1024 may include one or more laser sources, a laser scanner, and one or more detectors, as well as other system components.

[0093] Camera 1025 can be used to capture multiple images of the surrounding environment of vehicle 100. Camera 1025 can be a still camera or a video camera.

[0094] Computer vision system 1026 can be used to process and analyze images captured by camera 1025 to identify objects and / or features in the environment surrounding vehicle 100. The objects and / or features may include traffic signals, road boundaries, obstacles, and road surface conditions. Computer vision system 1026 may use object recognition algorithms, structure from motion (SFM) algorithms, video tracking, and other computer vision techniques. In some embodiments, computer vision system 1026 can be used to map the environment, track objects, estimate object velocities, etc.

[0095] The control system 103 is used to perform operations controlling the vehicle 100 and its components. The control system 103 may include various components, such as... Figure 2 As shown, the control system 103 may include a steering system 1031, a throttle 1032, a braking unit 1033, a route planning system 1034, and an obstacle avoidance system 1035.

[0096] The direction of travel of the vehicle 100 can be adjusted by operating the steering system 1031. For example, in one embodiment, this can be a steering wheel system. The accelerator 1032 is used to control the operating speed of the engine 1011 and thus the speed of the vehicle 100.

[0097] Braking unit 1033 is used to control the deceleration of vehicle 100. Braking unit 1033 can use friction to slow down wheel 1014. In other embodiments, braking unit 1033 can convert the kinetic energy of wheel 1014 into electric current. Braking unit 1033 may also take other forms to slow down the rotational speed of wheel 1014 to control the speed of vehicle 100.

[0098] The route planning system 1034 is used to determine the driving route of the vehicle 100. In some embodiments, the route control system 1034 may combine data from the sensor system 102 (e.g., the positioning system 1021 and the computer vision system 1026) and one or more predetermined maps to plan a driving route for the vehicle 100 that avoids potential obstacles in the environment.

[0099] The obstacle avoidance system 1035 is used to generate control values ​​for throttle, brake and steering angle based on the driving route or driving trajectory output by the route planning system 1034, thereby controlling the steering system 1031, throttle 1032 and braking unit 1033.

[0100] Of course, in one instance, the control system 103 may include additional or alternative components besides those shown and described. Alternatively, some of the components shown above may be reduced.

[0101] Vehicle 100 interacts with external sensors, other vehicles, other computer systems, or users via one or more peripheral devices 104. Peripheral devices 104 may include a wireless communication system 1041, an on-board computer 1042, a microphone 1043, and / or a speaker 1044.

[0102] In some embodiments, peripheral device 104 provides a means for a user of vehicle 100 to interact with user interface 106. For example, on-board computer 1042 can provide information to the user of vehicle 100. User interface 106 can also operate on-board computer 1042 to receive user input. In one implementation, on-board computer 1042 can be operated via a touchscreen. In other cases, peripheral device 104 provides a means for vehicle 100 to communicate with other devices located within the vehicle. For example, microphone 1043 can receive audio (e.g., voice commands or other audio input) from the user of vehicle 100. Similarly, speaker 1044 can output audio to the user of vehicle 100.

[0103] The wireless communication system 1041 can communicate wirelessly with one or more devices directly or via a communication network. For example, the wireless communication system 1041 can communicate using third-generation mobile communication technology (3G), such as code division multiple access (CDMA) or Global System for Mobile Communications (GSM), or fourth-generation mobile communication technology (4G), such as LTE (long term evolution). Alternatively, it can communicate using fifth-generation mobile communication technology (5G). The wireless communication system 1041 can communicate using wireless fidelity (WiFi) and wireless local area network (WLAN). In some embodiments, the wireless communication system 1041 can communicate directly with devices using an infrared link, Bluetooth, or Zigbee. Other wireless protocols, such as various vehicle communication systems, for example, wireless communication system 1041 may include one or more dedicated short range communications (DSRC) devices, which may include public and / or private data communication devices between vehicles and / or roadside stations.

[0104] Power source 105 can provide power to various components of vehicle 100. In one embodiment, power source 105 can be a rechargeable lithium-ion or lead-acid battery. One or more such battery packs can be configured as a power source and provide power to various components of vehicle 100. In some embodiments, power source 105 and energy source 1012 can be implemented together, as in a fully electric vehicle.

[0105] Some or all of the functions of vehicle 100 are controlled by computer system 107. Computer system 107 may include at least one processor 1071, which executes instructions 1073 stored in a non-transitory computer-readable medium such as memory 1072. Computer system 107 may also be multiple computing devices that control individual components or subsystems of vehicle 100 in a distributed manner.

[0106] Processor 1071 can be any conventional processor, such as a commercially available central processing unit (CPU). Alternatively, processor 1071 can be a special-purpose device such as an application-specific integrated circuit (ASIC) or other hardware-based processor. Although Figure 2 The processor 1071, memory 1072, and other components of the computer system 107 are functionally illustrated; however, those skilled in the art will understand that processor 1071 and memory 1072 may actually include multiple other processors or memories not located in the same physical housing. For example, memory 1072 may be a hard disk drive or other storage media located in a housing different from that of computer system 107. Therefore, references to processor 1071 will be understood to include references to a collection of processors or memories that may operate in parallel or not. Unlike using a single processor to perform the steps described herein, some components, such as steering and deceleration components, may each have their own processor that performs calculations only related to the component's specific function; or subsystems such as motion system 101, sensor system 102, and control system 103 may also have their own processors to perform calculations for the relevant tasks of the corresponding subsystem, thereby achieving the corresponding function.

[0107] In all the aspects described herein, processor 1071 may be located remotely from vehicle 100 and may communicate wirelessly with vehicle 100. In other aspects, some of the processes described herein are executed on a processor located within vehicle 100, while others are executed by a remote processor, including taking the necessary steps to perform a single operation.

[0108] In some embodiments, memory 1072 may contain instructions 1073 (e.g., program logic) that can be executed by processor 1071 to perform various functions of vehicle 100, including those described above. Memory 1072 may also contain additional instructions, including instructions for sending data to, receiving data from, interacting with, and / or controlling one or more of the mobility system 101, sensor system 102, control system 103, and peripheral devices 104.

[0109] In addition to instruction 1073, memory 1072 may also store other relevant data, such as map information, route information, vehicle position, direction, speed, and other related information. This information may be used by vehicle 100 or specifically by computer system 107 during operation of vehicle 100 in autonomous, semi-autonomous, and / or manual modes.

[0110] User interface 106 is used to provide information to or receive information from users of vehicle 100. Optionally, user interface 106 may include one or more input / output devices within a set of peripheral devices 104, such as wireless communication system 1041, vehicle computer 1042, microphone 1043, and speaker 1044.

[0111] Computer system 107 can control the functions of vehicle 100 based on input received from various subsystems (e.g., mobility system 101, sensor system 102, and control system 103) and from user interface 106. In some embodiments, computer system 107 is operable to provide control over many aspects of vehicle 100 and its subsystems.

[0112] Optionally, one or more of these components may be installed separately from or associated with vehicle 100. For example, memory 1072 may exist partially or completely separately from vehicle 100. The components may be communicatively coupled together in a wired and / or wireless manner.

[0113] Optionally, the components described above are merely examples. In actual applications, components in each of the above modules may be added or removed as needed. Figure 2 This should not be construed as a limitation on the embodiments of this application.

[0114] Autonomous vehicles traveling on roads, such as vehicle 100 above, can identify objects in their surrounding environment to determine adjustments to their driving trajectory, including adjustments to their speed. These objects can be other vehicles, traffic control equipment, or other types of objects. In some examples, each identified object can be considered independently, and the trajectory planning for the autonomous vehicle, including the speed adjustment, can be determined based on the object's individual characteristics, such as its current speed, acceleration, and distance from the vehicle.

[0115] Optionally, the computing device associated with vehicle 100 (such as...) Figure 2The computer system 107 and computer vision system 1026 can predict the behavior of the identified objects based on the characteristics of the identified objects and the state of the surrounding environment (e.g., traffic, rain, ice on the road, etc.). Optionally, each identified object depends on the behavior of each other, so all identified objects can also be considered together to predict the behavior of a single identified object. The vehicle 100 can plan its driving trajectory (including speed) based on the predicted behavior of the identified objects. And based on the planning result, it determines what state the vehicle will need to adjust to (in terms of speed adjustment, such as acceleration, deceleration, or stopping) for the autonomous vehicle. In this process, other factors can also be considered to determine the driving trajectory of the vehicle 100, such as the lateral position of the vehicle 100 in the road, the curvature of the road, the proximity of static and dynamic objects, etc.

[0116] Based on the planning results, in addition to adjusting the speed of the autonomous vehicle, the computing device can also provide instructions to modify the steering angle of the vehicle 100 so that the autonomous vehicle follows a given trajectory and / or maintains a safe lateral and longitudinal distance from objects near the autonomous vehicle (e.g., cars in adjacent lanes on the road).

[0117] The aforementioned vehicle 100 can be a car, truck, motorcycle, bus, ship, airplane, helicopter, lawnmower, recreational vehicle, amusement park vehicle, construction equipment, tram, golf cart, train, and handcart, etc., and this application embodiment does not impose any special limitations.

[0118] This application provides a method for determining the parking position of a vehicle. This method can be applied to vehicles, such as those described above. Figure 2 The vehicle 100 is shown below. The following describes the method for determining the parking position of the vehicle using two embodiments as examples.

[0119] Example 1: Determine the vehicle's parking location based on real-time traffic information.

[0120] Please refer to Figure 3 , Figure 3 The following is a flowchart illustrating a method for determining the parking location of a vehicle, provided as an embodiment of this application. The method may include the following steps:

[0121] 201. Determine the target area based on the vehicle's destination location, where the maximum distance between the target area and the destination location is less than a distance threshold.

[0122] The target area is the region near the destination location, which may or may not include the destination location. For example, the target area can be a circular or rectangular area with the destination location as its geometric center. When the target area is circular, its radius is less than or equal to a distance threshold; when the target area is rectangular, half of its diagonal is less than or equal to the distance threshold. The distance threshold can be 40 meters, 50 meters, or 75 meters, etc. This application embodiment does not limit the shape of the target area or the specific value of the distance threshold.

[0123] The destination location can be determined by the vehicle (e.g., peripheral device 104 in vehicle 100) based on the received driving instructions. Users can send driving instructions through peripheral device 104 or an application running on a terminal device (e.g., a mobile phone, computer, or wristband). When a user sends a driving instruction through an application running on peripheral device 104, the vehicle directly receives the instruction. When a user sends a driving instruction through an application running on a terminal device, the terminal device can send the instruction to the vehicle through the application's backend server.

[0124] After determining the destination location, the vehicle determines a driving route based on the destination location and travels from the current location to the destination location according to the determined driving route. This process can be carried out by... Figure 2 The route planning system 1034 in the vehicle 100 shown is executed. During the vehicle's journey, when the distance between the current location and the destination location is determined to be less than or equal to a certain distance via positioning, process 201 can be initiated; that is, process 201 is initiated when the vehicle approaches the destination location. For example, this certain distance can be 40 meters, 50 meters, or 75 meters, etc., and this embodiment does not limit this. Alternatively, the vehicle can also initiate process 201 immediately after determining the destination location; this embodiment does not limit the execution time of this process.

[0125] 202. Obtain real-time traffic information for the target area at the current moment. The real-time traffic information includes at least one road element.

[0126] Real-time traffic information typically includes dynamic road elements, such as obstacles and / or road surface conditions. Optionally, real-time traffic information may include at least one of the following road elements: drivable area, vehicles, roadblocks, pedestrians, guardrails, flooded areas, muddy areas, and potholes. Vehicles, roadblocks, pedestrians, and guardrails are all considered obstacles, while flooded areas, muddy areas, and potholes are considered road surface conditions.

[0127] Optionally, real-time traffic information can be provided by Figure 2The sensor system 102 in the vehicle 100 shown identifies the target area. The sensor system 102 can first acquire an image of the target area, and then process and analyze the image of the target area to obtain real-time traffic information of the target area.

[0128] For example, sensor system 102 can acquire images of the target area via radar 1023 or camera 1025, and then process and analyze the images of the target area via computer vision system 1026 to obtain real-time traffic information of the target area. Computer vision system 1026 can obtain real-time traffic information through at least one of the following processing and analysis methods: obstacle detection, lane line detection, drivable area segmentation, water accumulation area segmentation, silt area segmentation, and pothole area segmentation.

[0129] Furthermore, after obtaining real-time traffic information through the sensor system 102, the real-time traffic information can be preprocessed to obtain the correspondence between each road element and its coordinates. Optionally, the preprocessing may include depth estimation, coordinate transformation, and logical operations. For example, the coordinates can be coordinates in the coordinate system used by the map stored in the vehicle.

[0130] Optionally, some road elements in real-time traffic information have the same meaning. For example, vehicles, roadblocks, pedestrians, or guardrails in real-time traffic information refer to areas outside the drivable area in the target region (also known as non-drivable areas). Therefore, real-time traffic information may not include vehicles, roadblocks, pedestrians, or guardrails, as long as the drivable area does not include vehicles, roadblocks, pedestrians, or guardrails.

[0131] 203. Based on real-time traffic information and the target machine learning model, determine the vehicle's current stopping position in the target area.

[0132] A target machine learning model is a model obtained by training on data and using algorithms to discover the patterns and essence of the data. Target machine learning models can include at least one of the following: Vision Transformers, CNN, Graph Transformers, and GNN. Target machine learning models are used to determine docking locations (also known as location determination models), and their inputs can be implemented in various ways; these implementations are described below.

[0133] For the first implementation method, please refer to Figure 4 , Figure 4 This application provides a schematic diagram of a method for determining a parking location based on real-time traffic information and a target machine learning model. The method includes the following steps:

[0134] 2031a. Input real-time traffic information into the target machine learning model.

[0135] 2032a. Obtain the docking position of the target machine learning model output.

[0136] In this implementation, the target area includes multiple candidate locations. The target machine learning model is used to determine the stopping priority corresponding to each of the multiple candidate locations in the target area based on the input real-time traffic information, and outputs the candidate location with the highest stopping priority. This stopping priority can be represented by letters or Arabic numerals. Optionally, the multiple candidate locations can be all locations in the target area. The target machine learning model can be obtained through a supervised learning training process, which is described below.

[0137] Optionally, a training region can be determined based on the training location. The maximum distance between the training region and the training location is less than a distance threshold, and the training region includes multiple locations. The initial model is trained using the real-time traffic information of the training region (i.e., the real-time traffic information corresponding to multiple locations in the training region) and the training labels corresponding to multiple locations in the training region to obtain the target machine learning model.

[0138] The method for obtaining real-time traffic information in the training area can refer to the aforementioned process 202, which will not be repeated here in this embodiment. It should be noted that the aforementioned processes 201 to 203 are executed during the autonomous driving process of the vehicle. During the model training process, the process of the vehicle automatically driving to the training location can be simulated to obtain real-time traffic information corresponding to multiple locations.

[0139] The process of obtaining training labels includes: acquiring label parameters corresponding to multiple locations through simulation, and determining the training label corresponding to the location based on the label parameters. The label parameters corresponding to the location include at least one of the following: the time taken to walk from the training location to the location (represented as Time), the delay of traffic flow at the location when a vehicle stops at the location (represented as Delay), the stopping status at the location (represented as Mood), and the probability of an accident occurring at the location (represented as Risk).

[0140] For example, taking any one of multiple locations as an example, the time taken to walk from the training location to the location (as per the training label) can be determined by simulating a scenario where a user walks from the training location to that location. Similarly, the delay in traffic flow at a location when a vehicle is parked (as per the training label) can be determined by simulating a scenario where other vehicles continue to move while a vehicle is parked at that location.

[0141] The parking status of the location in the training labels can be determined by the operator. The parking status can include at least one of the following: unparkable, parkable, or parkable but with poor road conditions (e.g., the candidate location is in a flooded area, a muddy area, or a pothole area). For example, the parking status value can include at least one of the following: inf, 0, and greater than 0 and less than or equal to 1, with each value corresponding to one of the aforementioned parking statuses. When the parking status value is inf, it indicates that the candidate location is unparkable. When the parking status value is 0, it indicates that the candidate location is parkable. When the parking status value is greater than 0 and less than or equal to 1, it indicates that the candidate location is parkable but with poor road conditions. The closer the parking status value is to 0, the better the road conditions at the candidate location; the closer the parking status value is to 1, the worse the road conditions at the candidate location. This parking status and its value are merely illustrative examples, and this application embodiment does not limit the scope of the application. The probability of an accident occurring at a location in the training labels can be determined by simulating scenarios where a user gets on or off the vehicle at any of those locations.

[0142] For example, the training label corresponding to a position can be calculated based on the label parameters corresponding to that position. The calculation method for the training label can include: score2 = n1T + n2D + n3M + n4R + n5, where n1 to n5 are user-defined parameters. T represents Time, D represents Delay, M represents Mood, and R represents Risk.

[0143] The model training process can be performed by the vehicle or by a separate model training device. In the latter case, after the model training device has trained the target machine learning model, it needs to send the target machine learning model to the vehicle.

[0144] Different types of models have different input data formats. Therefore, when the format of real-time traffic information differs from the input data format of the target machine learning model, the real-time traffic information needs to be encoded to match the input format requirements of the target machine learning model. The encoded data is then input into the target machine learning model to obtain the parking location output by the model.

[0145] In the second implementation, the target machine learning model requires input in the form of feature maps, such as VisionTransformers and CNN models. In this case, real-time traffic information data needs to be encoded into feature maps. Please refer to [reference needed]. Figure 5 , Figure 5 This application provides an embodiment of another method for determining a parking location based on real-time traffic information and a target machine learning model. The method includes the following steps:

[0146] 2031b. Obtain at least one feature map based on real-time traffic information.

[0147] One feature map is used to indicate at least one road element in the real-time traffic information. For example, a raster map corresponding to the target area can be determined from a bird's-eye view (BEV) perspective. Then, based on at least one road element in the real-time traffic information, the position of the center point of each raster is determined to correspond to at least one road element, resulting in a feature map indicating that at least one road element.

[0148] 2032b. Input at least one feature map into the target machine learning model to obtain the docking position output by the target machine learning model.

[0149] In this implementation, the target region includes multiple candidate locations. The target machine learning model is used to determine the docking priority of each candidate location in the target region based on at least one feature map as input, and outputs the candidate location with the highest docking priority. The target machine learning model can be obtained through a supervised learning training process, which is described below.

[0150] Optionally, a training area can be determined based on the training location, and real-time traffic information of the training area can be obtained. Then, at least one feature map is obtained based on the real-time traffic information of the training area, and the initial model is trained using the at least one feature map and training labels corresponding to multiple locations in the training area to obtain the target machine learning model. The method for obtaining the real-time traffic information of the training area can refer to the aforementioned process 202, the process for obtaining at least one feature map based on the real-time traffic information of the training area can refer to the aforementioned process 2031b, and the process for obtaining training labels and the training process can both refer to the aforementioned process 2031a. These details are not elaborated upon here in the embodiments of this application.

[0151] In the third implementation, the input format of the target machine learning model is required to be a dataset, such as a GraphTransformers model or a GNN model. In this case, the real-time traffic information data needs to be encoded into a dataset. The dataset can include non-Euclidean data formats, which refer to discretely distributed data formats.

[0152] Please refer to Figure 6 , Figure 6 A schematic diagram illustrating another method for determining a parking location based on real-time traffic information and a target machine learning model, provided in an embodiment of this application, is shown. This method includes the following steps:

[0153] 2031c. Determine the initial candidate location set corresponding to each road element in the real-time traffic information. The initial candidate location set includes at least one candidate location among multiple candidate locations in the target area.

[0154] For example, for each road element in real-time traffic information, its corresponding initial candidate location set can be the key locations within that road element. For instance, for the road element of a roadblock, the center location of each roadblock can be determined as the initial candidate location set for that roadblock. For the road element of a flooded area, the center location of each flooded area can be determined as the initial candidate location set for that flooded area.

[0155] 2032c. Merge the initial candidate location sets corresponding to all road elements in the real-time traffic information to obtain the candidate location set.

[0156] 2033c. Based on the road elements corresponding to the candidate locations in the candidate location set in the real-time traffic information, obtain the data sets corresponding to the candidate locations in the candidate location set. The data set corresponding to any candidate location in the candidate location set is used to indicate the road element corresponding to any candidate location in the real-time traffic information.

[0157] The data set corresponding to any candidate location in the candidate location set can be represented as {U, V, F1, F2, F3, F4, F5, F5, F6, F7, F8}, where U and V represent the coordinates of any candidate location, and F1 to F8 represent at least one road element corresponding to any candidate location.

[0158] 2034c. Input the data sets corresponding to the candidate positions in the candidate position set into the target machine learning model to obtain the docking position output by the target machine learning model.

[0159] In this implementation, the target area includes multiple candidate locations. The target machine learning model is used to determine the docking priority of each candidate location in the input candidate location set based on the corresponding datasets, and outputs the candidate location with the highest docking priority. The target machine learning model can be obtained through a supervised learning training process, which is described below.

[0160] Optionally, a training area can be determined based on the training location, and real-time traffic information of the training area can be obtained. Then, an initial location set corresponding to each road element in the real-time traffic information of the training area is determined. The initial location set includes at least one of multiple locations in the training area. The initial location sets corresponding to all road elements in the real-time traffic information of the training area are merged to obtain a location set. Based on the road elements corresponding to the locations in the location set, a data set corresponding to each location in the location set is obtained. The data set corresponding to any location in the location set is used to indicate the road element corresponding to any location in the real-time traffic information of the training area. Finally, the initial model is trained using the data set corresponding to each location in the location set and training labels to obtain the target machine learning model. The method for obtaining the real-time traffic information of the training area can refer to the aforementioned process 202. The process of obtaining the data set corresponding to each location in the location set based on the real-time traffic information of the training area can refer to the aforementioned processes 2031c to 2033c. The process of obtaining training labels and the training process can both refer to the aforementioned process 2031a. These details are not elaborated upon here in this embodiment.

[0161] It should be noted that, in the three implementation methods mentioned above, real-time traffic information typically includes a wide variety of road elements, resulting in a complex amount of information. The target machine learning model can process these diverse road elements, thereby accurately and efficiently determining the optimal stopping location from multiple candidate locations. Specifically, during training, the target machine learning model can adaptively determine the importance of various road elements and the redundancy among them based on the training labels. This allows the trained target machine learning model to accurately determine the importance and redundancy of various road elements corresponding to each candidate location, and thus determine and output the optimal stopping location from multiple candidate locations based on the importance and redundancy of these road elements.

[0162] In the fourth implementation, the target area includes multiple candidate locations; please refer to [reference needed]. Figure 7 , Figure 7 A schematic diagram illustrating another method for determining a parking location based on real-time traffic information and a target machine learning model, provided in this application embodiment, is shown. This method includes the following steps:

[0163] 2031d. Based on real-time traffic information and destination location, determine one or more feature data corresponding to each candidate location among multiple candidate locations.

[0164] The one or more feature data includes at least one of the following feature data: parking availability feature, distance feature, traffic obstruction feature, and traffic flow feature. Wherein, the parking availability feature indicates the parking status of the candidate location, the distance feature indicates the distance between the destination location and the candidate location, the traffic obstruction feature indicates the degree to which a vehicle obstructs traffic flow when parked at the candidate location, and the traffic flow feature indicates the traffic flow at the candidate location.

[0165] The dockability feature can be represented as F1. The meaning of F1 can be referred to the docking status of the location in the training labels. This application will not elaborate on the specific implementation here.

[0166] The following explanation uses real-time traffic information, including drivable areas, vehicles, roadblocks, pedestrians, guardrails, flooded areas, muddy areas, and potholes, as examples. When a vehicle determines that a candidate location belongs to a non-drivable area based on real-time traffic information, the F1 value is set to inf. When a candidate location is determined to be a drivable area and not a flooded, muddy, or pothole area, the F1 value is set to 0. When a candidate location is determined to be a drivable area and belongs to at least one of a flooded, muddy, or pothole area, the F1 value is set to a value greater than 0 and less than or equal to 1.

[0167] Distance features can be represented by two distances, F2 and F3. F2 represents the distance between the target location and the candidate location in the first direction, and F3 represents the distance between the target location and the candidate location in the second direction, with an angle between the first and second directions. For example, the first direction could be the direction along the length of the vehicle, and the second direction could be the direction along the width of the vehicle.

[0168] The values ​​of F2 and F3 can be obtained from the coordinates of the target position and the candidate position. The value of F2 is the absolute value of the difference between the x-coordinate of the target position and the x-coordinate of the candidate position, and the value of F3 is the absolute value of the difference between the y-coordinate of the target position and the y-coordinate of the candidate position.

[0169] For example, please refer to Figure 8 , Figure 8 This is a schematic diagram illustrating a scenario where a vehicle is parked at a candidate location, as provided in an embodiment of this application. Figure 8 The diagram shows vehicle 301, destination location 302, road 303, first direction a1, and second direction a2. The distance between destination location 302 and vehicle 301 in the first direction a1 is b, and the distance in the second direction is c. The value of F2 is b, and the value of F3 is c.

[0170] Traffic obstruction features can be represented as F4, which can be obtained from the width of the road where the candidate location is located and the width of the vehicle. For example, F4 can be the ratio of the vehicle width to the width of the road where the candidate location is located. Vehicles can determine the road where the candidate location is located based on its coordinates, and thus obtain the width of the road where the candidate location is located. For example, as described above... Figure 8 As shown, the width of vehicle 301 is d, the width of road 303 where the candidate position is located is e, and the value of F4 is d / e.

[0171] Traffic flow characteristics can be represented as F5, which can be obtained from the coordinates of candidate locations. For example, a vehicle can determine the traffic flow corresponding to the coordinates of a candidate location based on the traffic flow, and thus obtain the value of F5.

[0172] 2032d. Input one or more feature data corresponding to multiple candidate positions into the target machine learning model to obtain the docking position output by the target machine learning model.

[0173] In this implementation, the target machine learning model is used to determine the docking priority of each of the multiple candidate locations based on one or more feature data corresponding to each of the input candidate locations, and outputs the candidate location with the highest docking priority. The target machine learning model can be obtained through a supervised learning training process, which is described below.

[0174] Optionally, a training area can be determined based on the training location, and real-time traffic information of the training area can be obtained. Then, based on the real-time traffic information of the training area and the target location, one or more feature data corresponding to each location in the training area are determined. Finally, the initial model is trained using the one or more feature data corresponding to each location and training labels to obtain the target machine learning model. The method for obtaining the real-time traffic information of the training area can refer to the aforementioned process 202; the process for obtaining one or more feature data corresponding to each location in the training area based on the real-time traffic information can refer to the aforementioned process 2031d; and the process for obtaining training labels and the training process can both refer to the aforementioned process 2031a. These details are not elaborated upon here in the embodiments of this application.

[0175] When the format of one or more feature data corresponding to multiple locations in the training region differs from the format of the input data of the target machine learning model, it is also necessary to encode the one or more feature data corresponding to multiple locations in the training region according to the type of the initial model. This process can be referred to the aforementioned... Figure 5 or Figure 6 The process is not described in detail in the embodiments of this application.

[0176] In this embodiment, processes 202 and 203 can be executed cyclically. Since real-time traffic information is constantly changing during automatic vehicle operation, cyclically executing processes 202 and 203 allows for the determination of the optimal stopping position based on the changing real-time traffic information. For example, the vehicle can re-execute processes 202 and 203 at intervals t, where t can be 0.25 seconds, 0.3 seconds, or 0.5 seconds, etc. This embodiment does not limit the value of t.

[0177] Once a parking location is determined, the vehicle can park accordingly. For example, it can be parked by... Figure 2 The control system 103 in the vehicle 100 shown controls the vehicle 100 to park according to the determined parking position. Afterwards, the user can be notified to get on or off the vehicle. For scenarios requiring notification to the user to get on, a boarding notification can be sent to the user via an application that sends user commands.

[0178] In summary, the vehicle parking location determination method provided in this application determines a target area based on the vehicle's destination location, then obtains real-time traffic information for the target area at the current moment. This real-time traffic information includes at least one road element. Finally, based on the real-time traffic information and a target machine learning model, the method determines the vehicle's parking location within the target area at the current moment. By determining the parking location based on real-time traffic information and the target machine learning model, the method considers changes in road conditions during the parking location determination process, enabling the vehicle to park at the determined location and making it more convenient and safer for users to get on and off the vehicle at the determined parking location.

[0179] Furthermore, real-time traffic information typically includes a wide variety of road elements, resulting in complex information. Objective machine learning models can process these diverse road elements to accurately and efficiently determine the optimal stopping location from multiple candidate locations. Specifically, during training, the objective machine learning model adaptively acquires the importance of various road elements and the redundancy among them based on the training labels. This allows the trained objective machine learning model to accurately determine the importance and redundancy of various road elements corresponding to each candidate location, thereby determining and outputting the optimal stopping location from multiple candidate locations based on these factors.

[0180] In addition, real-time traffic information can include at least one of the following road elements: drivable area, vehicles, roadblocks, pedestrians, guardrails, flooded areas, muddy areas, and potholes. This reduces the probability of poor road conditions at a given parking location (e.g., in a flooded, muddy, or pothole area), making it more convenient and safer for users to get on and off the vehicle, thus improving their overall experience.

[0181] The order of the methods provided in the embodiments of this application can be adjusted appropriately, and the process can also be added or removed as appropriate. Any variations that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the protection scope of this application, and the embodiments of this application do not limit this.

[0182] Example 2: Determine the vehicle's parking location based on real-time traffic information and map information.

[0183] Please refer to Figure 9 , Figure 9 A flowchart illustrating another method for determining the parking location of a vehicle provided in this application embodiment, the method may include the following steps:

[0184] 401. Determine the target area based on the vehicle's destination location, and the maximum distance between the target area and the destination location is less than the distance threshold.

[0185] This process 401 can refer to the aforementioned process 201, and will not be described again in the embodiments of this application.

[0186] 402. Obtain map information of the target area at the current time. The map information includes at least one road element.

[0187] Optionally, map information can be obtained from high-definition maps. High-definition maps typically represent road elements using at least one map layer, with each map layer representing one type of road element. For example, map information may include at least one of the following road elements: lane lines, sidewalks, curbs, road gradients, speed limits, traffic flow, and drivable areas.

[0188] Map information can be pre-stored in the vehicle, for example, it can be stored in... Figure 2 The map information is stored in memory 1072 within the computer system 107 of the vehicle 100 shown. Memory 1072 transmits the map information of the target area at the current time to the vehicle 100 via the vehicle communication network. Alternatively, the map information can be stored in the cloud, and the vehicle retrieves the map information from the cloud; this embodiment does not limit this approach.

[0189] 403. Obtain real-time traffic information for the target area at the current moment. The real-time traffic information includes at least one road element.

[0190] The process described in step 403 can be referred to in step 202 above, and will not be repeated here in this embodiment. After obtaining real-time traffic information, the vehicle needs to store the real-time traffic information in order to determine the parking location later. The vehicle can use a map to store the real-time traffic information or store the real-time traffic information separately. This embodiment does not limit the storage method of the real-time traffic information.

[0191] When using maps to store real-time traffic information, the map information can be updated using the real-time traffic information to obtain updated map information. The updated map information includes road elements that existed before the update and road elements from the real-time traffic information.

[0192] For any road element in real-time traffic information, if the road element already exists in the map information, the road element in the map information is updated to match the road element in the real-time traffic information. If the road element does not exist in the map information, it is added to the map information. Taking a high-definition map as an example, for any road element in real-time traffic information, if a map layer representing the road element exists in the high-definition map, the map layer is updated based on the road element. If a map layer representing the road element does not exist in the high-definition map, a new map layer is added to the high-definition map, and the representation rules for the new map layer are defined. Then, the added map layer is updated based on the road element.

[0193] Some road elements in real-time traffic information and map information have the same meaning. Vehicles can represent road elements with the same meaning using different map layers or the same map layer. This application does not limit this.

[0194] Utilizing maps to store real-time traffic information is advantageous, as map data structures are well-suited for this purpose. Furthermore, updated map information can be used for subsequent autonomous driving and automatic parking (e.g., route planning or determining the next stopping location), thus improving the utilization rate of real-time traffic information.

[0195] 404. Based on real-time traffic information, map information, and the target machine learning model, determine the vehicle's current stopping position in the target area.

[0196] In process 404, if real-time traffic information is stored in the map in process 403, the stopping location can be determined based on the updated map information and the target machine learning model.

[0197] There are multiple ways to implement the input of a target machine learning model. The following describes each of these methods.

[0198] In the first implementation, real-time traffic information and map information can be input into the target machine learning model to obtain the parking location output by the model. The target machine learning model is used to determine the parking priority of multiple candidate locations in the target area based on the input real-time traffic information and map information, and outputs the candidate location with the highest parking priority. The target machine learning model can be obtained through a supervised learning training process, which is described below.

[0199] Optionally, a training area can be determined based on the training location, and real-time traffic information and map information of the training area can be obtained. Then, the initial model is trained using the real-time traffic information, map information, and training labels corresponding to multiple locations within the training area to obtain the target machine learning model. The method for obtaining the real-time traffic information of the training area can refer to the aforementioned processes 202 and 403, the method for obtaining the map information of the training area can refer to the aforementioned process 402, and the process for obtaining the training labels and the training process can both refer to the aforementioned process 2031a. These details are not elaborated upon here in the embodiments of this application.

[0200] In the second implementation, the input format for the target machine learning model is a feature map. At least one feature map can be obtained based on real-time traffic information and map information. This feature map is then input into the target machine learning model to obtain the output parking location. A feature map indicates at least one road element from the real-time traffic information and map information. The process of obtaining at least one feature map from the real-time traffic information and map information can be referred to in the aforementioned process 2031b, and will not be elaborated upon here.

[0201] In this implementation, the role of the target machine learning model can be referred to in the aforementioned process 2032b, and will not be elaborated further in this embodiment. The target machine learning model can be obtained through a supervised learning training process, and the training process of the target machine learning model is described below. Optionally, a training area can be determined based on the training location, and real-time traffic information and map information of the training area can be obtained. Then, at least one feature map is obtained based on the real-time traffic information and map information of the training area, and the initial model is trained using at least one feature map and training labels corresponding to multiple locations in the training area to obtain the target machine learning model. The method for obtaining the real-time traffic information of the training area can refer to the aforementioned processes 202 and 403, the method for obtaining the map information of the training area can refer to the aforementioned process 402, the process for obtaining at least one feature map based on the real-time traffic information of the training area can refer to the aforementioned process 2031b, and the process for obtaining training labels and the training process can both refer to the aforementioned process 2031a, and will not be elaborated further in this embodiment.

[0202] In the third implementation, the input format of the target machine learning model is required to be a dataset, which can determine the initial candidate location sets corresponding to road elements in the real-time traffic information and map information, respectively. The initial candidate location sets corresponding to all road elements in the real-time traffic information and map information are merged to obtain a candidate location set. Based on the road elements corresponding to the candidate locations in the candidate location sets in the real-time traffic information and map information, a dataset corresponding to each candidate location in the candidate location set is obtained. The dataset corresponding to any candidate location in the candidate location set is used to indicate the road element corresponding to any candidate location in the real-time traffic information and map information. Then, the dataset corresponding to each candidate location in the candidate location set is input into the target machine learning model to obtain the parking location output by the target machine learning model. The process of obtaining the dataset corresponding to each candidate location in the candidate location set based on the real-time traffic information and map information can be referred to the aforementioned processes 2031c to 2033c, and will not be elaborated upon here in the embodiments of this application.

[0203] In this implementation, the role of the target machine learning model can be referred to in the aforementioned process 2034c, and will not be repeated here in the embodiments of this application. The target machine learning model can be obtained through a supervised learning training process, and the training process of the target machine learning model is described below.

[0204] Optionally, a training area can be determined based on the training location, and real-time traffic information and map information of the training area can be obtained. Then, the initial location sets corresponding to the road elements in the real-time traffic information and map information of the training area are determined, and the initial location sets corresponding to all road elements in the real-time traffic information and map information of the training area are merged to obtain a location set. Based on the road elements corresponding to the locations in the location set in the real-time traffic information and map information of the training area, a data set corresponding to each location in the location set is obtained. Finally, the initial model is trained using the data set corresponding to each location in the location set and the training labels to obtain the target machine learning model. The method for obtaining the real-time traffic information of the training area can refer to the aforementioned processes 202 and 403; the method for obtaining the map information of the training area can refer to the aforementioned process 402; the process for obtaining the data set corresponding to each location in the location set based on the real-time traffic information and map information of the training area can refer to the aforementioned process 2031b; the process for obtaining the training labels and the training process can both refer to the aforementioned process 2031a. The embodiments of this application will not be elaborated upon here.

[0205] In the fourth implementation, one or more feature data corresponding to each of the multiple candidate locations can be determined based on real-time traffic information, map information, and the destination location. Then, the one or more feature data corresponding to each of the multiple candidate locations are input into the target machine learning model to obtain the parking location output by the target machine learning model. The process of obtaining the one or more feature data and determining the one or more feature data corresponding to each of the multiple candidate locations based on real-time traffic information, map information, and the destination location can refer to the aforementioned process 2031d, and will not be elaborated upon here. Map information may include: drivable areas and traffic flow. It should be noted that, in this case, the F1 value can be obtained based on real-time traffic information and map information, and the F5 value can be obtained through the coordinates of the candidate location and map information. For example, the vehicle can determine the traffic flow corresponding to the coordinates of the candidate location based on the traffic flow in the map information, and then obtain the F5 value.

[0206] In this implementation, the role of the target machine learning model can be referred to in the aforementioned process 2032d, and will not be repeated here in the embodiments of this application. The target machine learning model can be obtained through a supervised learning training process, and the training process of the target machine learning model is described below.

[0207] Optionally, a training area can be determined based on the training location, and real-time traffic information and map information of the training area can be obtained. Then, based on the real-time traffic information, map information, and destination location of the training area, one or more feature data corresponding to each location within the training area are determined. Finally, the initial model is trained using the one or more feature data corresponding to each location and training labels to obtain the target machine learning model. The method for obtaining the real-time traffic information of the training area can refer to the aforementioned processes 202 and 403; the method for obtaining the map information of the training area can refer to the aforementioned process 402; the process for obtaining one or more feature data corresponding to each location within the training area based on the real-time traffic information can refer to the aforementioned process 2031d; and the process for obtaining training labels and the training process can both refer to the aforementioned process 2031a. These details are not elaborated upon here in the embodiments of this application.

[0208] In this embodiment of the application, the aforementioned processes 402 to 404 can be executed cyclically. For example, the vehicle can re-execute the aforementioned processes 402 to 404 every interval t.

[0209] In summary, the vehicle parking location determination method provided in this application determines a target area based on the vehicle's destination location, then acquires real-time traffic information and map information of the target area at the current moment. The real-time traffic information includes at least one road element. Finally, based on the real-time traffic information, map information, and a target machine learning model, the vehicle's parking location within the target area at the current moment is determined. By determining the parking location based on real-time traffic information, map information, and the target machine learning model, the method considers changes in traffic conditions and map information during the parking location determination process, enabling vehicles to park at the determined parking locations and making it more convenient and safer for users to get on and off the vehicle at the determined parking locations.

[0210] Furthermore, real-time traffic and map information typically includes a wide variety of road elements, resulting in complex information. Objective machine learning models can process these diverse road elements to accurately and efficiently determine the optimal stopping location from multiple candidate locations. Specifically, during training, the objective machine learning model adaptively acquires the importance of various road elements and the redundancy among them based on the training labels. This allows the trained objective machine learning model to accurately determine the importance and redundancy of various road elements corresponding to each candidate location, thereby determining and outputting the optimal stopping location from multiple candidate locations based on these factors.

[0211] In addition, real-time traffic information can include at least one of the following road elements: drivable area, vehicles, roadblocks, pedestrians, guardrails, flooded areas, muddy areas, and potholes. This reduces the probability of poor road conditions at a given parking location (e.g., in a flooded, muddy, or pothole area), making it more convenient and safer for users to get on and off the vehicle, thus improving their overall experience.

[0212] The order of the methods provided in this application can be adjusted appropriately, and the processes can be added or removed as needed. For example, process 402 can be omitted, and the parking location can be determined directly based on real-time traffic information. Alternatively, the order of processes 402 and 403 can be reversed. Any variations that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the protection scope of this application. This application does not limit these variations.

[0213] The foregoing has described the method for determining the parking position of a vehicle according to embodiments of this application. It is understood that, in order to achieve the above functions, the vehicle includes corresponding hardware structures and / or software modules for performing each function. Those skilled in the art should readily recognize that, based on the algorithm steps of the examples described in conjunction with the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0214] This application embodiment can divide the vehicle into functional modules according to the above method example. For example, each function can be divided into its own functional module, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware or as a software functional module. It should be noted that the module division in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0215] For example, the vehicle parking location determination device may be the vehicle itself, or it may be a chip or other combination device or component that has the functions of the vehicle parking location determination device. Figure 10 This is a block diagram of a vehicle parking position determination device provided in an embodiment of this application. When each functional module is divided according to its corresponding function, the vehicle parking position determination device 500 includes:

[0216] The first determining module 501 is used to determine a target area based on the destination location of the vehicle, wherein the maximum distance between the target area and the destination location is less than a distance threshold.

[0217] The first acquisition module 502 is used to acquire real-time traffic information of the target area at the current moment, wherein the real-time traffic information includes at least one road element.

[0218] The second determining module 503 is used to determine the parking position of the vehicle in the target area at the current moment based on the real-time traffic information and the target machine learning model.

[0219] In conjunction with the above scheme, the real-time traffic information includes at least one of the following road elements: drivable area, vehicles, roadblocks, pedestrians, guardrails, waterlogged areas, muddy areas, and potholes.

[0220] Please refer to the above solutions. Figure 11 , Figure 11A block diagram of another vehicle parking position determination device provided in this application embodiment, in Figure 10 Based on this, the vehicle parking position determination device 500 also includes:

[0221] The second acquisition module 504 is used to acquire map information of the target area at the current time, the map information including at least one road element;

[0222] Please refer to Figure 12 , Figure 12 A block diagram of a second determining module 503 provided in an embodiment of this application, the second determining module 503 including:

[0223] The determining unit 5031 is used to determine the parking location based on the real-time traffic information, the map information, and the target machine learning model.

[0224] Please refer to the above solutions. Figure 13 , Figure 13 A block diagram of another vehicle parking position determination device provided in this application embodiment, in Figure 11 Based on this, the vehicle parking position determination device 500 also includes:

[0225] Update module 505 is used to update the map information using the real-time traffic information to obtain updated map information, wherein the updated map information includes road elements in the real-time traffic information;

[0226] The determining unit 5031 is specifically used to determine the docking location based on the updated map information and the target machine learning model.

[0227] In conjunction with the above scheme, the determining unit 5031 is specifically used to input the real-time traffic information and the map information into the target machine learning model to obtain the parking location output by the target machine learning model.

[0228] In conjunction with the above scheme, the determining unit 5031 is specifically used to obtain at least one feature map based on the real-time traffic information and the map information, wherein the feature map is used to indicate at least one road element in the real-time traffic information and the map information; and input the at least one feature map into the target machine learning model to obtain the parking position output by the target machine learning model.

[0229] In conjunction with the above scheme, the target area includes multiple candidate locations. The determining unit 5031 is specifically used to determine the initial candidate location sets corresponding to the road elements in the real-time traffic information and the map information, respectively. The initial candidate location sets include at least one candidate location among the multiple candidate locations. The initial candidate location sets corresponding to all road elements in the real-time traffic information and the map information are merged to obtain a candidate location set. Based on the road elements corresponding to the candidate locations in the candidate location sets in the real-time traffic information and the map information, a data set corresponding to each candidate location in the candidate location set is obtained. The data set corresponding to each candidate location in the candidate location set is input into the target machine learning model to obtain the parking location output by the target machine learning model.

[0230] In conjunction with the above scheme, the target area includes multiple candidate locations. The determining unit 5031 is specifically used to determine one or more feature data corresponding to each candidate location among the multiple candidate locations based on the real-time traffic information, the map information, and the destination location. The one or more feature data includes at least one of the following feature data: parking availability feature, distance feature, traffic obstruction feature, and traffic flow feature. The one or more feature data corresponding to each of the multiple candidate locations are input into the target machine learning model to obtain the parking location output by the target machine learning model. Wherein, the parking availability feature represents the parking status of the candidate location, the distance feature represents the distance between the destination location and the candidate location, the traffic obstruction feature represents the degree of obstruction of traffic flow when the vehicle is parked at the candidate location, and the traffic flow feature represents the traffic flow of the candidate location.

[0231] Please refer to the above solutions. Figure 14 , Figure 14 A block diagram of another vehicle parking position determination device provided in this application embodiment, in Figure 10 Based on this, the vehicle parking position determination device 500 also includes:

[0232] The loop module 506 is used to repeatedly execute the parking position determination process until the vehicle stops moving;

[0233] The process of determining the docking location includes:

[0234] Obtain real-time traffic information for the target area at the current moment;

[0235] Based on the real-time traffic information and the target machine learning model, the vehicle's stopping position in the target area at the current moment is determined.

[0236] Figure 15This is a schematic diagram of a vehicle parking position determination device provided in an embodiment of this application. The vehicle parking position determination device 600 can be a vehicle or a chip or functional module within a vehicle. Figure 15 As shown, the vehicle parking location determination device 600 includes a processor 601, a transceiver 602, and a communication line 603.

[0237] Among them, processor 601 is used to perform such as Figure 3 and Figure 9 In any step of the method embodiment shown, when performing data transmission such as receiving the destination location, the transceiver 602 and communication line 603 may be invoked to complete the corresponding operation.

[0238] Furthermore, the vehicle parking location determination device 600 may also include a memory 604. The processor 601, memory 604, and transceiver 602 can be connected via a communication line 603.

[0239] The processor 601 can be a central processing unit (CPU), a network processor (NP), a digital signal processor (DSP), a microprocessor, a microcontroller, a programmable logic device (PLD), or any combination thereof. The processor 601 can also be other devices with processing capabilities, such as circuits, devices, or software modules, without limitation.

[0240] Transceiver 602 is used to communicate with other devices or other communication networks, such as Ethernet, radio access network (RAN), wireless local area network (WLAN), etc. Transceiver 602 can be a module, circuit, transceiver, or any device capable of enabling communication.

[0241] The transceiver 602 is mainly used for transmitting and receiving signals / data, and may include a transmitter and a receiver to send and receive signals / data respectively; operations other than signal / data transmission and reception are implemented by the processor, such as information processing and calculation.

[0242] Communication line 603 is used to transmit information between the components included in the vehicle parking position determination device 600.

[0243] In one design, the processor can be viewed as a logic circuit, and the transceiver as an interface circuit.

[0244] Memory 604 is used to store instructions. These instructions can be computer programs.

[0245] The memory 604 can be volatile memory or non-volatile memory, or it can include both. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DRRAM). Memory 604 can also be a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed discs, laser discs, optical discs, digital universal discs, Blu-ray discs, etc.), magnetic disk storage media, or other magnetic storage devices. It should be noted that the memory in the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0246] It should be noted that the memory 604 can exist independently of the processor 601, or it can be integrated with the processor 601. The memory 604 can be used to store instructions, program code, or some data, etc. The memory 604 can be located inside or outside the vehicle parking position determining device 600, without limitation. The processor 601 is used to execute the instructions stored in the memory 604 to implement the method provided in the above embodiments of this application.

[0247] In one example, processor 601 may include one or more CPUs, for example Figure 15CPU0 and CPU1 in the CPU.

[0248] As an optional implementation, the vehicle parking location determination device 600 includes multiple processors, for example, in addition to Figure 15 In addition to processor 601, it may also include processor 607.

[0249] As an optional implementation, the vehicle parking location determination device 600 also includes an output device 605 and an input device 606. For example, the input device 606 is a device such as a keyboard, mouse, microphone, or joystick, and the output device 605 is a device such as a display screen or speaker.

[0250] It should be noted that the vehicle parking location determination device 600 can be a terminal device, network device, relay device, embedded device, chip system, or other similar device. Figure 15 Equipment with a similar structure. Furthermore... Figure 15 The components shown do not constitute a limitation on the vehicle parking position determination device 600, except... Figure 15 In addition to the components shown, the vehicle parking position determination device 600 may include a ratio Figure 15 This may indicate more or fewer components, or combinations of certain components, or different component arrangements.

[0251] The processor and transceiver described in this application can be implemented on integrated circuits (ICs), analog ICs, radio frequency integrated circuits, mixed-signal ICs, application-specific integrated circuits (ASICs), printed circuit boards (PCBs), electronic devices, etc. The processor and transceiver can also be manufactured using various IC process technologies, such as complementary metal-oxide semiconductors (CMOS), n-metal-oxide-semiconductor (NMOS), positive-channel metal-oxide semiconductors (PMOS), bipolar junction transistors (BJTs), bipolar CMOS (BiCMOS), silicon germanium (SiGe), gallium arsenide (GaAs), etc.

[0252] It should be noted that the vehicle parking location determination device 600 can be any type of handheld or fixed device, such as a laptop or tablet computer, mobile phone, smartphone, tablet or tablet computer, camera, desktop computer, set-top box, television, display device, digital media player, video game console, video streaming device (e.g., content service server or content distribution server), broadcast receiving device, broadcast transmitting device, and monitoring equipment, etc., and may or may not use any type of operating system. The vehicle parking location determination device 600 can also be a device in a cloud computing scenario, such as a virtual machine in a cloud computing scenario. In some cases, the vehicle parking location determination device 600 may be equipped with components for wireless communication. Therefore, the vehicle parking location determination device 600 can be a wireless communication device. Figure 15 Equipment with a similar structure. Furthermore... Figure 15 The components shown do not constitute a limitation on the vehicle parking position determination device 600, except... Figure 15 In addition to the components shown, the vehicle parking position determination device 600 may include a ratio Figure 15 This may indicate more or fewer components, or combinations of certain components, or different component arrangements.

[0253] The vehicle parking position determination device 600 can also be a chip system, which can be composed of chips or include chips and other discrete components. Furthermore, the actions, terminology, etc., involved in the various embodiments of this application can be referenced interchangeably without limitation.

[0254] Figure 16 This is a schematic diagram of another vehicle parking position determination device provided in an embodiment of this application. This vehicle parking position determination device is applicable to the scenarios shown in the above method embodiments. For ease of explanation, Figure 16 Only the main components of the vehicle parking position determination device are shown, including a processor, memory, control circuitry, and input / output devices. The processor is primarily used to process communication protocols and data, execute software programs, and process the data within those programs. The memory is mainly used to store the software programs and data. The control circuitry is primarily used for power supply and the transmission of various electrical signals. The input / output devices are primarily used to receive user input data and output data to the user.

[0255] When the vehicle's parking position determination device is a vehicle, the control circuit can be a motherboard. The memory includes storage media such as hard drives, RAM, and ROM. The processor can include a baseband processor and a central processing unit (CPU). The baseband processor is mainly used for processing communication protocols and data, while the CPU is mainly used for controlling the entire signal testing device, executing software programs, and processing software program data. Input / output devices include a display screen, keyboard, and mouse. The control circuit can further include or connect to transceiver circuits or transceivers, such as network cable interfaces, for sending or receiving data or signals, such as for data transmission and communication with other devices. Furthermore, it can also include an antenna for transmitting and receiving wireless signals for data / signal transmission with other devices.

[0256] According to the method provided in the embodiments of this application, this application also provides a computer program product, which includes computer program code. When the computer program code is run on a computer, it causes the computer to execute any of the methods described in the embodiments of this application.

[0257] This application also provides a computer-readable storage medium. All or part of the processes in the above method embodiments can be executed by a computer or an information processing device, using computer programs or instructions to control related hardware. The computer program or set of instructions can be stored in the computer-readable storage medium. When executed, the computer program or set of instructions can include the processes described in the above method embodiments. The computer-readable storage medium can be an internal storage unit of the vehicle in any of the foregoing embodiments, such as a vehicle's hard drive or memory. The computer-readable storage medium can also be an external storage device of the vehicle, such as a plug-in hard drive, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped in the vehicle. Further, the computer-readable storage medium can include both internal storage units and external storage devices of the vehicle. The computer-readable storage medium is used to store the computer program or instructions and other programs and data required by the vehicle. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0258] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0259] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0260] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0261] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0262] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0263] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0264] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for determining the parking location of a vehicle, characterized in that, The method includes: The target area is determined based on the vehicle's destination location, and the maximum distance between the target area and the destination location is less than a distance threshold. Obtain real-time traffic information for the target area at the current moment, wherein the real-time traffic information includes at least one road element; Based on the real-time traffic information and the target machine learning model, the vehicle's stopping position in the target area at the current moment is determined; The method further includes: Obtain map information of the target area at the current time, wherein the map information includes at least one road element; Determining the vehicle's current stopping position in the target area based on the real-time traffic information and the target machine learning model includes: The parking location is determined based on the real-time traffic information, the map information, and the target machine learning model; The target area includes multiple candidate locations. Determining the parking location based on the real-time traffic information, the map information, and the target machine learning model includes: Determine the initial candidate location sets corresponding to the road elements in the real-time traffic information and the map information, respectively, wherein the initial candidate location sets include at least one candidate location from the plurality of candidate locations; The initial candidate location sets corresponding to all road elements in the real-time traffic information and the map information are merged to obtain the candidate location set; Based on the road elements corresponding to the candidate locations in the candidate location set in the real-time traffic information and the map information, a data set corresponding to each candidate location in the candidate location set is obtained. The data set corresponding to any candidate location in the candidate location set is used to indicate the road element corresponding to any candidate location in the real-time traffic information and the map information. The data sets corresponding to the candidate locations in the candidate location set are input into the target machine learning model to obtain the docking location output by the target machine learning model.

2. The method according to claim 1, characterized in that, The real-time traffic information includes at least one of the following road elements: drivable area, vehicles, roadblocks, pedestrians, guardrails, waterlogged areas, muddy areas, and potholes.

3. The method according to claim 1, characterized in that, The method further includes: The map information is updated using the real-time traffic information to obtain updated map information, which includes road elements from the real-time traffic information. Determining the parking location based on the real-time traffic information, the map information, and the target machine learning model includes: The docking location is determined based on the updated map information and the target machine learning model.

4. The method according to any one of claims 1 to 3, characterized in that, The method further includes: The process of determining the parking location is repeated until the vehicle stops moving. The process of determining the docking location includes: Obtain real-time traffic information for the target area at the current moment; Based on the real-time traffic information and the target machine learning model, the vehicle's stopping position in the target area at the current moment is determined.

5. A vehicle parking position determination device, characterized in that, The device includes: The first determining module is used to determine a target area based on the vehicle's destination location, wherein the maximum distance between the target area and the destination location is less than a distance threshold. The first acquisition module is used to acquire real-time traffic information of the target area at the current moment, and the real-time traffic information includes at least one road element. The second determining module is used to determine the stopping position of the vehicle in the target area at the current moment based on the real-time traffic information and the target machine learning model; The device further includes: The second acquisition module is used to acquire map information of the target area at the current time, the map information including at least one road element; The target area includes multiple candidate locations. The second determining module includes: a determining unit, configured to determine initial candidate location sets corresponding to road elements in the real-time traffic information and the map information, wherein the initial candidate location sets include at least one candidate location among the multiple candidate locations; merge the initial candidate location sets corresponding to all road elements in the real-time traffic information and the map information to obtain a candidate location set; obtain a data set corresponding to each candidate location in the candidate location set based on the road elements corresponding to each candidate location in the candidate location set in the real-time traffic information and the map information; input the data set corresponding to each candidate location in the candidate location set into the target machine learning model to obtain the stop location output by the target machine learning model.

6. The apparatus according to claim 5, characterized in that, The real-time traffic information includes at least one of the following road elements: drivable area, vehicles, roadblocks, pedestrians, guardrails, waterlogged areas, muddy areas, and potholes.

7. The apparatus according to claim 5, characterized in that, The device further includes: The update module is used to update the map information using the real-time traffic information to obtain updated map information, wherein the updated map information includes road elements in the real-time traffic information; The determining unit is specifically used to determine the docking location based on the updated map information and the target machine learning model.

8. The apparatus according to any one of claims 5 to 7, characterized in that, The device further includes: The loop module is used to repeatedly execute the parking position determination process until the vehicle stops moving; The process of determining the docking location includes: Obtain real-time traffic information for the target area at the current moment; Based on the real-time traffic information and the target machine learning model, the vehicle's stopping position in the target area at the current moment is determined.

9. A vehicle parking position determination device, characterized in that, include: One or more processors; Memory, used to store one or more computer programs or instructions; When the one or more computer programs or instructions are executed by the one or more processors, the one or more processors perform the method as described in any one of claims 1 to 4.

10. A computer-readable storage medium, characterized in that, It includes a computer program or instructions that, when executed on a computer, cause the computer to perform the method of any one of claims 1 to 4.

11. A computer program product, characterized in that, The computer program product includes computer program code that, when run on a computer, causes the computer to perform the method of any one of claims 1 to 4.

12. A vehicle, characterized in that, The vehicle includes a control system and a vehicle parking position determination device as described in any one of claims 5 to 8; The vehicle parking position determination device is used to determine the parking position of the vehicle in the target area; The control system is used to control the parking of the vehicle according to the determined parking position.

Images