Vehicle positioning method and device, electronic equipment and storage medium

Abstract

The invention discloses a vehicle positioning method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a reference track in a scene map of a scene where a vehicle is located at present in response to a triggering condition; determining a first feature point of the reference trajectory based on pose attribute information of a reference trajectory point in the reference trajectory; determining an increment feature point based on the pose attribute information of an increment driving track point located in the current positioning window in the current driving track of the vehicle; matching the incremental feature points with the first feature points to obtain incremental feature point pairs; registering the reference track and the driving track based on the first feature point sequence and the second feature point sequence; and in response to matching of the registered reference trajectory and the driving trajectory, determining a current positioning result of the vehicle based on the reference trajectory point corresponding to the tail end driving trajectory point in the registered driving trajectory, so that effective positioning of the vehicle can be realized in a scene with poor satellite distribution conditions or few visible satellites.

Description

Technical Field

[0001] The present disclosure relates to intelligent driving technology, data processing technology, and positioning technology, and particularly relates to a vehicle positioning method and device, an electronic device, and a storage medium. Background Art

[0002] In scenarios such as intelligent driving scenarios or vehicle navigation, it is necessary to position a vehicle. Existing vehicle positioning technologies mainly rely on the Global Navigation Satellite System (GNSS). In an outdoor environment, a GNSS receiver can use the signals of all positioning satellites, so its signal is stronger, and the positioning data is more accurate and reliable. Summary of the Invention

[0003] Existing vehicle positioning technologies that rely on GNSS cannot effectively position a vehicle when the satellite distribution conditions are poor or the number of visible stars is small, because the satellite signals are blocked or interrupted. To solve the above technical problems, embodiments of the present disclosure provide a vehicle positioning method and device, an electronic device, and a storage medium to achieve effective vehicle positioning in scenarios where the satellite distribution conditions are poor or the number of visible stars is small.

[0004] An aspect of an embodiment of the present disclosure provides a vehicle positioning method, including: in response to a trigger condition, obtaining a reference trajectory in a scene map of a scene where the vehicle is currently located; determining a first feature point of the reference trajectory based on pose attribute information of reference trajectory points in the reference trajectory; determining incremental feature points based on pose attribute information of incremental driving trajectory points located within a current positioning window in the vehicle's current driving trajectory; matching the incremental feature points with the first feature point to obtain incremental feature point pairs, where the incremental feature point pairs include the matched incremental feature points and the first feature points; registering the reference trajectory and the driving trajectory based on a first feature point sequence formed by the first feature points in multiple groups of feature point pairs in chronological order, and a second feature point sequence formed by the second feature points in the multiple groups of feature point pairs in chronological order; where the multiple groups of feature point pairs at least include some of the feature point pairs obtained by matching the first feature points of the reference trajectory with the second feature points of the driving trajectory, and the multiple groups of feature point pairs at least include some of the incremental feature point pairs; in response to the registration of the reference trajectory and the driving trajectory being matched, determining the current positioning result of the vehicle based on the reference trajectory point corresponding to the end driving trajectory point in the registered driving trajectory.

[0005] Another aspect of the embodiments of the present disclosure provides a vehicle positioning device, including: a first acquisition module, configured to acquire a reference trajectory in a scene map of the current scene where the vehicle is located in response to a trigger condition; a first determination module, configured to determine a first feature point of the reference trajectory based on the pose attribute information of the reference trajectory points in the reference trajectory; a second determination module, configured to determine incremental feature points based on the pose attribute information of the incremental driving trajectory points located within the current positioning window in the current driving trajectory of the vehicle; a matching module, configured to match the incremental feature points with the first feature points to obtain incremental feature point pairs, where the incremental feature point pairs include the matched incremental feature points and the first feature points; a registration module, configured to register the reference trajectory and the driving trajectory based on a first feature point sequence formed by the first feature points in multiple groups of feature point pairs in chronological order, and a second feature point sequence formed by the second feature points in the multiple groups of feature point pairs in chronological order; where the multiple groups of feature point pairs at least include some of the feature point pairs obtained by matching the first feature points of the reference trajectory with the second feature points of the driving trajectory, and the multiple groups of feature point pairs at least include some of the incremental feature point pairs; a positioning module, configured to, in response to the matching of the registered reference trajectory and the driving trajectory, determine the current positioning result of the vehicle based on the reference trajectory point corresponding to the end driving trajectory point in the registered driving trajectory.

[0006] Another aspect of the embodiments of the present disclosure provides an electronic device, including: a memory, configured to store a computer program product; a processor, configured to execute the computer program product stored in the memory, and when the computer program product is executed, implement the vehicle positioning method according to any embodiment of the present disclosure.

[0007] Another aspect of the embodiments of the present disclosure provides a computer-readable storage medium, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, implement the method according to any embodiment of the present disclosure.

[0008] Another aspect of the embodiments of the present disclosure provides a computer program product, and when the instruction processor in the computer program product executes, implement the vehicle positioning method according to any embodiment of the present disclosure.

[0009] Based on the embodiments of the present disclosure, in response to a trigger condition, a reference trajectory in a scene map of the current scene where the vehicle is located is obtained, and a first feature point of the reference trajectory is determined based on the pose attribute information of the reference trajectory points in the reference trajectory. An incremental feature point is determined based on the pose attribute information of the incremental driving trajectory points located within the current positioning window in the current driving trajectory of the vehicle. The incremental feature point is matched with the first feature point to obtain an incremental feature point pair (including the matched incremental feature point and the first feature point). Based on the first feature point sequence formed by the first feature points in multiple groups of feature point pairs in chronological order and the second feature point sequence formed by the second feature points in multiple groups of feature point pairs in chronological order, the reference trajectory and the driving trajectory are registered. The multiple groups of feature point pairs at least include some of the feature point pairs obtained by matching the first feature points of the reference trajectory with the second feature points of the driving trajectory, and the multiple groups of feature point pairs at least include some of the incremental feature point pairs. In response to the matching of the registered reference trajectory and the driving trajectory, the current positioning result of the vehicle is determined based on the reference trajectory point corresponding to the end driving trajectory point in the registered driving trajectory. Thus, based on the feature points determined by the pose attribute information of the reference trajectory points in the reference trajectory and the feature points determined by the pose attribute information of the incremental driving trajectory points within the positioning window, through an incremental trajectory matching method, effective and dynamic positioning of the vehicle is achieved in scenarios with poor satellite distribution conditions or few visible satellites, and there is no need to deploy a large number of positioning base stations and other hardware devices in the above scenarios, resulting in low hardware costs and effectively reducing the overall cost of the positioning system. In addition, frequent data interaction between the vehicle and the positioning base station is effectively avoided, and the positioning efficiency can be effectively improved.

[0010] The technical solutions of the present disclosure will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 is an exemplary application scenario diagram of an embodiment of the present disclosure.

[0012] Figure 2 is a schematic flowchart of a vehicle positioning method provided by an exemplary embodiment of the present disclosure.

[0013] Figure 3 is a schematic flowchart of determining the first feature point of the reference trajectory in an exemplary embodiment of the present disclosure.

[0014] Figure 4 is a schematic diagram of a slope feature point in an exemplary embodiment of the present disclosure.

[0015] Figure 5 is a schematic flowchart of determining the first feature point of the reference trajectory in another exemplary embodiment of the present disclosure.

[0016] Figure 6A schematic diagram of steering feature points in an exemplary embodiment of the present disclosure.

[0017] Figure 7 A schematic flowchart of determining the first feature point of the reference trajectory in another exemplary embodiment of the present disclosure.

[0018] Figure 8 A schematic diagram of reference trajectory points and first valid trajectory points within a positioning window in an exemplary embodiment of the present disclosure.

[0019] Figure 9 A schematic diagram of global position feature points in an exemplary embodiment of the present disclosure.

[0020] Figure 10 A schematic flowchart of matching the incremental feature point with the first feature point in an exemplary embodiment of the present disclosure.

[0021] Figure 11 A schematic flowchart of registering the reference trajectory and the driving trajectory in an exemplary embodiment of the present disclosure.

[0022] Figure 12 A schematic flowchart of a vehicle positioning method provided in another exemplary embodiment of the present disclosure.

[0023] Figure 13 A schematic flowchart of a vehicle positioning method provided in another exemplary embodiment of the present disclosure.

[0024] Figure 14 A schematic flowchart of a vehicle positioning method provided in yet another exemplary embodiment of the present disclosure.

[0025] Figure 15 An application schematic diagram of adjusting the positioning reference point in an exemplary embodiment of the present disclosure.

[0026] Figure 16 A schematic flowchart of a vehicle positioning method provided in still another exemplary embodiment of the present disclosure.

[0027] Figure 17 A schematic flowchart of obtaining the reference trajectory in an exemplary embodiment of the present disclosure.

[0028] Figure 18 A schematic flowchart of a vehicle positioning method provided in another exemplary embodiment of the present disclosure.

[0029] Figure 19 A schematic flowchart of suppressing the reference trajectory in an exemplary embodiment of the present disclosure.

[0030] Figure 20 An application schematic diagram of suppressing the reference trajectory in an exemplary embodiment of the present disclosure.

[0031] Figure 21 It is a structural block diagram of a vehicle positioning device provided by an exemplary embodiment of the present disclosure.

[0032] Figure 22 It is a structural block diagram of a vehicle positioning device provided by another exemplary embodiment of the present disclosure.

[0033] Figure 23 It is a structural diagram of an electronic device provided by an exemplary embodiment of the present disclosure. Detailed implementation manners

[0034] To explain the present disclosure, exemplary embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all the embodiments. It should be understood that the present disclosure is not limited by the exemplary embodiments.

[0035] It should be noted that: unless otherwise specifically stated, the relative arrangements of components and steps, numerical expressions and values set forth in these embodiments do not limit the scope of the present disclosure.

[0036] It should be understood that the present disclosure emphasizes the differences between various embodiments. The same or similar parts can be referred to each other. For the sake of brevity, they will not be described one by one.

[0037] Technologies, methods and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods and devices should be regarded as part of the specification.

[0038] Overview of the present disclosure

[0039] Existing vehicle positioning technologies mainly rely on GNSS and need to receive signals from a certain number of effective satellites simultaneously to achieve accurate positioning of the vehicle. In an outdoor environment, a GNSS receiver can use the signals of all positioning satellites, so its signal is stronger and the positioning data is more accurate and reliable. However, in some scenarios with poor satellite distribution or few visible stars, for example, in indoor scenarios such as underground parking lots, logistics warehouses, large shopping malls, garages at airports or stations, multi-story stereoscopic garages, etc., or when the vehicle is driving in tunnels, between high-rise buildings in the city, etc., the satellite signals are blocked or interrupted, and effective positioning of the vehicle cannot be achieved.

[0040] To solve the above problems, in the related art, a large number of positioning base stations are deployed in the above scenarios to achieve vehicle positioning. The hardware cost is relatively high, increasing the overall cost of the positioning system. Moreover, in this implementation method, frequent data interaction is required between the vehicle and the positioning base station, thereby limiting the positioning efficiency and resulting in a relatively low positioning efficiency.

[0041] Exemplary application scenarios

[0042] Embodiments of the present disclosure can be used for positioning autonomous mobile devices (also referred to as agents), such as vehicles, robots, and drones, in scenarios where satellite signals are blocked or interrupted. For example, it can include, but is not limited to, positioning autonomous mobile devices in indoor scenarios such as underground parking lots, logistics warehouses, large shopping malls, garages at airports or stations, and multi-story stereoscopic garages, or positioning in near-indoor scenarios where the main mobile device is traveling between tunnels or high-rise buildings in the city. The vehicles therein can include, for example, but are not limited to cars, buses, logistics vehicles, engineering vehicles, etc., any vehicle with functions such as passenger carrying, cargo carrying, and transportation. Embodiments of the present disclosure do not limit the objects used for positioning and the specific application scenarios.

[0043] Figure 1 is an exemplary application scenario diagram of the embodiments of the present disclosure. In this embodiment, taking a vehicle as an example, an exemplary application of the embodiments of the present disclosure is described. As Figure 1As shown, an odometer module 120, a GNSS module 130, a storage module 140, and a computing platform 150 are deployed on the vehicle 110. Among them, the odometer module 120 can receive signals transmitted through the vehicle chassis sensors, such as signals transmitted through an Inertial Measurement Unit (IMU), a Wheel Speed Encoder, etc. For example, signals such as acceleration and heading angular velocity transmitted through the IMU, and signals such as wheel speed pulses transmitted through the wheel speed encoder, perform dead reckoning, and output the pose information of the vehicle 110 (referred to as the ego vehicle) in a specified local coordinate system (such as the ego vehicle coordinate system). At the same time, the GNSS module 130 can acquire satellite signals, based on the satellite signal propagation characteristics and the geometric triangulation principle, achieve high-precision position solution through the cooperation of multiple satellites, and output the positioning information of the ego vehicle in the global position coordinate system (i.e., the earth coordinate system). For example, it can include but is not limited to three-dimensional position coordinates, speed, and time and other information. The global position coordinate system among them can be, for example, the Universal Transverse Mercator (UTM) coordinate system. By converting the longitude and latitude coordinates on the earth's surface into a plane rectangular coordinate system, high-precision spatial data expression can be achieved. The storage module 140 can store the reference trajectory data in the scene map of at least one scene. The scene map of the at least one scene can be created by the ego vehicle, or can be created by other vehicles, and then the ego vehicle can obtain its reference trajectory data from other vehicles through the vehicle network, or can be uploaded to the cloud server by other vehicles and then obtained by the ego vehicle from the cloud server. The embodiments of the present disclosure do not limit the manner of obtaining the reference trajectory data in the scene map. The at least one scene among them can include but is not limited to indoor or near-indoor scenes such as underground parking lots, logistics warehouses, large shopping malls, garages in airports or stations, multi-story stereoscopic garages, tunnels, and between high-rise buildings in the city. In addition, the storage module 140 can also cache the driving trajectory data during the vehicle's current driving.

[0044] During the driving process of vehicle 110, the computing platform 140 can obtain the pose information output by the odometer module 120, the positioning information output by the GNSS module 130, and the reference trajectory in the scene map stored in the storage module 140 to locate the ego vehicle. Specifically, the computing platform 150 can, in response to a trigger condition, obtain the reference trajectory in the scene map of the scene where the ego vehicle is currently located, and determine the first feature point of the reference trajectory based on the pose attribute information of the reference trajectory points in the reference trajectory; determine the incremental feature points based on the pose attribute information of the incremental driving trajectory points within the current positioning window in the vehicle's current driving trajectory, and match the incremental feature points with the first feature point to obtain an incremental feature point pair, where the incremental feature point pair includes the matched incremental feature point and the first feature point; then, based on the first feature point sequence formed by the first feature points in multiple groups of feature point pairs in chronological order, and the second feature point sequence formed by the second feature points in multiple groups of feature point pairs in chronological order, register the reference trajectory and the driving trajectory, where the multiple groups of feature point pairs at least include some of the feature point pairs obtained by matching the first feature points of the reference trajectory with the second feature points of the driving trajectory, and the multiple groups of feature point pairs at least include some of the incremental feature point pairs; furthermore, in response to the matching of the registered reference trajectory and the driving trajectory, determine the current positioning result of the ego vehicle based on the reference trajectory point corresponding to the end driving trajectory point in the registered driving trajectory.

[0045] Thus, based on the pose attribute information of the reference trajectory points in the reference trajectory when creating the scene map by the ego vehicle or other vehicles and the pose attribute information of the incremental driving trajectory points within the positioning window, through an incremental trajectory matching method, effective positioning of the vehicle is achieved in scenarios with poor satellite distribution or few visible satellites, without the need to deploy a large number of hardware devices such as positioning base stations in the above scenarios, with low hardware costs, which can effectively control the overall cost of the positioning system, and furthermore, frequent data interaction between the vehicle and the positioning base station is effectively avoided, which can effectively improve the positioning efficiency.

[0046] For example, when the embodiment of the present disclosure is applied to an indoor scene such as an underground parking lot, based on the reference trajectory data of the underground parking lot map (as the scene map) pre-stored in vehicle 110, starting from the entrance of the underground parking lot or a certain position outside the entrance of the underground parking lot and aiming at the target parking space in the underground parking lot, vehicle 110 is located during the process of driving from the starting point to the destination point to navigate the route of vehicle 110 from the starting point to the destination point.

[0047] Figure 1This is only an implementation of an exemplary application scenario of the present disclosure. Those skilled in the art can know based on the records of the embodiments of the present disclosure that the embodiments of the present disclosure can also adopt any other feasible implementation with arbitrary deformation. For example, the computing platform 150 can also be deployed on a cloud server or a terminal device (such as a mobile phone terminal, a tablet computer, a PC, etc.). By communicating with the vehicle driving control system, it can obtain the pose information output by the odometer module 120, the positioning information output by the GNSS module 130, and the reference trajectory in the scene map stored in the storage module 140, and then locate the vehicle itself and feedback to the vehicle driving control system of the vehicle, so that the vehicle driving control system can perform downstream tasks based on the positioning result, such as vehicle driving path planning, vehicle driving control, and other downstream tasks.

[0048] Exemplary method

[0049] Figure 2 FIG. is a schematic flowchart of a vehicle positioning method provided by an exemplary embodiment of the present disclosure. The embodiments of the present disclosure can be applied to implement the positioning of any autonomous mobile device (such as a vehicle, a robot, a drone, etc.) in a scenario where satellite signals are blocked or interrupted. The embodiments of the present disclosure can be applied to any electronic device with data processing capabilities, such as but not limited to a terminal device with data processing capabilities (such as an in-vehicle terminal, a mobile phone terminal, a tablet computer, a PC, etc.), a cloud server, a vehicle driving control system, or a computing platform in other autonomous mobile device control systems. By obtaining the reference trajectory data in the scene map of the scene and the driving trajectory data of the autonomous mobile device during this driving, the positioning result of the autonomous mobile device is determined, so as to perform downstream tasks such as vehicle driving path planning and vehicle driving control accordingly. The embodiments of the present disclosure are described by taking the autonomous mobile device as a vehicle as an example. Those skilled in the art can know based on the records of the embodiments of the present disclosure that the positioning of other arbitrary autonomous mobile devices (such as robots, drones, etc.) can be implemented with reference. As Figure 2 shown, the vehicle positioning method of the embodiments of the present disclosure includes:

[0050] 210. In response to a trigger condition, obtain the reference trajectory in the scene map of the scene where the vehicle is currently located.

[0051] In the embodiments of the present disclosure, the trigger condition can be a condition preset for executing the vehicle positioning method of the embodiments of the present disclosure. In some implementation manners, the trigger condition can include, for example, but not limited to at least one of the following: receiving a positioning instruction sent by a user; detecting that the vehicle (i.e., the current vehicle, also referred to as the ego vehicle) enters a scene with poor satellite distribution conditions or few visible stars, such as indoor scenes such as underground parking lots, logistics warehouses, large shopping malls, garages in airports or stations, multi-story stereoscopic garages, etc., or driving in near-indoor scenes such as tunnels or between high-rise buildings in the city.

[0052] In a specific implementation, when the vehicle enters or is about to enter a specified scenario (such as a scenario with poor satellite distribution conditions or few visible stars), for example, when entering an underground parking lot, the user can send a positioning instruction to the electronic device for implementing the embodiments of the present disclosure, and then the electronic device can execute the operation 210 in response to the positioning instruction. Alternatively, the electronic device can also execute the operation 210 when it detects that the vehicle enters a specified scenario (such as a scenario with poor satellite distribution conditions or few visible stars), for example, when it detects that the vehicle enters the entrance of the underground parking lot from the ground.

[0053] In the embodiments of the present disclosure, the reference trajectory, that is, the trajectory in the scenario map, can be the driving trajectory when the ego vehicle or other vehicles create the scenario map. Each scenario map can include at least one reference trajectory. For example, when an underground parking lot includes multiple floors of garages, each floor of the garage can correspond to one reference trajectory, and then the scenario map of the underground parking lot can include multiple reference trajectories corresponding to multiple floors of garages. Another example is that when an underground parking lot includes multiple entrances, each entrance to the corresponding parking space can correspond to one reference trajectory, and then the scenario map of the underground parking lot can include multiple reference trajectories from multiple entrances to the corresponding parking spaces. If there are multiple reference trajectories in the scenario map of the scenario where the vehicle is currently located, subsequent operations can be executed respectively for each reference trajectory, and specific details can refer to the relevant introduction in the following embodiments. It can be understood that for the case where each floor of the garage corresponds to one reference trajectory, the reference trajectory can be the trajectory from the garage entrance to the corresponding parking space.

[0054] In the embodiments of the present disclosure, the ego vehicle can pre-store the scenario map data of the scenario where it is currently located. Then, in this operation 210, in response to the trigger condition, the reference trajectory in the scenario map of the scenario where it is currently located can be directly obtained from the scenario map data stored by the ego vehicle. Alternatively, the ego vehicle can also, in response to the trigger condition, when it does not store the scenario map data of the scenario where it is currently located, based on the current position information of the ego vehicle, obtain the reference trajectory in the scenario map of the scenario where it is currently located from the cloud. Alternatively, the ego vehicle can also, in response to the trigger condition, when it does not store the scenario map data of the scenario where it is currently located, based on the current position information of the ego vehicle, obtain the reference trajectory in the scenario map of the scenario where it is currently located from other vehicles that can interact through the vehicle network through the vehicle network. The embodiments of the present disclosure do not limit the specific implementation manner of obtaining the reference trajectory in the scenario map of the scenario where the ego vehicle is currently located.

[0055] 220. Based on the pose attribute information of the reference trajectory points in the reference trajectory, determine the first feature point of the reference trajectory.

[0056] In the embodiments of the present disclosure, the reference trajectory point is the trajectory point in the reference trajectory.

[0057] In some of these implementation manners, the pose attribute information may include the pose information in a specified local coordinate system. The specified local coordinate system may be, for example, the vehicle coordinate system corresponding to the starting point (which may be referred to as the reference datum point) of the vehicle creating the scene map on the reference trajectory, or any coordinate system such as the world coordinate system. The coordinate systems can be transformed into each other. For example, the pose information in the vehicle coordinate system can be transformed into the world coordinate system based on the pre-determined transformation parameters between the coordinate systems, and vice versa. The pose information in the specified local coordinate system may include, for example, but is not limited to, three-dimensional coordinate information and attitude information in the specified local coordinate system. The attitude information may include, for example, but is not limited to at least one of the following: yaw angle (also known as yaw, Yaw) information (i.e., yaw angle or yaw angle), pitch angle (Pitch) information (i.e., pitch angle), and roll angle (Roll) information (i.e., roll angle).

[0058] In addition, the pose attribute information may also selectively include the global position information in the global position coordinate system (i.e., the earth coordinate system). The global position coordinate system may be, for example, the UTM coordinate system. Correspondingly, the global position information in the UTM coordinate system is the position information of the UTM coordinate point, which consists of three parts: zone number (i.e., zone identifier), easting (also known as abscissa, representing the offset relative to the central meridian of the UTM zone where it is located), and northing (also known as ordinate, representing the offset relative to the equator). The UTM coordinate system is a planar coordinate system that divides the earth's surface into 60 longitudinal zones, each zone covering 6 degrees of longitude, from 180 degrees west longitude to 180 degrees east longitude. Each zone has a unique identifier (i.e., zone identifier). The UTM coordinates and the latitude and longitude coordinates can be transformed into each other. By using the UTM coordinate points, distances can be more conveniently measured and calculated in the planar coordinate system.

[0059] In some of these implementation manners, the first feature point is the feature point of the reference trajectory.

[0060] In some of these implementation manners, in the operation 220, during the positioning process of the vehicle's current driving, based on the pose attribute information of the reference trajectory points, the first feature point of the reference trajectory can be determined immediately, such as the slope feature point, the steering feature point, and the global position feature point. Or, in some of these implementation manners, after the first feature point of the reference trajectory is determined in the positioning process of the vehicle's historical driving, the first feature point of the reference trajectory can be stored. Then, in the operation 220, the stored first feature point of the reference trajectory can be directly obtained. When the stored first feature point of the reference trajectory is not obtained, the first feature point of the reference trajectory can be determined immediately based on the pose attribute information of the reference trajectory points. The specific implementation manner for determining the first feature point of the reference trajectory in the embodiments of the present disclosure is not limited.

[0061] Based on the pose attribute information of the incremental driving trajectory points within the current positioning window in the vehicle's current driving trajectory, determine the incremental feature points.

[0062] In the embodiments of the present disclosure, the positioning window is a window used for positioning the vehicle once during vehicle driving, determined based on a preset positioning interval. The preset positioning interval can be a driving distance (such as 2m), a time interval (such as 0.01s, 0.3s), etc. Accordingly, the vehicle is positioned once every 2m during driving, or the vehicle is positioned once every 0.3s during driving. In practical applications, information such as the real-time positioning requirements and the computing power resources of the computing platform can be comprehensively considered to determine the positioning interval. When the computing power resources support it, the positioning interval can be set small enough to achieve real-time or quasi-real-time positioning of the vehicle.

[0063] In the embodiments of the present disclosure, the incremental driving trajectory points are the trajectory points within the current positioning window in the driving trajectory corresponding to the vehicle's current driving (hereinafter referred to as: driving trajectory points), that is, the newly added driving trajectory points within the current positioning window.

[0064] In some implementation manners, the pose attribute information of the incremental driving trajectory points may include the pose information of the incremental driving trajectory points in a specified local coordinate system. The specified local coordinate system can be, for example, the vehicle coordinate system corresponding to the starting point of the vehicle's current driving trajectory (which can be called the positioning reference point), or any coordinate system such as the world coordinate system, and the coordinate systems can be transformed with each other. The pose information of the incremental driving trajectory points in the specified local coordinate system may include, for example, but is not limited to, the three-dimensional coordinate information and the attitude information of the incremental driving trajectory points in the specified local coordinate system. The attitude information may include yaw angle information, pitch angle information, and roll angle information (i.e., roll angle). The three-dimensional coordinate information and the attitude information of the incremental driving trajectory points in the specified local coordinate system can be output by an odometer module deployed on the vehicle. The odometer module can receive signals transmitted through a vehicle chassis sensor (such as an IMU), for example, signals such as acceleration and heading angular velocity transmitted through the IMU, and signals such as wheel speed pulses transmitted through a wheel speed encoder, perform dead reckoning, and output the pose information of the trajectory point where the vehicle is currently located in the specified local coordinate system (such as the ego-vehicle coordinate system).

[0065] In addition, the pose attribute information of the incremental driving trajectory points may also selectively include the global position information (such as the coordinate position information of a UTM coordinate point) of the incremental driving trajectory points in the global position coordinate system (i.e., the earth coordinate system, such as the UTM coordinate system). This pose attribute information can be obtained through GNSS deployed on the vehicle.

[0066] In the embodiments of the present disclosure, this operation 230 can be sequentially performed for each positioning window. There is no execution order restriction between operation 230 and operation 220, and they can be executed simultaneously or sequentially with any time difference. The embodiments of the present disclosure do not limit this.

[0067] 240. Match the incremental feature points with the first feature points to obtain incremental feature point pairs.

[0068] Match the incremental feature points with the first feature points. The successfully matched feature point pairs are the incremental feature point pairs. Each incremental feature point pair includes a successfully matched incremental feature point and a first feature point.

[0069] 250. Based on the first feature point sequence formed by the first feature points in multiple groups of feature point pairs in time sequence, and the second feature point sequence formed by the second feature points in multiple groups of feature point pairs in time sequence, register the reference trajectory and the driving trajectory.

[0070] In the embodiments of the present disclosure, the second feature points are the feature points of the driving trajectory. The second feature points may include the historical incremental feature points corresponding to each historical positioning window before the current positioning window in the current driving trajectory and the incremental feature points corresponding to the current positioning window. The multiple groups of feature point pairs at least include some of the feature point pairs obtained by matching the first feature points of the reference trajectory with the second feature points of the driving trajectory, that is, the multiple groups of feature point pairs may include all the feature point pairs obtained by matching the first feature points with the second feature points, or may only include any part of the feature point pairs obtained by matching the first feature points with the second feature points; and the multiple groups of feature point pairs at least include some of the incremental feature point pairs, that is, the multiple groups of feature point pairs may include the incremental feature point pairs or any part of the incremental feature point pairs. The embodiments of the present disclosure do not limit this.

[0071] In some of these implementation manners, any trajectory registration algorithm can be adopted to register the reference trajectory and the driving trajectory. For example, the Umeyama algorithm can be used to register the reference trajectory and the driving trajectory to obtain a registration result. The Umeyama algorithm is an algorithm for calculating the optimal similarity transformation (including rotation, scaling, and translation) between two sets of point clouds or trajectories. By finding a set of parameters (rotation matrix R, scaling coefficient c, and translation vector t), one of the trajectories to be registered can be transformed to align as accurately as possible with the other trajectory after the transformation. Specifically, the Umeyama algorithm can be used to calculate a set of optimal transformation parameters (rotation matrix R, scaling coefficient c, and translation vector t) based on the first feature point sequence and the second feature point sequence, and use the calculated transformation parameters to transform one of the trajectories (driving trajectory or reference trajectory), so that the error between the transformed trajectory (driving trajectory or reference estimate) and the other trajectory (reference trajectory or driving trajectory) is minimized; evaluate the similarity between the transformed trajectory and the other trajectory, such as the root mean square error (RMSE) or other similarity metrics, which is used to characterize the alignment degree between the transformed trajectory and the other trajectory.

[0072] 260. In response to the matching of the registered reference trajectory and the driving trajectory, determine the current positioning result of the vehicle based on the reference trajectory point corresponding to the end driving trajectory point in the registered driving trajectory.

[0073] In some of these implementation manners, the matching of the registered reference trajectory and the driving trajectory means that the similarity between the registered reference trajectory and the driving trajectory meets a preset condition.

[0074] In the embodiments of the present disclosure, operations 230-260 are operations iteratively executed based on each positioning window during the current driving of the vehicle when the triggering condition is met, that is, for each additional positioning window, operations 230-260 are executed until the end of the current driving task or the above triggering condition is no longer met.

[0075] Based on this embodiment, in response to a trigger condition, a reference trajectory in the scene map of the current scene where the vehicle is located is obtained, and a first feature point of the reference trajectory is determined based on the pose attribute information of the reference trajectory points in the reference trajectory. An incremental feature point is determined based on the pose attribute information of the incremental travel trajectory points located within the current positioning window in the vehicle's current travel trajectory. The incremental feature point is matched with the first feature point to obtain an incremental feature point pair. Based on the first feature point sequence formed by the first feature points in multiple groups of feature point pairs in chronological order and the second feature point sequence formed by the second feature points in multiple groups of feature point pairs in chronological order, the reference trajectory and the travel trajectory are registered. In response to the matching of the registered reference trajectory and the travel trajectory, the current positioning result of the vehicle is determined based on the reference trajectory point corresponding to the end travel trajectory point in the registered travel trajectory. Thus, based on the feature points determined from the pose attribute information of the reference trajectory points in the reference trajectory and the feature points determined from the pose attribute information of the incremental travel trajectory points within the positioning window, through an incremental trajectory matching method, effective and dynamic positioning of the vehicle is achieved in scenarios with poor satellite distribution conditions or few visible stars, and there is no need to deploy a large number of positioning base stations and other hardware devices in the above scenarios. The hardware cost is low, which can effectively reduce the overall cost of the positioning system. In addition, frequent data interaction between the vehicle and the positioning base station is effectively avoided, and the positioning efficiency can be effectively improved.

[0076] Optionally, in some implementation manners, the types of the first feature point, the second feature point, and the incremental feature point may include, for example, but are not limited to, at least one of a slope feature point, a steering feature point, and a global position feature point. Since the slope feature point, the steering feature point, and the global position feature point can characterize the driving road features or position information of the corresponding location, based on the information of the slope feature point, the steering feature point, and the global position feature point, it is helpful to determine the positions represented by the respective feature points.

[0077] Figure 3 It is a schematic flowchart of determining the first feature point of the reference trajectory in an exemplary embodiment of the present disclosure. As Figure 3 shown, on the basis of the embodiment shown in Figure 2 in some implementation manners, when the type of the feature point (at least one of the first feature point, the second feature point, and the incremental feature point) is a slope feature point, operation 220 may include:

[0078] 2210. For the reference trajectory points in the reference trajectory, by means of a sliding window, based on the pitch angle information in the pose attribute information of the reference trajectory points within the sliding window, the first reference trajectory point and the second reference trajectory point corresponding to the ramp entry point and the ramp exit point of the ramp in the reference trajectory are determined.

[0079] In some of these implementation manners, for each reference trajectory point, the pitch angle information of the vehicle that creates the scene map at this reference trajectory point, which is calculated by the vehicle chassis sensor, can be obtained. For example, the pitch angle of the vehicle estimated by integrating the IMU of the vehicle is used as the plane slope value of the vehicle at this reference trajectory point. By using the pitch angles of multiple reference trajectory points within a sliding window and combining a preset pitch angle threshold, it is determined whether the vehicle has passed through a ramp and a flat ground, and the reference trajectory points corresponding to the peak and valley points of the pitch angle in the time-sequentially arranged pitch angle sequence are found as the entry ramp point and the exit ramp point. Among them, the length of the sliding window can be set according to actual needs, and the length of the sliding window can be adaptively adjusted according to the vehicle speed. When the vehicle is at high speed, the window can be enlarged to reduce the misjudgment rate. For example, the size of the sliding window can be set to a preset number of frames. Each frame samples a reference trajectory point, so the size of the sliding window is the reference trajectory points corresponding to the preset number of frames N (for example, 8 frames). For example, when the preset number of frames is 8, that is, the pitch angle information in the pose attribute information of 8 reference trajectory points is obtained in the way of a sliding window to determine the entry ramp point and the exit ramp point corresponding to the ramp in the reference trajectory. The pitch angle threshold can be preset. For example, it can be +5° / -5°, corresponding to the uphill threshold and the downhill threshold respectively.

[0080] In a specific example, the average value of the pitch angles in the state without a ramp can be pre-statistically calculated as a reference value. When the pitch angle fluctuation within the sliding window is within the range of the reference value ± the pitch angle threshold, it is determined that the corresponding reference trajectory point is in a flat ground state. When the average value of the pitch angles within the sliding window continuously exceeds the pitch angle threshold, such as the uphill threshold (such as +5°) or the downhill threshold (such as -5°) for N frames (for example, 3 to 5 frames), it is determined that the vehicle has entered a ramp state. From the time-sequential pitch angle sequences corresponding to each sliding window, local maximum values (peaks) and minimum values (valleys) are found by the first-order difference or second derivative method; when the pitch angle within the sliding window continuously exceeds the uphill threshold, it is determined that the vehicle is currently in an upward trend. If the pitch angle first rises from the reference value and exceeds the pitch angle threshold, corresponding to the starting point of the first significant peak (valley-to-peak transition), the reference trajectory point corresponding to this pitch angle is determined as the entry ramp point; when the pitch angle rapidly drops and exceeds the downhill threshold, it is determined that the vehicle has entered the exit ramp state. If the pitch angle drops from the peak to below the reference value, corresponding to the end point of the last significant peak (peak-to-valley transition), the reference trajectory point corresponding to this pitch angle is determined as the exit ramp point.

[0081] 2211, determining a first slope feature point based on a first reference trajectory point and a second reference trajectory point.

[0082] In this embodiment, the first feature point includes a first slope feature point, and the first slope feature point is a slope feature point determined based on a first reference trajectory point and a second reference trajectory point.

[0083] In some of these implementation manners, the first reference trajectory point and the second reference trajectory point can be determined as the first slope feature points. Alternatively, in some other implementation manners, the first reference trajectory point and the second reference trajectory point, and at least one reference trajectory point between the first reference trajectory point and the second reference trajectory point corresponding to the same ramp can be determined as the first slope feature points.

[0084] Correspondingly, operation 230 includes the following operations performed successively for each positioning window:

[0085] 2310. For the incremental driving trajectory points and adjacent historical driving trajectory points, in a sliding window manner, based on the pitch angle information in the pose attribute information of the driving trajectory points within the sliding window, determine the first driving trajectory point and the second driving trajectory point corresponding to the ramp entry point and the ramp exit point in the driving trajectory respectively.

[0086] Among them, the adjacent historical driving trajectory points include at least one driving trajectory point whose time sequence in the driving trajectory is before the incremental driving trajectory points.

[0087] In this embodiment, the implementation manner in operation 2210 can be referred to, and based on the pitch angle information in the pose attribute information of the driving trajectory points within the sliding window, determine the first driving trajectory point and the second driving trajectory point corresponding to the ramp entry point and the ramp exit point in the driving trajectory respectively, which will not be elaborated here.

[0088] 2311. Based on the first driving trajectory point and the second driving trajectory point, determine the second slope feature point.

[0089] In this embodiment, the second feature point and the incremental feature point include the second slope feature point, and the second slope feature point is the slope feature point determined based on the first driving trajectory point and the second driving trajectory point.

[0090] In some of these implementation manners, the first driving trajectory point and the second driving trajectory point can be determined as the second slope feature points. Alternatively, in some other implementation manners, the first driving trajectory point and the second driving trajectory point, and at least one driving trajectory point between the first driving trajectory point and the second driving trajectory point corresponding to the same ramp can be determined as the second slope feature points.

[0091] As Figure 4 shown, it is a schematic diagram of slope feature points in an exemplary embodiment of the present disclosure.

[0092] Based on this embodiment, by means of a sliding window, based on the pitch angle information in the pose attribute information of trajectory points (including reference trajectory points and driving trajectory points), slope feature points in the corresponding trajectories (reference trajectory, driving trajectory) are determined. Since the slope feature points can characterize the slope features of the driving road at the corresponding locations, matching the reference trajectory and the driving trajectory based on the matching relationship between the slope features of the reference trajectory points and the driving trajectory points helps to determine the reference trajectory points corresponding to the end driving trajectory points of each positioning window, thereby determining the positioning result of the vehicle.

[0093] Figure 5 It is a schematic flow chart for determining the first feature point of the reference trajectory in another exemplary embodiment of the present disclosure. As Figure 5 shown, on the basis of the embodiment shown in Figure 2 in some implementation manners, when the type of the feature point (at least one of the first feature point, the second feature point, and the incremental feature point) is a steering feature point, operation 220 may include:

[0094] 2212, based on the yaw angle information in the pose attribute information of the reference trajectory points, determine that the third reference trajectory point corresponding to the vehicle steering in the reference trajectory is the first steering feature point.

[0095] In this embodiment, the first feature point includes the first steering feature point. The third reference trajectory point is the trajectory point corresponding to the vehicle steering in the reference trajectory determined based on the yaw angle information in the pose attribute information of the reference trajectory points.

[0096] The yaw angle of the vehicle is a key parameter describing the body posture when the vehicle steers, and it reflects the rotation angle of the vehicle around the vertical axis.

[0097] In some implementation manners, the yaw angle of the vehicle at each trajectory point can be estimated by trajectory position difference, that is, by continuously recording the position coordinates of the vehicle at different times (i.e., different trajectory points) and calculating the change amount between these position coordinates, estimating the yaw angle of the vehicle. Furthermore, by detecting the position point where the change rate of the yaw angle (i.e., the yaw angular velocity, also called the yaw rate) is greater than the preset angular velocity threshold as the trajectory point corresponding to the vehicle steering.

[0098] In some other implementations, other sensor data can be combined, for example, the direct measurement values of the accelerometer and gyroscope in the IMU regarding the vehicle acceleration and angular velocity, and the direct measurement values of the wheel speed encoder regarding the vehicle wheel speed pulses, to more accurately estimate the yaw angle. For example, the wheel speed difference estimation method can be used to estimate the yaw angular velocity based on the front and rear wheel speed differences, and then integrate to obtain the yaw angle; or, the lateral acceleration - vehicle speed method can be used to calculate the yaw angular velocity by measuring the lateral acceleration and vehicle speed of the vehicle, and then integrate to obtain the yaw angle; or, directly obtain the yaw angular velocity from the IMU, and then integrate to obtain the yaw angle, and so on. The embodiments of the present disclosure do not limit the implementation manners of obtaining the yaw angle information.

[0099] Correspondingly, operation 230 includes the following operations sequentially performed for each positioning window:

[0100] 2312. Based on the yaw angle information in the pose attribute information of the incremental driving trajectory points and the adjacent historical driving trajectory points, determine that the third driving trajectory point corresponding to the vehicle steering in the driving trajectory is the second steering feature point.

[0101] In this embodiment, the second feature point and the incremental feature point include the second steering feature point, and the adjacent historical driving trajectory points include at least one driving trajectory point in the driving trajectory whose time sequence is before the incremental driving trajectory points. The third driving trajectory point is the driving trajectory point corresponding to the vehicle steering in the driving trajectory determined based on the yaw angle information in the pose attribute information of the incremental driving trajectory points and the adjacent historical driving trajectory points.

[0102] In this embodiment, for the acquisition method of the yaw angle information, reference can be made to the implementation manner in operation 2212, which will not be elaborated here.

[0103] As Figure 6 shown, it is a schematic diagram of the steering feature point in an exemplary embodiment of the present disclosure.

[0104] Based on this embodiment, the steering feature points in the corresponding trajectories (reference trajectory, driving trajectory) can be determined based on the yaw angle information in the pose attribute information of the trajectory points (including reference trajectory points and driving trajectory points). Since the steering feature points can characterize the method features of the driving roads at the corresponding locations, matching the reference trajectory and the driving trajectory based on the matching relationship between the direction features of the reference trajectory points and the driving trajectory points helps to determine the reference trajectory points corresponding to the end driving trajectory points of each positioning window, thereby determining the positioning result of the vehicle.

[0105] Figure 7 is a schematic flowchart of determining the first feature point of the reference trajectory in another exemplary embodiment of the present disclosure. As Figure 7 shown, in Figure 2Based on the illustrated embodiments, in some implementations, the above relative position information is position information in a specified local coordinate system, and the above global position information is position information in a global position coordinate system. When the type of the feature point (at least one of the first feature point, the second feature point, and the incremental feature point) is a global position feature point, operation 220 may include:

[0106] 2213. For each reference trajectory point within each positioning window, determine a first valid trajectory point based on the satellite signal information in the pose attribute information of the reference trajectory point and the relationship between the first variation and the second variation.

[0107] Wherein, the first variation is the variation of the relative position information in the pose attribute information of the reference trajectory point with respect to the relative position information in the pose attribute information of the previous adjacent reference trajectory point, and the second variation is the variation of the global position information in the pose attribute information of the reference trajectory point with respect to the global position information in the pose attribute information of the previous adjacent reference trajectory point.

[0108] In some of these implementations, satellite signal information can be obtained from GNSS. The satellite signal information can include, for example, but is not limited to, global position information of each observation point, the number of satellites in view (e.g., the number of satellites in satellite navigation systems such as the Global Positioning System (GPS), Beidou, etc.), signal quality, Real Time Kinematic (RTK) solution status, and other information. The signal quality can include, for example, but is not limited to, at least one of signal strength, carrier-to-noise ratio (C / N0), signal-to-noise ratio (SNR), dilution of precision (DOP), etc. The RTK solution status is a key indicator for measuring positioning accuracy and signal stability. The RTK solution status can include, for example, but is not limited to, at least one of fixed solution (RTK Fixed), float solution (also known as floating solution, RTK Float), single-point solution, pseudorange solution, no solution, known state, differential solution, static solution baseline state, etc. For at least one reference trajectory point within each positioning window, based on the satellite signal information of the at least one reference trajectory point, observation points (corresponding to the reference trajectory points) with better satellite signals are selected from the at least one reference trajectory point. For example, observation points with high SNR, the number of satellites in view meeting a preset number, and good geometric distribution (i.e., low DOP value), and / or observation points with an RTK solution status of fixed solution or float solution are selected to reduce multipath effects and atmospheric delay errors. Then, outlier observation points are removed based on the Euclidean distance between the observation points. For example, observation points with signal interruption, jumps, or outliers are removed. Regarding the calculation method of the Euclidean distance, reference can be made to related technologies and will not be introduced here. After that, for each reference trajectory point corresponding to the remaining observation points after removing the outlier observation points, the change amount of the relative position information in the pose attribute information of the reference trajectory point relative to the relative position information in the pose attribute information of the previous adjacent reference trajectory point is obtained as the first change amount. At the same time, the change amount of the global position information in the pose attribute information of the reference trajectory point relative to the global position information in the pose attribute information of the previous adjacent reference trajectory point is obtained as the second change amount. It is determined whether the relationship between the first change amount and the second change amount meets a preset condition. For example, the ratio between the first change amount and the second change amount is close to 1 or the difference from 1 is within a preset difference range (e.g., less than 0.05). Whether the relationship between the first change amount and the second change amount meets the preset condition indicates that the change in the relative position and the change in the global position of two adjacent reference trajectory points are highly consistent, and the position of this reference trajectory point is relatively accurate and stable. Each reference trajectory point with the ratio between the first change amount and the second change amount meeting the preset condition is selected as a first valid trajectory point. Thus, valid observation points can be screened out as valid trajectory points for each positioning window respectively.

[0109] As Figure 8 shown, it is a schematic diagram of reference trajectory points and first valid trajectory points within a positioning window in an exemplary embodiment of the present disclosure.

[0110] In practical applications, there may be deviations in the global position information in satellite signal information. For example, during vehicle driving, based on the position information in the pose information output by the odometer module at two consecutive moments (such as moment A and B), the vehicle has traveled forward 100 meters (as the first relative quantity); while based on the UTM coordinate points corresponding to the above two consecutive moments A and B output by GNSS, the vehicle has traveled forward 50 meters (as the second relative quantity), and the ratio between the two is 2 and not close to 1, indicating that the consistency between the mileage information measured by the odometer module and the distance information measured by GNSS is low, and there is an error in this measurement of the position. Selecting each reference trajectory point where the ratio between the first change quantity and the second change quantity meets the preset condition as the first effective trajectory point helps to accurately determine the global position feature point, thereby achieving the accuracy of the vehicle positioning result.

[0111] 2214. Cluster the first effective trajectory points corresponding to each positioning window respectively to obtain the first clustering result corresponding to each positioning window.

[0112] In some implementation manners, a spatial clustering algorithm can be adopted, such as the density-based unsupervised clustering algorithm (Density-Based Spatial Clustering of Applications with Noise, DBSCAN) or other density clustering algorithms, to cluster the effective observation points to obtain the first clustering result. Specifically, a spatial clustering algorithm can be used to perform spatial grouping on the effective observation points, screen out the areas where the point density is higher than the preset threshold (referred to as dense areas), and determine the effective observation points in this dense area as dense effective observation points (i.e., the first clustering result); alternatively, it is also possible to selectively retain the observation points with uniform time intervals in the same dense area to avoid central deviation caused by local oversampling.

[0113] 2215. Respectively determine the center of the first clustering result corresponding to each positioning window as the first global position feature point.

[0114] In this embodiment, the first feature point includes the first global position feature point, and this first global position feature point is used to represent the center of the first clustering result corresponding to each positioning window.

[0115] In some implementation manners, for each positioning window respectively, determine the center of the first clustering result corresponding to this positioning window as the first global position feature point.

[0116] In some other implementations, for the valid observation points in the dense area corresponding to each positioning window, the weighted geometric center coordinates (such as longitude and latitude) can be calculated with the accuracy (such as the horizontal dilution of precision) of each valid observation point in the dense area as the weight, and the geometric center is used as the first global position feature point.

[0117] Correspondingly, operation 230 includes the following operations sequentially performed for each positioning window:

[0118] 2313. Determine a second valid trajectory point based on the satellite signal information in the pose attribute information of the incremental driving trajectory points and the relationship between the third variation and the fourth variation.

[0119] Wherein, the third variation is the variation of the relative position information in the pose attribute information of the incremental driving trajectory point with respect to the relative position information in the pose attribute information of the previous adjacent driving trajectory point, and the fourth variation is the variation of the global position information in the pose attribute information of the incremental driving trajectory point with respect to the global position information in the pose attribute information of the previous adjacent driving trajectory point.

[0120] Among them, the valid trajectory point corresponding to the incremental driving trajectory point can be determined as the second valid trajectory point in a manner similar to operation 2213, which will not be elaborated here.

[0121] 2314. Cluster the second valid trajectory points to obtain a second clustering result.

[0122] Among them, the second valid trajectory points can be clustered in a manner similar to operation 2214 to obtain a second clustering result, which will not be elaborated here.

[0123] 2315. Determine the center of the second clustering result as the second global position feature point.

[0124] In the embodiments of the present disclosure, the second feature point and the incremental feature point include the second global position feature point, and the second global position feature point is used to represent the center of each second clustering result.

[0125] Among them, the center of the second clustering result can be determined as the second global position feature point in a manner similar to operation 2215, which will not be elaborated here.

[0126] As Figure 9 shown, it is a schematic diagram of the global position feature point in an exemplary embodiment of the present disclosure.

[0127] Based on this embodiment, for the trajectory points (including reference trajectory points and incremental driving trajectory points) within each positioning window, through processing such as observation point screening based on satellite signal information, outlier observation point elimination, consistency screening of the change amount of relative position information and the change amount of global position, and spatial clustering, high-precision global position feature points can be extracted from dense valid observation points. Since the global position feature points can effectively represent the positions of the corresponding locations, matching the reference trajectory and the driving trajectory based on the matching relationship between the global position feature points of the reference trajectory points and the driving trajectory points helps to determine the reference trajectory points corresponding to the end driving trajectory points of each positioning window, improving the reliability and accuracy of the vehicle positioning result.

[0128] Optionally, in some implementation manners, the types of the first feature point, the second feature point, and the incremental feature point may respectively include at least one of a slope feature point, a steering feature point, and a global position feature point. Figure 10 It is a schematic flowchart of matching the incremental feature point with the first feature point in an exemplary embodiment of the present disclosure. As Figure 10 shown, based on any of the embodiments shown in Figures 2 - 9 this embodiment, in this embodiment, operation 240 may include:

[0129] The type of the incremental feature point can be determined. In response to the type of the incremental feature point being a slope feature point, operations 2410-2412 are executed; in response to the type of the incremental feature point being a steering feature point, operations 2413-2415 are executed; in response to the type of the incremental feature point being a global position feature point, operations 2416-2418 are executed.

[0130] For example, in a specific example, the length of the reference trajectory is 100 meters, the number of the first feature points of the determined reference trajectory is 10, and the 10 first feature points form a first feature point sequence in chronological order. Among the 10 first feature points, there are 4 slope feature points, 3 turning feature points, and 3 global position feature points. The length of the vehicle's current driving trajectory is 80 meters, the number of the second feature points of the determined driving trajectory is 45, and the 45 second feature points form a second feature point sequence in chronological order. Among them, the number of incremental feature points within the current positioning window is 5, including 2 slope feature points and 3 turning feature points. When matching the incremental feature points with the first feature points, the types of the first feature points in the first feature point sequence can be determined first, and 4 slope feature points are respectively extracted to form a first slope feature point sequence, 3 turning feature points are respectively extracted to form a first turning feature point sequence, and 3 global position feature points are respectively extracted to form a first global feature point sequence. At the same time, it is determined that the types of the incremental feature points are 2 slope feature points and 3 turning feature points. For these 2 slope feature points, operations 2410-2412 are respectively executed, that is, these 2 slope feature points are respectively matched with the first slope feature points in the first slope feature point sequence to determine the first slope feature points they match. For these 3 turning feature points, operations 2413-2415 are respectively executed, that is, these 3 turning feature points are respectively matched with the first turning feature points in the first turning feature point sequence to determine the first turning feature points they match.

[0131] 2410. Determine a first slope feature point sequence based on the first slope feature points among the first feature points.

[0132] In some implementation manners, the first slope feature points with the type of slope feature points can be determined from the first feature points, and the determined first slope feature points are formed into a sequence based on the chronological order of entering the current scene to obtain a first slope feature point sequence.

[0133] 2411. Based on the pose attribute information of the incremental feature points and the pose attribute information of the first slope feature points in the first slope feature point sequence, match the incremental feature points with the first slope feature points in the first slope feature point sequence to obtain a first similarity.

[0134] Among them, the first similarity is used to characterize the matching degree between the incremental feature points and the first slope feature points. The higher the value of the first similarity, the higher the matching degree between the two.

[0135] In some of these implementations, for each slope feature point (i.e., the second slope feature point) in the incremental feature points, based on the slope attribute of the second slope feature point (such as the entry slope point, the exit slope point, the ramp or flat ground, or the pitch angle information), the slope attribute of each first slope feature point in the first slope feature point sequence, and the distance between the second slope feature point and each first slope feature point, the second slope feature point and each first slope feature point are matched. For example, based on the slope attribute and the feature point spacing between the second slope feature point and each first slope feature point, the Mahalanobis Distance between the second slope feature point and each first slope feature point can be calculated, and the calculated Mahalanobis Distance is used as the first similarity.

[0136] 2412. Based on the first similarity, determine the first slope feature point that the incremental feature point matches.

[0137] In some of these implementations, for each second slope feature point, one first slope feature point with the largest first similarity can be determined as the first slope feature point that the second slope feature point matches.

[0138] After that, perform operation 250.

[0139] 2413. Based on the first steering feature point in the first feature points, determine the first steering feature point sequence.

[0140] In some implementations, the first steering feature point of the type of steering feature point can be determined from the first feature points, and the determined first steering feature points are formed into a sequence based on the timing of entering the current scene to obtain the first steering feature point sequence.

[0141] 2414. Based on the pose attribute information of the incremental feature points and the pose attribute information of the first steering feature points in the first steering feature point sequence, match the incremental feature points with the first steering feature points in the first steering feature point sequence to obtain the second similarity.

[0142] Among them, the second similarity is used to characterize the matching degree between the incremental feature points and the first slope feature points. The higher the value of the second similarity, the higher the matching degree between the two.

[0143] In some of these implementations, for each steering feature point (i.e., the second steering feature point) in the incremental feature points, based on the steering attribute of the second steering feature point (such as whether it steers, or yaw angle information), the steering attributes of each first steering feature point in the first steering feature point sequence, and the distances between the second steering feature point and each first slope feature point, the second steering feature point is matched with each first steering feature point. For example, based on the steering attributes and the feature point spacings between the second steering feature point and each first steering feature point, the Mahalanobis Distance between the second steering feature point and each first steering feature point can be calculated, and the calculated Mahalanobis Distance is used as the second similarity.

[0144] 2415, Based on the second similarity, determine the first steering feature point that the incremental feature point matches.

[0145] In some of these implementations, for each second steering feature point, the first steering feature point with the largest second similarity can be determined as the first steering feature point that the second steering feature point matches.

[0146] After that, perform operation 250.

[0147] 2416, Based on the first global position feature point in the first feature points, determine the first global position feature point sequence.

[0148] In some implementations, the first global position feature point of the type of global position feature point can be determined from the first feature points, and the determined first global position feature points are formed into a sequence based on the timing of entering the current scene to obtain the first global position feature point sequence.

[0149] 2417, Based on the relative position information and global position information in the pose attribute information of the incremental feature points, and the relative position information and global position information in the pose attribute information of the first global position feature points in the first global feature point sequence, determine the similarity between the incremental feature point and the first global position feature points in the first global position feature point sequence to obtain the third similarity.

[0150] Among them, the third similarity is used to characterize the matching degree between the incremental feature point and the first global feature point. The higher the value of the third similarity, the higher the matching degree between the two.

[0151] In some of these implementation manners, the product of the absolute position distance function value and the relative position distance scale function value between the second global feature point and each first global feature point in the incremental feature points (i.e., the second global feature points) can be determined respectively based on the relative position information and global position information of the second global feature points, and the relative position information and global position information in the pose attribute information of each first global feature point in the first global feature point sequence, as the similarity between the two (i.e., the third similarity).

[0152] In some of these implementation manners, a normal distribution function with a mean of 0 can be used as the absolute position distance function, and based on the following formula (1), the absolute position distance function value between the second global feature point and the first global feature point can be determined:

[0153]

[0154] Where d represents the distance determined based on the global position information between the second global feature point and the first global feature point (referred to as: absolute position distance), and σ represents the standard deviation size between the second global feature point and the first global feature point sequence.

[0155] In some of these implementation manners, based on the following formula (2), the relative position distance scale function value between the second global feature point and the first global feature point can be determined:

[0156]

[0157] Where S represents the ratio of the relative distance of the second global feature point to the relative distance of the first global feature point, and the β parameter is a similarity coefficient set in advance for controlling the similarity scale, and the value of β is a value greater than 1.

[0158] Among them, the relative distance of the second global feature point is the position information of the second global feature point in the specified local coordinate system, that is, the distance of the second global feature point relative to the origin of the specified local coordinate system (the origin of the local coordinate system corresponding to the positioning reference point). The relative distance of the first global feature point is the position information of the first global feature point in the specified local coordinate system, that is, the distance of the first global feature point relative to the origin of the specified local coordinate system (the origin of the local coordinate system corresponding to the reference benchmark point). For example, if the specified local coordinate system takes the entrance of the underground parking lot as the coordinate origin, then the ratio of the relative distance of the second global feature point to the relative distance of the first global feature point represents the ratio of the relative distance of the second global feature point and the relative distance of the first global feature point to the entrance of the underground parking lot respectively. Suppose the distance of the second global feature point relative to the entrance of the underground parking lot is 60m, and the distance of the first global feature point relative to the entrance of the underground parking lot is 40m respectively, then the ratio of the relative distance of the second global feature point to the relative distance of the first global feature point is: 60 / 40 = 1.5.

[0159] In this embodiment, for two global position feature points, the product of the absolute position distance function value and the relative position distance scale function value of the two global position feature points is used as the similarity between the two, and the similarity between the two can be determined by integrating the global position and the relative position, which helps to improve the objectivity and accuracy of the similarity determination result.

[0160] 2418. Based on the third similarity, determine the first global position feature point that matches the incremental feature point.

[0161] In some implementation manners, one first global feature point with the largest third similarity to each second global feature point can be determined respectively as the first global feature point that matches the second global feature point.

[0162] After that, perform operation 250.

[0163] Optionally, in some other implementation manners, when the types of the incremental feature points are two or three of the slope feature point, the steering feature point and the global position feature point at the same time, based on the types of the incremental feature points, the above methods can be used respectively to determine the similarity with the first feature points of the corresponding types, and based on the weights corresponding to the types of the pre-set feature points, the similarities corresponding to the corresponding types are weighted and calculated to obtain the comprehensive similarity, and the first feature point that matches the incremental feature point is determined based on the comprehensive similarity.

[0164] Based on this embodiment, since the slope feature points, turning feature points, and global position feature points can be used to reflect different features of the corresponding trajectory points, matching the corresponding type of feature points (slope feature points, turning feature points, global position feature points) of the reference trajectory and the driving trajectory respectively helps to improve the effectiveness and matching efficiency of the feature point matching result, saves the computing resources consumed in matching between different types of feature points, improves the utilization rate of computing resources, and thus reduces the computing power requirements for the computing platform.

[0165] Figure 11 It is a schematic flowchart of registering the reference trajectory and the driving trajectory in an exemplary embodiment of the present disclosure. As Figure 11 shown, on the basis of any of the illustrated embodiments, in this embodiment, operation 250 may include: Figures 2 - 10 In any of the illustrated embodiments, in this embodiment, operation 250 may include:

[0166] 2510, Downsample the first feature point sequence and the second feature point sequence respectively according to a preset sampling method, and correspondingly obtain a first sampling point sequence and a second sampling point sequence.

[0167] In some implementation manners, the first feature point sequence and the second feature point sequence may be downsampled respectively according to a preset sampling frequency, for example, a sampling frequency of sampling one point every five points. The size of the preset sampling frequency may be set according to requirements such as vehicle speed, accuracy of the trajectory matching result, and computing power resources, and may be modified as needed. The embodiments of the present disclosure do not limit the size of the preset sampling frequency.

[0168] In some implementation manners, the first feature point sequence and the second feature point sequence may be downsampled respectively according to a preset Euclidean distance. The size of the preset Euclidean distance may be set according to requirements such as vehicle speed, accuracy of the trajectory matching result, and computing power resources, and may be modified as needed. The embodiments of the present disclosure do not limit the size of the preset Euclidean distance.

[0169] Optionally, in some implementation manners, abnormal feature points in the first feature point sequence and the second feature point sequence may be removed first, for example, feature points with a large difference in Euclidean distance from the previous and subsequent feature points. For example, the Euclidean distance between the previous and subsequent adjacent feature points of a certain feature point is 8, while the Euclidean distance between this feature point and the previous adjacent feature point is 10, and the Euclidean distance between this feature point and the subsequent adjacent feature point is 8. Then this feature point may be a mutation point. First, remove this feature point from the corresponding feature point sequence and remove the corresponding feature point from the other feature point sequence, and then perform downsampling to avoid the influence of abnormal feature points on the trajectory matching result.

[0170] 2511, Register the reference trajectory and the driving trajectory based on the first sampling point sequence and the second sampling point sequence.

[0171] In some of these implementations, any trajectory registration algorithm can be adopted to register the reference trajectory and the driving trajectory. For example, the Umeyama algorithm can be used to register the reference trajectory and the driving trajectory to obtain a registration result. For details, reference can be made to the introduction in the above embodiments, which will not be elaborated here.

[0172] Based on this embodiment, by registering the reference trajectory and the driving trajectory with the first sampling point sequence and the second sampling point sequence obtained by downsampling the first feature point sequence and the second feature point sequence, the number of feature points participating in trajectory matching can be reduced, computing resources can be saved, thereby reducing the computing power requirement for the computing platform, and at the same time, the efficiency of trajectory matching can be improved, and further the positioning efficiency can be enhanced.

[0173] Figure 12 is a schematic flowchart of a vehicle positioning method provided by another exemplary embodiment of the present disclosure. As Figure 12 shown, on the basis of any of the Figures 2 - 11 illustrated embodiments, in this embodiment, it may further include:

[0174] 310. In response to a trigger condition, obtain the pose attribute information of the driving trajectory points corresponding to the trigger condition in the driving trajectory.

[0175] Among them, during the process from responding to the trigger condition to obtaining the vehicle driving trajectory, the vehicle may have passed through multiple driving trajectory points. Then, the driving trajectory points corresponding to the trigger condition are at least one driving trajectory point that the vehicle has passed through since the trigger condition.

[0176] 320. Based on the pose attribute information of the driving trajectory points corresponding to the trigger condition, determine an initial feature point.

[0177] Among them, the initial feature point is a feature point determined from the pose attribute information of at least one driving trajectory point that the vehicle has passed through during the current driving since the trigger condition. The initial feature point may be one or more.

[0178] 330. Based on the initial feature point, determine a positioning reference point.

[0179] In this embodiment, the positioning reference point is the starting point of the vehicle's current driving trajectory that participates in trajectory matching. The second feature points include the positioning reference point and the feature points of the driving trajectory that are chronologically after the positioning reference point, that is, the positioning reference point and the feature points of the driving trajectory that are chronologically after the positioning reference point are the feature points participating in trajectory matching.

[0180] In some of these implementation manners, one of the initial feature points can be determined as the positioning reference point. Specifically, when there is only one initial feature point, this initial feature point can be directly determined as the positioning reference point. When there are multiple initial feature points, in accordance with a preset manner, for example, in accordance with the vehicle driving time sequence, one of the multiple initial feature points with the earliest time sequence can be selected as the positioning reference point, or one of the multiple initial feature points with the middle time sequence can be selected as the positioning reference point, or one of the multiple initial feature points with the latest time sequence can be selected as the positioning reference point. Alternatively, one can also be randomly selected from the multiple initial feature points as the initial feature point, and the embodiments of the present disclosure do not limit this.

[0181] In the embodiments of the present disclosure, since the positioning reference point is determined from the initial feature points, the type of this positioning reference point can be a slope feature point, a steering feature point, or a global position feature point on the driving trajectory with good satellite signals (for example, strong satellite signals in an outdoor scene), and the embodiments of the present disclosure do not limit the type of the positioning reference point.

[0182] 340, determine the reference trajectory point matched with the positioning reference point as the reference benchmark point.

[0183] In the embodiments of the present disclosure, the reference benchmark point is the starting point participating in the trajectory matching in the corresponding reference trajectory.

[0184] In some of these implementation manners, based on the pose attribute information of the positioning reference point and the pose attribute information of the reference trajectory point, a reference trajectory point consistent with this pose attribute information can be determined as the reference benchmark point. For example, based on the global position information in the pose attribute information of the positioning reference point, from the pre-stored scene map data, one reference trajectory point with the same or closest global position information as that of this positioning reference point can be selected as the reference benchmark point.

[0185] Alternatively, in some other implementation manners, multiple reference trajectory points whose distance between the global position information and the global position information of this positioning reference point is within a preset distance range can be selected as candidate reference trajectory points, and based on a similar manner of matching the incremental feature points and the first feature points in the above embodiments, calculate the similarity between this positioning reference point and each candidate reference trajectory point. According to the type of this positioning reference point, this similarity can be at least one of the first similarity, the second similarity, and the third similarity. According to the similarity between this positioning reference point and each candidate reference trajectory point, select the candidate reference trajectory point with the maximum similarity as the reference benchmark point.

[0186] In the embodiments of the present disclosure, since the reference benchmark point is a reference trajectory point that matches the positioning benchmark point, and the type of the positioning benchmark point can be a slope feature point, a steering feature point, or a global position feature point on the driving trajectory with good satellite signals, accordingly, the type of the reference benchmark point can also be a slope feature point, a steering feature point, or a global position feature point on the driving trajectory with good satellite signals. The embodiments of the present disclosure do not limit the type of the reference benchmark point.

[0187] 350. Determine the positioning benchmark point and the feature points in the initial feature points whose time sequence is after the positioning benchmark point as the incremental feature points corresponding to the first positioning window, and execute operations 230-260.

[0188] Correspondingly, in this embodiment, operation 220 may include:

[0189] Based on the reference benchmark point and the pose attribute information of the reference trajectory points in the reference trajectory whose time sequence is after the reference benchmark point, determine the first feature point of the reference trajectory.

[0190] Based on this embodiment, the positioning benchmark point of the vehicle's current driving trajectory and the corresponding reference benchmark point can be determined based on the trigger condition, so as to determine the second feature point of the driving trajectory starting from the positioning benchmark point and the first feature point of the reference trajectory starting from the reference benchmark point, so as to perform trajectory matching and precise positioning based on the second feature point and the first feature point, and improve the effectiveness of the positioning result.

[0191] Figure 13 is a schematic flowchart of a vehicle positioning method provided by another exemplary embodiment of the present disclosure. As Figure 13 shown, on the basis of any of the embodiments shown, in this embodiment, after operation 250, it may further include: Figures 2 - 12

[0192] 410. Obtain the similarity between the registered reference trajectory and the driving trajectory to obtain the fourth similarity.

[0193] Among them, the similarity between the registered reference trajectory and the driving trajectory is used to characterize the matching degree between the registered reference trajectory and the driving trajectory. The larger the calculated similarity value, the higher the matching degree between the registered reference trajectory and the driving trajectory.

[0194] In some implementation manners, the similarity Sim between the registered reference trajectory and the driving trajectory can be obtained in the following manner abs :

[0195] Sim abs = exp(-0.5 * mse) ∈ (0, 1) Formula (3)

[0196] ​Among them, mse represents the mean square error of the Euclidean distance between the feature point pairs corresponding to the reference trajectory and the positioning trajectory after registration.

[0197] 420, determine the relationship between the fourth similarity and the first preset threshold.

[0198] In this operation 420, the relationship between whether the fourth similarity is greater than the first preset threshold can be determined.

[0199] In response to the fourth similarity being greater than the first preset threshold, perform operation 430. Otherwise, in response to the fourth similarity not being greater than the first preset threshold, operation 460 can be selectively performed. The specific value of this first preset threshold can be set according to actual needs and can be updated as required.

[0200] 430, determine that the registered reference trajectory and the driving trajectory match.

[0201] Correspondingly, operation 260 may include:

[0202] 440, determine the matching relationship between the end driving trajectory point and the corresponding reference trajectory point.

[0203] In this operation 440, it can be determined whether the end driving trajectory point matches the corresponding reference trajectory point.

[0204] In response to the end driving trajectory point matching the corresponding reference trajectory point, perform operation 450. Otherwise, in response to the end driving trajectory point not matching the corresponding reference trajectory point, operation 460 can be selectively performed.

[0205] In some implementation manners, based on the global position information in the position attribute information of the end driving trajectory point and the global position information of the corresponding reference trajectory point, the Euclidean distance between the end driving trajectory point and the corresponding reference trajectory point can be calculated. Based on whether this Euclidean distance is greater than a preset distance threshold, it can be determined whether the end driving trajectory point matches the corresponding reference trajectory point. It can be determined that the end driving trajectory point matches the corresponding reference trajectory point when this Euclidean distance is greater than the preset distance threshold; otherwise, if this Euclidean distance is not greater than the preset distance threshold, it is determined that the end driving trajectory point does not match the corresponding reference trajectory point. Regarding the calculation method of the Euclidean distance, relevant technologies can be referred to and will not be introduced here.

[0206] In some other implementations, the Euclidean distance between the end driving trajectory point and the corresponding reference trajectory point can be calculated based on the three-dimensional coordinate information of the end driving trajectory point in the specified local coordinate system (hereinafter referred to as the first three-dimensional coordinate information) and the three-dimensional coordinate information of the corresponding reference trajectory point in the specified local coordinate system (hereinafter referred to as the second three-dimensional coordinate information). Based on whether the Euclidean distance is greater than a preset distance threshold, it is determined whether the end driving trajectory point matches the corresponding reference trajectory point. When the Euclidean distance is greater than the preset distance threshold, it can be determined that the end driving trajectory point matches the corresponding reference trajectory point; otherwise, if the Euclidean distance is not greater than the preset distance threshold, it is determined that the end driving trajectory point does not match the corresponding reference trajectory point.

[0207] In some other implementations, the Euclidean distance between the end driving trajectory point and the corresponding reference trajectory point can be calculated based on the global position information in the position attribute information of the end driving trajectory point and the global position information of the corresponding reference trajectory point to obtain the first Euclidean distance. At the same time, the Euclidean distance between the end driving trajectory point and the corresponding reference trajectory point can be calculated based on the first three-dimensional coordinate information of the end driving trajectory point and the second three-dimensional coordinate information of the corresponding reference trajectory point to obtain the second Euclidean distance; the comprehensive Euclidean distance is determined based on the first Euclidean distance and the second Euclidean distance. For example, weighted calculation is performed on the two based on a preset weight, or the two are summed, or the average value of the two is calculated as the comprehensive Euclidean distance; based on whether the comprehensive Euclidean distance is greater than the preset distance threshold, it is determined whether the end driving trajectory point matches the corresponding reference trajectory point. When the comprehensive Euclidean distance is greater than the preset distance threshold, it can be determined that the end driving trajectory point matches the corresponding reference trajectory point; otherwise, if the comprehensive Euclidean distance is not greater than the preset distance threshold, it is determined that the end driving trajectory point does not match the corresponding reference trajectory point.

[0208] 450, determine that the reference trajectory point corresponding to the end driving trajectory point is the current positioning result of the vehicle.

[0209] After determining that the corresponding reference trajectory point is the current positioning result of the vehicle, the current position of the vehicle can be displayed at the reference trajectory point on the corresponding scene map, so as to display the visual positioning result of the current position of the vehicle for the user.

[0210] After that, for the next positioning window, operations 230 and subsequent processes can be iteratively executed.

[0211] 460, determine the current position yaw of the vehicle.

[0212] If it is determined that the current position of the vehicle is yawed, a prompt message indicating that the current position of the vehicle has been yawed can be output to the user. At the same time, the position obtained by dead reckoning by the odometer module on the vehicle can be displayed on the reference map as the current position of the vehicle.

[0213] Based on this embodiment, when the registered reference trajectory and the driving trajectory match, and the end driving trajectory points match the corresponding reference trajectory points, the corresponding reference trajectory points can be used as the current positioning result of the vehicle, so as to achieve effective positioning of the vehicle's position; when the registered reference trajectory and the driving trajectory do not match and / or it is determined that the end driving trajectory points do not match the corresponding reference trajectory points, it is determined that the vehicle's current position is yawed, that is, it deviates from the reference trajectory.

[0214] Figure 14 It is a schematic flowchart of a vehicle positioning method provided by another exemplary embodiment of the present disclosure. As Figure 14 shown, on the basis of the embodiment shown in Figure 13 , in this embodiment, after operation 410, it may further include:

[0215] 510. According to a preset adjustment method, adjust the positioning reference point based on the initial feature points, and adjust the second feature point sequence based on the adjusted positioning reference point.

[0216] In some implementation manners, the positioning reference point can be adjusted according to a preset candidate offset amount.

[0217] In a specific example, the candidate offset amount can be a set of serial number parameters preset according to empirical values for offsetting the feature points serving as the positioning reference point. For example, [-5, -3, -1, 0, 1, 3, 5], where each serial number parameter is used to represent the offset serial number relative to the initially determined positioning reference point, and the feature point corresponding to the offset serial number is used as the new positioning reference point. The initially determined positioning reference point is the positioning reference point determined through the above operation 330. The serial number parameter "0" indicates that the offset serial number relative to the initially determined positioning reference point is 0, that is, the initially determined positioning reference point is not adjusted. Among the serial number parameters, "-" indicates forward offset adjustment relative to the initially determined positioning reference point according to the time sequence, and vice versa, indicating backward offset adjustment relative to the initially determined positioning reference point according to the time sequence. For example, the offset serial number "-3" means that among the feature points corresponding to the driving trajectory, the initially determined positioning reference point is adjusted forward by 3 serial numbers (i.e., the feature point with serial number -3) according to the time sequence as the new positioning reference point, so it is determined that the 3rd feature point before the initially determined positioning reference point among the feature points corresponding to the driving trajectory is the adjusted positioning reference point; the offset serial number "5" means that among the feature points corresponding to the driving trajectory, the initially determined positioning reference point is adjusted backward by 5 serial numbers (i.e., the feature point with serial number 5) according to the time sequence as the new positioning reference point, so it is determined that the 5th feature point after the initially determined positioning reference point among the feature points corresponding to the driving trajectory is the adjusted positioning reference point.

[0218] In another specific example, the candidate offset can be a set of distance parameters preset according to empirical values. Each distance parameter in the set can be set based on the possible distance deviation between the positioning reference point and the reference reference point in actual applications. For example, the candidate offset can be [-12, -9, -5, 0, 5, 9, 12], and the unit can be m. Among them, each distance parameter is used to represent the offset distance relative to the positioning reference point determined for the first time. Among the feature points corresponding to the driving trajectory, the feature points corresponding to the adjustment of the positioning reference point determined for the first time according to this offset distance are used as the new positioning reference point. The distance parameter "0" means that the offset distance relative to the positioning reference point determined for the first time is 0, that is, the positioning reference point determined for the first time is not adjusted. Among the distance parameters, "-" means forward offset adjustment relative to the positioning reference point determined for the first time according to the time sequence, and vice versa, it means backward offset adjustment relative to the positioning reference point determined for the first time according to the time sequence. For example, the distance parameter "5" means that among the feature points corresponding to the driving trajectory, the feature point whose Euclidean distance from the positioning reference point determined for the first time is 5 meters after the time sequence of the positioning reference point determined for the first time is the adjusted positioning reference point.

[0219] After that, a second feature point sequence is formed by the adjusted positioning reference point and the feature points in the driving trajectory whose time sequence is after the adjusted positioning reference point. As Figure 15 shown, it is an application schematic diagram of adjusting the positioning reference point in the embodiments of the present disclosure.

[0220] 520. Downsample the adjusted second feature point sequence to obtain a third sampling point sequence.

[0221] 530. Based on the second sampling point sequence and the third sampling point sequence, register the reference trajectory and the driving trajectory.

[0222] 540. Obtain the similarity between the registered reference trajectory and the driving trajectory to obtain a fifth similarity.

[0223] 550. Determine the relationship between the excess value of the fifth similarity relative to the fourth similarity and the second preset threshold.

[0224] In this operation 550, it can be determined whether the excess value of the fifth similarity relative to the fourth similarity is greater than the second preset threshold.

[0225] In response to the excess value of the fifth similarity relative to the fourth similarity being greater than the second preset threshold, that is, the similarity between the registered reference trajectory and the driving trajectory changes significantly, indicating that there is a large deviation in the time series of the feature point sequences participating in the matching between the reference trajectory and the driving trajectory. Iteratively execute operations 510 - 550 until the excess value of the fifth similarity obtained relative to the fifth similarity obtained in the previous adjacent iteration is not greater than the second preset threshold. Use the fifth similarity obtained in the last iteration as the fourth similarity and execute operation 420.

[0226] In response to the excess value of the fifth similarity relative to the fourth similarity being not greater than the second preset threshold, directly execute operation 420.

[0227] In some implementation manners, an element value in the candidate offset amount (for example, offset serial number 0 or distance parameter 0) can be preset as a default value (referred to as the initial default value). For each positioning window, based on the first determined positioning reference point and the default value in the candidate offset amount, the second feature point sequence participating in the trajectory matching for the first time can be determined. For example, when the initial default value is offset serial number 0, it means using the first determined positioning reference point as the positioning reference point, and forming the second feature point sequence with this positioning reference point and the feature points in the driving trajectory whose time series is after this positioning reference point. After that, each time operation 510 in operations 510 - 550 is iteratively executed, based on the element value corresponding to the initial default value, an element value can be sequentially selected from the candidate offset amount in a sequential or reverse order, or first sequential and then reverse, or first reverse and then sequential manner. Based on the selected element value, determine the current positioning reference point, and form the second feature point sequence with this positioning reference point and the feature points in the driving trajectory whose time series is after this positioning reference point, and execute operations 520 - 550.

[0228] In other implementation manners, for any positioning window, when the excess value of the fifth similarity obtained in a certain iteration (denoted as the mth iteration, where m is an integer greater than 0) of executing operations 510 - 550 relative to the fifth similarity obtained in the previous adjacent iteration is not greater than the second preset threshold, update the initial default value used to determine the positioning reference point in the subsequent positioning window (that is, the positioning window after this any positioning window) to the element value used to determine the positioning reference point when operation 510 is executed in the mth iteration. For example, for the current positioning window, when the element value selected from the candidate offset amount when operation 510 is executed in the mth iteration is offset serial number 3, and the excess value of the fifth similarity obtained relative to the fifth similarity obtained in the previous adjacent iteration is not greater than the second preset threshold, then in the subsequent positioning window, use the third feature point in the driving trajectory corresponding to the feature points whose time series is after the above first determined positioning reference point (offset serial number 0) as the positioning reference point for the second feature point sequence participating in the trajectory matching for the first time in the subsequent positioning window.

[0229] In some other implementation manners, if the exceeded value of the fifth similarity relative to the fourth similarity obtained by adjusting the second feature point sequence through operation 510 for the first time is not greater than the second preset threshold, then in subsequent positioning windows, the initial default value is still used as the default value for determining the positioning reference point for the first time.

[0230] Based on this embodiment, the positioning reference point can be dynamically adjusted. When the exceeded value of the similarity between the reference trajectory and the driving trajectory after the adjustment of the positioning reference point relative to the similarity before the adjustment is greater than the second preset threshold, the positioning reference point is optimized and adjusted to align the reference trajectory and the driving trajectory in time sequence, which helps to improve the accuracy of the matching result between the reference trajectory and the driving trajectory, thereby improving the accuracy of the vehicle positioning result.

[0231] Figure 16 It is a schematic flowchart of a vehicle positioning method provided by still another exemplary embodiment of the present disclosure. As Figure 16 shown, on the basis of the embodiment shown in Figure 13 in this embodiment, after operation 410, the following steps may further be included:

[0232] 610. Adjust the reference reference point according to a preset adjustment method, and adjust the first feature point sequence based on the adjusted reference reference point.

[0233] In some of the implementation manners, the reference reference point can be adjusted according to a preset candidate offset amount, and the specific implementation manner can refer to operation 510, which will not be elaborated here.

[0234] 620. Downsample the adjusted first feature point sequence to obtain a fourth sampling point sequence.

[0235] 630. Register the reference trajectory and the driving trajectory based on the fourth sampling point sequence and the second sampling point sequence.

[0236] 640. Obtain the similarity between the registered reference trajectory and the driving trajectory to obtain a sixth similarity.

[0237] 650. Determine the relationship between the exceeded value of the sixth similarity relative to the fourth similarity and the second preset threshold.

[0238] In this operation 650, it can be determined whether the exceeded value of the sixth similarity relative to the fourth similarity is greater than the second preset threshold.

[0239] In response to the fact that the excess value of the sixth similarity relative to the fourth similarity is greater than the second preset threshold, that is, the similarity between the reference trajectory and the driving trajectory after alignment changes greatly, indicating that there is a large deviation in the timing of the feature point sequence involved in the matching between the reference trajectory and the driving trajectory, operations 610-650 are iteratively performed until the excess value of the obtained sixth similarity relative to the sixth similarity obtained the previous time is not greater than the second preset threshold, the sixth similarity obtained for the last time is taken as the fourth similarity, and operation 420 is performed.

[0240] In response to the excess value of the sixth similarity relative to the fourth similarity being not greater than the second preset threshold, operation 420 is directly performed.

[0241] The specific implementation of operation 650 may refer to operation 550 and will not be described in detail here.

[0242] Based on this embodiment, the reference reference point can be dynamically adjusted. When the similarity between the reference trajectory and the driving trajectory after the reference reference point is adjusted exceeds the similarity obtained last time by more than a second preset threshold, the reference positioning reference point is optimized and adjusted to align the reference trajectory and the driving trajectory in time sequence, which helps to improve the accuracy of the matching results between the reference trajectory and the driving trajectory, thereby improving the accuracy of the vehicle positioning results.

[0243] Figure 17 FIG. 1 is a flow chart of obtaining a reference trajectory in an exemplary embodiment of the present disclosure. Figure 17 As shown, in Figures 2 - 16 Based on any of the illustrated embodiments, in this embodiment, operation 210 may include:

[0244] 2110. In response to a trigger condition, obtain the position and posture attribute information of the current driving trajectory point of the vehicle.

[0245] 2111, determine a target trajectory point corresponding to posture attribute information corresponding to at least one scene map stored in the vehicle, which is consistent with the posture attribute information of the current driving trajectory point of the vehicle.

[0246] Among them, the target trajectory point is a trajectory point corresponding to the posture attribute information determined from the posture attribute information corresponding to at least one scene map stored in the vehicle and consistent with the posture attribute information of the current driving trajectory point of the vehicle.

[0247] In some of these implementation manners, the pose attribute information of the currently located driving trajectory point can be compared with the pose attribute information corresponding to at least one stored scene map, and the trajectory point whose pose attribute information in the at least one stored scene map is consistent with the pose attribute information of the currently located driving trajectory point is determined as the target trajectory point. The pose attribute information of the target trajectory point is consistent with the pose attribute information of the currently located driving trajectory point. For example, it may be that the global position information of the target trajectory point is the same as the global position information of the currently located driving trajectory point, or the global position information of the target trajectory point in the at least one stored scene map is the closest to the global position information of the currently located driving trajectory point, or the Euclidean distance between the global position information of the target trajectory point in the at least one stored scene map and the global position information of the currently located driving trajectory point is less than a preset distance threshold, etc. The embodiments of the present disclosure do not limit this.

[0248] 2112, determine that the trajectory where the target trajectory point is located is the reference trajectory.

[0249] Based on this embodiment, based on a trigger condition, for example, when receiving a positioning instruction sent by a user or detecting that the vehicle enters a scene with poor satellite distribution conditions or few visible stars, the reference trajectory can be determined from at least one scene map stored in the vehicle based on the pose attribute information of the currently located driving trajectory point of the vehicle, so as to position the vehicle based on the reference trajectory.

[0250] Figure 18 It is a schematic flowchart of a vehicle positioning method provided by another exemplary embodiment of the present disclosure. As Figure 18 shown, on the basis of any of the Figures 13 - 17 illustrated embodiments, in this embodiment, in response to the reference trajectory obtained based on operation 210 being one, after obtaining the fourth similarity between the registered reference trajectory and the driving trajectory through operation 410, operation 420 can be directly executed. In response to the reference trajectories obtained based on operation 210 being multiple, for each reference trajectory, operation 220 can be started until the fourth similarity between each registered reference trajectory and the driving trajectory is obtained through operation 410. Correspondingly, in this embodiment, after obtaining the fourth similarity between the registered reference trajectory and the driving trajectory through operation 410, it may further include:

[0251] 710, suppress each registered reference trajectory based on a preset suppression exit condition according to the fourth similarity corresponding to each registered reference trajectory, and obtain the matchable trajectory.

[0252] For example, in some of these implementation manners, the reference trajectory with the fourth similarity not greater than a first preset threshold (also referred to as the absolute threshold score_thre) can be suppressed to obtain the matchable trajectory.

[0253] After that, for the matchable trajectories, i.e., replacing the reference trajectories in the above embodiments with the matchable trajectories, operation 420 is performed. If there are multiple matchable trajectories, operation 420 can be performed for each matchable trajectory separately.

[0254] In the embodiments of the present disclosure, the matchable trajectories determined for each positioning window can be used as the reference trajectories corresponding to the next positioning window and participate in the positioning process of the next positioning window.

[0255] Based on this embodiment, for any positioning window, when there are multiple reference trajectories, the registered reference trajectories can be suppressed through a preset suppression exit condition, the reference trajectories with lower matching degrees can be exited, and better matchable trajectories can be selected, so as to save the computing resources required for trajectory matching based on all reference trajectories, thereby reducing the computing power requirements for the computing platform, improving the trajectory matching efficiency, and further improving the positioning efficiency.

[0256] Figure 19 It is a schematic flowchart of suppressing the reference trajectories in an exemplary embodiment of the present disclosure. As Figure 19 shown, on the basis of the embodiment shown in Figure 18 in some implementation manners, operation 710 may include:

[0257] 7110, determining the normalized similarity corresponding to each registered reference trajectory based on the fourth similarity corresponding to each registered reference trajectory.

[0258] Among them, the fourth similarity corresponding to each registered reference trajectory i can be called the absolute similarity and can be obtained through the following formula (4).

[0259] The normalized similarity corresponding to each registered reference trajectory can be called the relative similarity.

[0260]

[0261] In some implementation manners, the normalized similarity corresponding to each registered reference trajectory i can be obtained in the following manner:

[0262]

[0263] Among them, the value of k is [0, n], n is the number of reference trajectories, and the value of n is an integer greater than 1.

[0264] 7111, suppressing each registered reference trajectory based on the first relative threshold according to the normalized similarity corresponding to each registered reference trajectory, to obtain at least one candidate trajectory.

[0265] Among them, the specific value of the first relative threshold can be set according to actual needs and can be updated as required.

[0266] In response to there being one candidate trajectory, perform operation 7112; in response to there being multiple candidate trajectories, perform operation 7113.

[0267] 7112. Determine that this candidate trajectory is a matchable trajectory.

[0268] 7113. Based on the normalized similarity and the fourth similarity corresponding to each candidate trajectory, suppress the multiple candidate trajectories based on the second relative threshold and the absolute threshold to obtain matchable trajectories.

[0269] Among them, the specific values of the second relative threshold and the absolute threshold can be set according to actual needs and can be updated as required.

[0270] In some implementation manners, the second relative threshold is greater than the first relative threshold.

[0271] For example, in some implementation manners, the second relative threshold is greater than the first relative threshold. The first relative threshold is denoted as lower_thre, the second relative threshold is denoted as higher_thre, and the absolute threshold is denoted as score_thre. In operations 7111 - 7112, reference trajectories with a normalized similarity not greater than the first relative threshold lower_thre can be suppressed. Specifically, the normalized similarity corresponding to each registered reference trajectory i can be determined respectively whether it is greater than the first relative threshold lower_thre, suppress the reference trajectories with a normalized similarity not greater than the first relative threshold lower_thre (i.e., exit), and use the reference trajectories with a normalized similarity greater than the first relative threshold lower_thre as candidate trajectories.

[0272] In operation 7113, candidate trajectories with a normalized similarity not greater than the second relative threshold higher_thre and a fourth similarity not greater than the absolute threshold score_thre can be suppressed. Specifically, in an optional example, the normalized similarity corresponding to each candidate trajectory can be determined respectively whether it is greater than the second relative threshold higher_thre, and use the candidate trajectories with a normalized similarity greater than the second relative threshold higher_thre as matchable trajectories; then, determine the remaining candidate trajectories respectively (i.e., the normalized similarity whether the fourth similarity corresponding to the candidate trajectory not greater than the second relative threshold higher_thre is greater than the absolute threshold score_thre, suppress the candidate trajectories with the fourth similarity not greater than the absolute threshold score_thre among the remaining candidate trajectories, and use the candidate trajectories with the fourth similarity greater than the absolute threshold score_thre among the remaining candidate trajectories as the matchable trajectories.

[0273] In another alternative example, the normalized similarity corresponding to each candidate trajectory can be determined separately whether it is greater than the second relative threshold higher_thre, and whether the fourth similarity is greater than the absolute threshold score_thre, and suppress the candidate trajectories with the normalized similarity not greater than the second relative threshold higher_thre and the fourth similarity not greater than the absolute threshold score_thre, and use the remaining candidate trajectories as the matchable trajectories.

[0274] In yet another alternative example, it can be determined separately whether the fourth similarity corresponding to each candidate trajectory is greater than the absolute threshold score_thre, and use the candidate trajectories with the fourth similarity greater than the absolute threshold score_thre as the matchable trajectories; then, determine separately the normalized similarity corresponding to the remaining candidate trajectories (i.e., the candidate trajectories with the fourth similarity not greater than the absolute threshold score_thre) whether it is greater than the second relative threshold higher_thre, and suppress the candidate trajectories with the normalized similarity not greater than the second relative threshold higher_thre among the remaining candidate trajectories, and use the candidate trajectories with the normalized similarity greater than the second relative threshold higher_thre among the remaining candidate trajectories as the matchable trajectories.

[0275] Figure 20 It is an application schematic diagram for suppressing the reference trajectory in an embodiment of the present disclosure. Figure 20 An example is shown in which the reference trajectories obtained based on operation 210 include three reference trajectories, namely reference trajectory A, reference trajectory B, and reference trajectory C, and their corresponding normalized similarities are normalized similarity A, normalized similarity B, and normalized similarity C, respectively. Based on Figure 19 the illustrated embodiment, trajectory suppression is performed to finally obtain the matchable trajectories.

[0276] Based on this embodiment, for any positioning window, when there are multiple reference trajectories, the reference trajectories with relatively low matching degrees can be suppressed based on the fourth similarity corresponding to the reference trajectories or the normalized similarity determined therefrom, and the matchable trajectories participating in trajectory matching within the current positioning window can be accurately determined, thereby improving the accuracy of the determination result of the matchable trajectories and further improving the accuracy of the vehicle positioning result.

[0277] Optionally, in some implementation manners, in response to there being multiple matchable trajectories obtained through operation 710, operation 260 is respectively executed for each matchable trajectory, that is, each matchable trajectory is respectively used as the registered reference trajectory, and the reference trajectory point in the matchable trajectory corresponding to the end driving trajectory point in the registered driving trajectory is determined as the current positioning result of the vehicle.

[0278] Correspondingly, after determining that the reference trajectory point in the matchable trajectory corresponding to the end driving trajectory point in the registered driving trajectory is the current positioning result of the vehicle, it may further include: updating the reference trajectory corresponding to the next positioning window to the matchable trajectory where the reference trajectory point corresponding to the end driving trajectory point is located, and iteratively executing operation 230 - operation 250.

[0279] Based on this embodiment, when there are multiple matchable trajectories obtained after suppression, these multiple matchable trajectories can participate in the positioning of the vehicle in the next positioning window. As the vehicle travels (i.e., the driving trajectory grows), the matchable trajectories participating in subsequent trajectory matching and positioning are dynamically updated based on the change in the matching degree between the matchable trajectories and the positioning trajectory in each positioning window, thereby achieving precise positioning of the vehicle while saving computing resources.

[0280] Exemplary device

[0281] Figure 21 is a structural block diagram of a vehicle positioning device provided by an exemplary embodiment of the present disclosure. The vehicle positioning device of this embodiment can be used to implement the above-mentioned positioning method embodiments of the present disclosure. As Figure 21 shown, the vehicle positioning device of this embodiment includes: a first acquisition module 810, a first determination module 820, a second determination module 830, a matching module 840, a registration module 850, and a positioning module 860. Among them:

[0282] The first acquisition module 810 is configured to acquire a reference trajectory in the scene map of the scene where the vehicle is currently located in response to a trigger condition.

[0283] The first determination module 820 is configured to determine a first feature point of the reference trajectory based on the pose attribute information of the reference trajectory points in the reference trajectory.

[0284] A second determination module 830, configured to determine incremental feature points based on the pose attribute information of the incremental driving trajectory points located within the current positioning window in the current driving trajectory of the vehicle.

[0285] A matching module 840, configured to match the incremental feature points with the first feature points to obtain incremental feature point pairs, where each incremental feature point pair includes a matched incremental feature point and a first feature point.

[0286] A registration module 850, configured to register the reference trajectory and the driving trajectory based on the first feature point sequence formed by the first feature points in multiple groups of feature point pairs in chronological order, and the second feature point sequence formed by the second feature points in multiple groups of feature point pairs in chronological order; where the multiple groups of feature point pairs at least include some of the feature point pairs obtained by matching the first feature points of the reference trajectory with the second feature points of the driving trajectory, and the multiple groups of feature point pairs at least include some of the incremental feature point pairs.

[0287] A positioning module 860, configured to, in response to the matching of the registered reference trajectory and the driving trajectory, determine the current positioning result of the vehicle based on the reference trajectory point corresponding to the end driving trajectory point in the registered driving trajectory.

[0288] In some of these implementation manners, the first determination module 820 is specifically configured to: for the reference trajectory points in the reference trajectory, in a sliding window manner, based on the pitch angle information in the pose attribute information of the reference trajectory points within the sliding window, determine the first reference trajectory point and the second reference trajectory point in the reference trajectory that respectively correspond to the ramp entry point and the ramp exit point of the ramp; based on the first reference trajectory point and the second reference trajectory point, determine a first slope feature point, and the first feature point includes the first slope feature point. And / or, the first determination module 820 is specifically configured to: based on the yaw angle information in the pose attribute information of the reference trajectory points, determine that the third reference trajectory point corresponding to the vehicle steering in the reference trajectory is the first steering feature point, and the first feature point includes the first steering feature point. And / or, the first determination module 820 is specifically configured to: respectively for the reference trajectory points within each positioning window, based on the satellite signal information in the pose attribute information of the reference trajectory points and the relationship between the first change amount and the second change amount, determine the first valid trajectory points; wherein, the first change amount is the change amount of the relative position information in the pose attribute information of the reference trajectory point relative to the relative position information in the pose attribute information of the previous adjacent reference trajectory point, and the second change amount is the change amount of the global position information in the pose attribute information of the reference trajectory point relative to the global position information in the pose attribute information of the previous adjacent reference trajectory point; the relative position information is the position information in a specified local coordinate system, and the global position information is the position information in the global position coordinate system; respectively cluster the first valid trajectory points corresponding to each positioning window to obtain the first clustering result corresponding to each positioning window; respectively determine the center of the first clustering result corresponding to each positioning window as the first global position feature point, and the first feature point includes the first global position feature point.

[0289] In some of these implementations, the second determination module 830 is specifically configured to: for the incremental driving trajectory points and adjacent historical driving trajectory points, in a sliding window manner, based on the pitch angle information in the pose attribute information of the driving trajectory points within the sliding window, determine the first driving trajectory point and the second driving trajectory point corresponding to the ramp entry point and the ramp exit point in the driving trajectory respectively, where the adjacent historical driving trajectory points include at least one driving trajectory point in the driving trajectory whose time sequence is before the incremental driving trajectory point; based on the first driving trajectory point and the second driving trajectory point, determine the second slope feature point, and the second feature point and the incremental feature point include the second slope feature point. And / or, the second determination module 830 is specifically configured to: based on the yaw angle information in the pose attribute information of the incremental driving trajectory points and adjacent historical driving trajectory points, determine the third driving trajectory point corresponding to the vehicle steering in the driving trajectory as the second steering feature point, and the second feature point and the incremental feature point include the second steering feature point, where the adjacent historical driving trajectory points include at least one driving trajectory point in the driving trajectory whose time sequence is before the incremental driving trajectory point. And / or, the second determination module 830 is specifically configured to: based on the satellite signal information in the pose attribute information of the incremental driving trajectory points and the relationship between the third change amount and the fourth change amount, determine the second valid trajectory point; where the third change amount is the change amount of the relative position information in the pose attribute information of the incremental driving trajectory point relative to the relative position information in the pose attribute information of the previous adjacent driving trajectory point, and the fourth change amount is the change amount of the global position information in the pose attribute information of the incremental driving trajectory point relative to the global position information in the pose attribute information of the previous adjacent driving trajectory point; cluster the second valid trajectory points to obtain a second clustering result; determine the center of the second clustering result as the second global position feature point, and the second feature point and the incremental feature point include the second global position feature point.

[0290] In some of these implementations, the types of the first feature point, the second feature point, and the incremental feature point respectively include at least one of a slope feature point, a turning feature point, and a global position feature point. Accordingly, the registration module 850 is specifically configured to: in response to the type of the incremental feature point being a slope feature point, determine a first slope feature point sequence based on the first slope feature points in the first feature points; based on the pose attribute information of the incremental feature point and the pose attribute information of the first slope feature points in the first slope feature point sequence, match the incremental feature point with the first slope feature points in the first slope feature point sequence to obtain a first similarity; the first slope feature point sequence is a sequence formed by the first slope feature points in the first feature points based on the timing of entering the scene; based on the first similarity, determine the first slope feature point that the incremental feature point matches. In response to the type of the incremental feature point being a turning feature point, determine a first turning feature point sequence based on the first turning feature points in the first feature points; based on the pose attribute information of the incremental feature point and the pose attribute information of the first turning feature points in the first turning feature point sequence, match the incremental feature point with the first turning feature points in the first turning feature point sequence to obtain a second similarity; the first turning feature point sequence is a sequence formed by the first turning feature points in the first feature points based on the timing of entering the scene; based on the second similarity, determine the first turning feature point that the incremental feature point matches. In response to the type of the incremental feature point being a global position feature point, determine a first global position feature point sequence based on the first global position feature points in the first feature points; based on the relative position information and global position information in the pose attribute information of the incremental feature point and the relative position information and global position information in the pose attribute information of the first global position feature points in the first global feature point sequence, determine the similarity between the incremental feature point and the first global position feature points in the first global position feature point sequence to obtain a third similarity; wherein, the first global position feature point sequence is a sequence formed by the first global position feature points in the first feature points based on the timing of entering the scene; based on the third similarity, determine the first global position feature point that the incremental feature point matches.

[0291] Figure 22 is a structural block diagram of a vehicle positioning device provided by another exemplary embodiment of the present disclosure. As Figure 22 shown, in Figure 21Based on the illustrated embodiments, in the embodiments of the present disclosure, the vehicle positioning device may further include: a second acquisition module 910, a third determination module 920, and a fourth determination module 930. Among them, the second acquisition module 910 is configured to acquire the similarity between the registered reference trajectory and the driving trajectory, and obtain a fourth similarity. The third determination module 920 is configured to determine the relationship between the fourth similarity and a first preset threshold. The fourth determination module 930 is configured to, in response to the fourth similarity being greater than the first preset threshold, determine that the registered reference trajectory and the driving trajectory match. Correspondingly, in this embodiment, the positioning module 860 is specifically configured to: determine the matching relationship between the end driving trajectory point and the corresponding reference trajectory point; and in response to the end driving trajectory point matching the corresponding reference trajectory point, determine the corresponding reference trajectory point as the current positioning result of the vehicle.

[0292] Optionally, in the vehicle positioning device provided in another exemplary embodiment of the present disclosure, the positioning module 860 may further be configured to: in response to the fourth similarity not being greater than the first preset threshold or the end driving trajectory point not matching the corresponding reference trajectory point, determine that the current position of the vehicle is yawed.

[0293] In some implementation manners, the registration module 850 includes: a sampling unit 8510 configured to downsample the first feature point sequence and the second feature point sequence respectively according to a preset sampling method, and correspondingly obtain a first sampling point sequence and a second sampling point sequence; and a registration unit 8511 configured to register the reference trajectory and the driving trajectory based on the first sampling point sequence and the second sampling point sequence.

[0294] Optionally, referring again to Figure 22 , in the vehicle positioning device provided in another exemplary embodiment of the present disclosure, it may further selectively include: a third acquisition module 940, a fifth determination module 950, and a sixth determination module 960. Among them, the third acquisition module 940 is configured to, in response to a trigger condition, acquire the pose attribute information of the driving trajectory point corresponding to the trigger condition in the driving trajectory. The second determination module 830 is further configured to determine an initial feature point based on the pose attribute information of the driving trajectory point corresponding to the trigger condition. The fifth determination module 950 is configured to determine a positioning reference point based on the initial feature point, and the second feature point includes the positioning reference point and the feature points of the driving trajectory whose time sequence is after the positioning reference point. The sixth determination module 960 is configured to determine the reference trajectory point that matches the positioning reference point as the reference benchmark point. The second determination module 830 is further configured to determine the positioning reference point and the feature points of the initial feature point whose time sequence is after the positioning reference point as the incremental feature points corresponding to the first positioning window. The first determination module 820 is specifically configured to: determine the first feature point of the reference trajectory based on the reference benchmark point and the pose attribute information of the reference trajectory points in the reference trajectory whose time sequence is after the reference benchmark point.

[0295] Optionally, referring again to Figure 22 , in the vehicle positioning device provided in another exemplary embodiment of the present disclosure, it may also optionally include: an adjustment module 970 and a seventh determination module 980. In some implementation manners, the adjustment module 970 is configured to adjust the positioning reference point determined by the fifth determination module 950 based on the initial feature points according to a preset adjustment manner, and adjust the second feature point sequence based on the adjusted positioning reference point. Accordingly, the sampling unit 8510 is further configured to downsample the adjusted second feature point sequence to obtain a third sampled point sequence; the registration unit 8511 is further configured to register the reference trajectory and the driving trajectory based on the first sampled point sequence and the third sampled point sequence; the second acquisition module 910 is further configured to acquire the similarity between the registered reference trajectory and the driving trajectory to obtain a fifth similarity. The seventh determination module 980 is configured to determine the relationship between the exceeding value of the fifth similarity relative to the fourth similarity and a second preset threshold. In response to the exceeding value of the fifth similarity relative to the fourth similarity being greater than the second preset threshold, it instructs the adjustment module 970 to iteratively execute the operation of adjusting the positioning reference point based on the initial feature points according to the preset adjustment manner and adjusting the second feature point sequence based on the adjusted positioning reference point until the exceeding value of the obtained fifth similarity relative to the fifth similarity obtained in the previous adjacent time is not greater than the second preset threshold, and use the finally obtained fifth similarity as the fourth similarity, and instruct the third determination module 920 to execute the operation of determining whether the fourth similarity is greater than the first preset threshold; in response to the exceeding value of the fifth similarity relative to the fourth similarity not being greater than the second preset threshold, it instructs the third determination module 920 to execute the operation of determining whether the fourth similarity is greater than the first preset threshold.

[0296] Alternatively, in some other implementations (not shown in the figures), an adjustment module 970 is configured to adjust a reference fiducial point according to a preset adjustment method, and adjust the first feature point sequence based on the adjusted reference fiducial point. Accordingly, a sampling unit 8510 is further configured to downsample the adjusted first feature point sequence to obtain a fourth sampling point sequence; a registration unit 8511 is further configured to register the reference trajectory and the driving trajectory based on the fourth sampling point sequence and the second sampling point sequence; a second acquisition module 910 is further configured to obtain a similarity between the registered reference trajectory and the driving trajectory to obtain a sixth similarity. A seventh determination module 980 is configured to determine the relationship between the exceeded value of the sixth similarity relative to the fourth similarity and a second preset threshold; in response to the exceeded value of the sixth similarity relative to the fourth similarity being greater than the second preset threshold, instruct the adjustment module 970 to iteratively execute the operation of adjusting the reference fiducial point according to the preset adjustment method and adjusting the first feature point sequence based on the adjusted reference fiducial point until the exceeded value of the obtained sixth similarity relative to the sixth similarity obtained in the previous adjacent time is not greater than the second preset threshold, use the finally obtained sixth similarity as the fourth similarity, and instruct a third determination module 920 to execute the operation of determining whether the fourth similarity is greater than a first preset threshold; in response to the exceeded value of the sixth similarity relative to the fourth similarity not being greater than the second preset threshold, instruct the third determination module 920 to execute the operation of determining whether the fourth similarity is greater than the first preset threshold.

[0297] Optionally, referring again to Figure 22 , in some of these implementations, the first acquisition module 810 may include: a first acquisition unit 8110, a first determination unit 8111, and a second determination unit 8112. Among them, the first acquisition unit 8110 is configured to acquire pose attribute information of a driving trajectory point where the vehicle is currently located in response to a trigger condition; the first determination unit 8111 is configured to determine a target trajectory point corresponding to the pose attribute information that is consistent with the pose attribute information of the driving trajectory point where the vehicle is currently located among the pose attribute information corresponding to at least one scene map stored in the vehicle; the second determination unit 8112 is configured to determine that the trajectory where the target trajectory point is located is the reference trajectory.

[0298] Optionally, in the vehicle positioning device provided in another exemplary embodiment of the present disclosure, the first determination module 820 is specifically configured to: in response to there being multiple reference trajectories, respectively execute, for each reference trajectory, an operation of determining a first feature point of the reference trajectory based on the pose attribute information of the reference trajectory points in the reference trajectory when creating the scene map. Accordingly, referring again to Figure 22, the vehicle positioning device in this embodiment may further include: an inhibition module 990, configured to inhibit each registered reference trajectory based on a preset inhibition exit condition according to the fourth similarity corresponding to each registered reference trajectory obtained by the second acquisition module 910, so as to obtain a matchable trajectory. Correspondingly, the third determination module 920 is specifically configured to perform an operation of determining whether the fourth similarity is greater than a first preset threshold for the matchable trajectory.

[0299] In some implementation manners, the third determination module 920 is specifically configured to: in response to there being multiple matchable trajectories, respectively perform an operation of determining whether the fourth similarity is greater than a first preset threshold for each matchable trajectory. Correspondingly, see again Figure 22 , the vehicle positioning device in this embodiment may further include: an update module 900, configured to update the reference trajectory corresponding to the next positioning window to the matchable trajectory where the reference trajectory point corresponding to the end driving trajectory point is located after the positioning module 860 determines that the corresponding reference trajectory point is the current positioning result of the vehicle.

[0300] In some implementation manners, the inhibition module 990 is specifically configured to: determine the normalized similarity corresponding to each registered reference trajectory based on the fourth similarity corresponding to each registered reference trajectory; inhibit each registered reference trajectory based on a first relative threshold according to the normalized similarity corresponding to each registered reference trajectory, so as to obtain at least one candidate trajectory; in response to there being one candidate trajectory, determine the candidate trajectory as a matchable trajectory; in response to there being multiple candidate trajectories, inhibit the multiple candidate trajectories based on a second relative threshold and an absolute threshold according to the normalized similarity and the fourth similarity corresponding to each candidate trajectory, so as to obtain a matchable trajectory.

[0301] It should be noted that the vehicle positioning device in the embodiments of the present disclosure corresponds to the vehicle positioning method in the embodiments of the present disclosure in terms of technical implementation. For the specific implementation of the vehicle positioning device, reference may be made to the relevant records in the vehicle positioning method embodiment part above; the vehicle positioning device in the embodiments of the present disclosure also corresponds to the vehicle positioning method in the embodiments of the present disclosure in terms of technical effects. The achievable effects of the vehicle positioning device may also be referred to the relevant records in the vehicle positioning method embodiment part above. To reduce redundancy, it will not be elaborated here.

[0302] Exemplary electronic device

[0303] The embodiments of the present disclosure further provide an electronic device, including: a memory, configured to store a computer program product; a processor, configured to execute the computer program product stored in the memory, and when the computer program product is executed, implement the vehicle positioning method in any of the above embodiments.

[0304] Figure 23The structural diagram of an electronic device provided by an embodiment of the present disclosure includes at least one processor 100 and a memory 200.

[0305] The processor 100 can be a central processing unit (CPU) or other forms of processing units with data processing capabilities and / or instruction execution capabilities, and can control other components in the electronic device to perform desired functions.

[0306] The memory 200 can include one or more computer program products, and the computer program products can include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory can include, for example, random access memory (RAM) and / or cache memory, etc. Non-volatile memory can include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions can be stored on the computer-readable storage media, and the processor 100 can run one or more computer program instructions to implement the vehicle positioning methods and / or other desired functions of various embodiments of the present disclosure described above.

[0307] In one example, the electronic device can further include: an input device 300 and an output device 400, and these components are interconnected through a bus system and / or other forms of connection mechanisms (not shown). The input device 300 can also include, for example, a keyboard, a mouse, etc. The output device 400 can output various information to the outside, which can include, for example, a display, a speaker, a printer, and a communication network and its connected remote output devices, etc.

[0308] Of course, for simplicity, Figure 23 only some of the components related to the present disclosure in the electronic device are shown, and components such as buses, input / output interfaces, etc. are omitted. In addition, according to specific application scenarios, the electronic device 10 can further include any other appropriate components.

[0309] Exemplary computer program product and computer-readable storage medium

[0310] In addition to the above methods and devices, an embodiment of the present disclosure can also provide a computer program product, including computer program instructions, and when the computer program instructions are run by a processor, the processor is caused to execute the steps in the vehicle positioning methods of various embodiments of the present disclosure described in the above "Exemplary Method" section.

[0311] A computer program product may write program code for performing the operations of the embodiments of the present disclosure in any combination of one or more programming languages. The programming languages include object-oriented programming languages such as Java, C++, etc., and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on the user's computing device, partially on the user's device, executed as a stand-alone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0312] In addition, an embodiment of the present disclosure may also be a computer-readable storage medium having computer program instructions stored thereon. When the computer program instructions are run by a processor, the processor is caused to execute the steps in the vehicle positioning methods of various embodiments of the present disclosure described in the above "Exemplary Method" section.

[0313] The computer-readable storage medium may employ any combination of one or more readable media. The readable media may be a readable signal medium or a readable storage medium. The readable storage medium, for example but not limited to, includes systems, devices, or components of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0314] The basic principles of the present disclosure have been described above in conjunction with specific embodiments. However, the advantages, benefits, effects, etc. mentioned in the present disclosure are only examples and not limitations, and it cannot be considered that they are essential for each embodiment of the present disclosure. In addition, the above-disclosed specific details are only for the purposes of illustration and facilitating understanding, rather than limitations. The above details do not limit the present disclosure to necessarily adopt the above specific details for implementation.

[0315] Those skilled in the art can make various changes and modifications to the present disclosure without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present disclosure and their equivalent technologies, the present disclosure is also intended to include these changes and modifications.

Claims

1. A vehicle positioning method, comprising: Obtaining a reference trajectory in a scene map of the current scene where the vehicle is located in response to a trigger condition; Determining a first feature point of the reference trajectory based on the pose attribute information of the reference trajectory points in the reference trajectory; Determining incremental feature points based on the pose attribute information of the incremental driving trajectory points located within the current positioning window in the vehicle's current driving trajectory; Matching the incremental feature points with the first feature points to obtain incremental feature point pairs, where the incremental feature point pairs include the matched incremental feature points and the first feature points; Registering the reference trajectory and the driving trajectory based on a first feature point sequence formed by the first feature points in multiple groups of feature point pairs in chronological order, and a second feature point sequence formed by the second feature points in the multiple groups of feature point pairs in chronological order; wherein, the multiple groups of feature point pairs at least include some of the feature point pairs obtained by matching the first feature points of the reference trajectory with the second feature points of the driving trajectory, and the multiple groups of feature point pairs at least include some of the incremental feature point pairs; In response to the matching of the registered reference trajectory and driving trajectory, determining the current positioning result of the vehicle based on the reference trajectory point corresponding to the end driving trajectory point in the registered driving trajectory.

2. The method according to claim 1, wherein Determining a first feature point of the reference trajectory based on the pose attribute information of the reference trajectory points in the reference trajectory, including: For the reference trajectory points in the reference trajectory, in a sliding window manner, determining a first reference trajectory point and a second reference trajectory point corresponding to the ramp entry point and ramp exit point of the reference trajectory respectively based on the pitch angle information in the pose attribute information of the reference trajectory points within the sliding window; Determining a first slope feature point based on the first reference trajectory point and the second reference trajectory point, where the first feature point includes the first slope feature point; And / or, Based on the yaw angle information in the pose attribute information of the reference trajectory points, determining a third reference trajectory point corresponding to the vehicle's turning in the reference trajectory as a first turning feature point, where the first feature point includes the first turning feature point; And / or, For the reference trajectory points in each positioning window respectively, determining first valid trajectory points based on the satellite signal information in the pose attribute information of the reference trajectory points and the relationship between a first change amount and a second change amount; wherein, the first change amount is the change amount of the relative position information in the pose attribute information of the reference trajectory points with respect to the relative position information in the pose attribute information of the previous adjacent reference trajectory point, and the second change amount is the change amount of the global position information in the pose attribute information of the reference trajectory points with respect to the global position information in the pose attribute information of the previous adjacent reference trajectory point; the relative position information is the position information in a specified local coordinate system, and the global position information is the position information in a global position coordinate system; Clustering the first valid trajectory points corresponding to each positioning window respectively to obtain a first clustering result corresponding to each positioning window; The center of the first clustering result corresponding to each positioning window is determined respectively as the first global position feature point, and the first feature points include the first global position feature point.

3. The method according to claim 1, wherein Based on the pose attribute information of the incremental driving trajectory points within the current positioning window in the vehicle's current driving trajectory, the incremental feature points of the driving trajectory are determined, including: For the incremental driving trajectory points and adjacent historical driving trajectory points, in a sliding window manner, based on the pitch angle information in the pose attribute information of the driving trajectory points within the sliding window, the first driving trajectory point and the second driving trajectory point corresponding to the ramp entry point and the ramp exit point of the driving trajectory are determined respectively, and the adjacent historical driving trajectory points include at least one driving trajectory point in the driving trajectory whose time sequence is before the incremental driving trajectory point; Based on the first driving trajectory point and the second driving trajectory point, a second slope feature point is determined, and the second feature point and the incremental feature points include the second slope feature point; and / or, Based on the yaw angle information in the pose attribute information of the incremental driving trajectory points and adjacent historical driving trajectory points, the third driving trajectory point corresponding to the vehicle's steering in the driving trajectory is determined as the second steering feature point, and the second feature point and the incremental feature points include the second steering feature point, and the adjacent historical driving trajectory points include at least one driving trajectory point in the driving trajectory whose time sequence is before the incremental driving trajectory point; and / or, Based on the satellite signal information in the pose attribute information of the incremental driving trajectory points and the relationship between the third change amount and the fourth change amount, a second valid trajectory point is determined; wherein, the third change amount is the change amount of the relative position information in the pose attribute information of the incremental driving trajectory points with respect to the relative position information in the pose attribute information of the previous adjacent driving trajectory point, and the fourth change amount is the change amount of the global position information in the pose attribute information of the incremental driving trajectory points with respect to the global position information in the pose attribute information of the previous adjacent driving trajectory point; Cluster the second valid trajectory points to obtain a second clustering result; Determine the center of the second clustering result as the second global position feature point, and the second feature point and the incremental feature points include the second global position feature point.

4. The method according to claim 1, wherein The types of the first feature point, the second feature point, and the incremental feature point respectively include at least one of a slope feature point, a steering feature point, and a global position feature point; Match the incremental feature points with the first feature points to obtain incremental feature point pairs, including: In response to the type of the incremental feature point being a slope feature point, determine a first slope feature point sequence based on the first slope feature points in the first feature points; based on the pose attribute information of the incremental feature points and the pose attribute information of the first slope feature points in the first slope feature point sequence, match the incremental feature points with the first slope feature points in the first slope feature point sequence to obtain a first similarity; Based on the first similarity, determine the first slope feature point that the incremental feature point matches; In response to the type of the incremental feature point being a steering feature point, determine a first steering feature point sequence based on the first steering feature points in the first feature points; based on the pose attribute information of the incremental feature point and the pose attribute information of the first steering feature points in the first steering feature point sequence, match the incremental feature point with the first steering feature points in the first steering feature point sequence to obtain a second similarity; Based on the second similarity, determine the first steering feature point that the incremental feature point matches; In response to the type of the incremental feature point being a global position feature point, determine a first global position feature point sequence based on the first global position feature points in the first feature points; based on the relative position information and global position information in the pose attribute information of the incremental feature point and the relative position information and global position information in the pose attribute information of the first global position feature points in the first global feature point sequence, determine the similarity between the incremental feature point and the first global position feature points in the first global position feature point sequence to obtain a third similarity; Based on the third similarity, determine the first global position feature point that the incremental feature point matches; 5. According to the method as claimed in any one of claims 1-4, wherein In response to the registered reference trajectory and the driving trajectory matching, determine the current positioning result of the vehicle based on the reference trajectory points corresponding to the end driving trajectory points in the registered driving trajectory, including: Obtain the similarity between the registered reference trajectory and the driving trajectory to obtain a fourth similarity; Determine the relationship between the fourth similarity and a first preset threshold; In response to the fourth similarity being greater than the first preset threshold, determine that the registered reference trajectory and the driving trajectory match; Determine the matching relationship between the end driving trajectory point and the corresponding reference trajectory point; In response to the end driving trajectory point and the corresponding reference trajectory point matching, determine the corresponding reference trajectory point as the current positioning result of the vehicle.

6. The method according to claim 5, wherein, Based on the first feature point sequence formed by the first feature points in multiple groups of feature point pairs in chronological order, and the second feature point sequence formed by the second feature points in multiple groups of feature point pairs in chronological order, register the reference trajectory and the driving trajectory, including: Downsample the first feature point sequence and the second feature point sequence respectively according to a preset sampling method to obtain a first sampling point sequence and a second sampling point sequence correspondingly; Based on the first sampling point sequence and the second sampling point sequence, register the reference trajectory and the driving trajectory.

7. The method according to claim 6, further comprising: In response to the trigger condition, obtain the pose attribute information of the driving trajectory points corresponding to the trigger condition in the driving trajectory; Based on the pose attribute information of the driving trajectory points corresponding to the trigger condition, determine an initial feature point; Based on the initial feature point, determine a positioning reference point, and the second feature points include the positioning reference point and the feature points of the driving trajectory that are chronologically after the positioning reference point; Determine the reference trajectory point that the positioning reference point matches as the reference benchmark point; Determine the positioning reference point and the feature points among the initial feature points whose time sequence is after the positioning reference point as the incremental feature points corresponding to the first positioning window, and perform the operation of matching the incremental feature points with the first feature points; Based on the pose attribute information of the reference trajectory points in the reference trajectory, determine the first feature point of the reference trajectory, including: Based on the reference reference point and the pose attribute information of the reference trajectory points in the reference trajectory whose time sequence is after the reference reference point, determine the first feature point of the reference trajectory.

8. The method according to claim 7, wherein, Determine the relationship between the fourth similarity and the first preset threshold, including: According to the preset adjustment method, adjust the positioning reference point based on the initial feature points, and adjust the second feature point sequence based on the adjusted positioning reference point; Downsample the adjusted second feature point sequence to obtain a third sampling point sequence; Based on the first sampling point sequence and the third sampling point sequence, register the reference trajectory and the driving trajectory; Obtain the similarity between the registered reference trajectory and the driving trajectory to obtain a fifth similarity; Determine the relationship between the exceeding value of the fifth similarity relative to the fourth similarity and the second preset threshold; In response to the exceeding value of the fifth similarity relative to the fourth similarity being greater than the second preset threshold, iteratively execute the operation of adjusting the positioning reference point based on the initial feature points according to the preset adjustment method, and adjusting the second feature point sequence based on the adjusted positioning reference point until the exceeding value of the obtained fifth similarity relative to the fifth similarity obtained in the previous adjacent time is not greater than the second preset threshold, and use the fifth similarity obtained last time as the fourth similarity, and determine the relationship between the fourth similarity and the first preset threshold; In response to the exceeding value of the fifth similarity relative to the fourth similarity not being greater than the second preset threshold, determine the relationship between the fourth similarity and the first preset threshold.

9. The method according to claim 7, wherein, Determine the relationship between the fourth similarity and the first preset threshold, including: According to the preset adjustment method, adjust the reference reference point, and adjust the first feature point sequence based on the adjusted reference reference point; Downsample the adjusted first feature point sequence to obtain a fourth sampling point sequence; Based on the fourth sampling point sequence and the second sampling point sequence, register the reference trajectory and the driving trajectory; Obtain the similarity between the registered reference trajectory and the driving trajectory to obtain a sixth similarity; Determine the relationship between the exceeding value of the sixth similarity relative to the fourth similarity and the second preset threshold; In response to the excess value of the sixth similarity relative to the fourth similarity being greater than the second preset threshold, iteratively execute the operation of adjusting the reference reference point according to the preset adjustment method and adjusting the first feature point sequence based on the adjusted reference reference point until the excess value of the obtained sixth similarity relative to the sixth similarity obtained in the previous adjacent time is not greater than the second preset threshold. Use the finally obtained sixth similarity as the fourth similarity to determine the relationship between the fourth similarity and the first preset threshold; In response to the excess value of the sixth similarity relative to the fourth similarity being not greater than the second preset threshold, determine the relationship between the fourth similarity and the first preset threshold.

10. The method according to claim 5, wherein, In response to a trigger condition, obtain a reference trajectory in the scene map of the current scene where the vehicle is located, including: In response to a trigger condition, obtain the pose attribute information of the driving trajectory point where the vehicle is currently located; Determine a target trajectory point corresponding to the pose attribute information that is consistent with the pose attribute information of the driving trajectory point where the vehicle is currently located among the pose attribute information corresponding to at least one scene map stored in the vehicle; Determine the trajectory where the target trajectory point is located as the reference trajectory.

11. The method according to claim 10, wherein, Based on the pose attribute information of the reference trajectory points in the reference trajectory, determine the first feature points of the reference trajectory, including: In response to there being multiple reference trajectories, respectively for each reference trajectory, based on the pose attribute information of the reference trajectory points in the reference trajectory, determine the first feature points of the reference trajectory; Determine the relationship between the fourth similarity and the first preset threshold, including: According to the fourth similarity corresponding to each registered reference trajectory, suppress each registered reference trajectory based on a preset suppression exit condition to obtain a matchable trajectory; For the matchable trajectory, determine the relationship between the fourth similarity and the first preset threshold.

12. The method according to claim 11, wherein, Determine the relationship between the fourth similarity and the first preset threshold, including: In response to there being multiple matchable trajectories, respectively for each matchable trajectory, determine the relationship between the fourth similarity and the first preset threshold; Determine the corresponding reference trajectory point as the current positioning result of the vehicle, including: Update the reference trajectory corresponding to the next positioning window to the matchable trajectory where the corresponding reference trajectory point matched by the end driving trajectory point is located.

13. The method according to claim 11, wherein, According to the fourth similarity corresponding to each registered reference trajectory, suppress each registered reference trajectory based on a preset suppression exit condition to obtain a matchable trajectory, including: Based on the fourth similarity corresponding to each registered reference trajectory, determine the normalized similarity corresponding to each registered reference trajectory; According to the normalized similarity corresponding to each registered reference trajectory, suppress each registered reference trajectory based on a first relative threshold to obtain at least one candidate trajectory; In response to there being one candidate trajectory, determine the candidate trajectory as the matchable trajectory; In response to there being multiple candidate trajectories, suppress the multiple candidate trajectories based on a second relative threshold and an absolute threshold according to the normalized similarity and the fourth similarity corresponding to each candidate trajectory to obtain the matchable trajectory.

14. A vehicle positioning device, comprising: A first acquisition module, configured to acquire a reference trajectory in a scene map of a current scene where the vehicle is located in response to a trigger condition; A first determination module, configured to determine a first feature point of the reference trajectory based on pose attribute information of reference trajectory points in the reference trajectory; A second determination module, configured to determine incremental feature points based on pose attribute information of incremental travel trajectory points located within a current positioning window in the vehicle's current travel trajectory; A matching module, configured to match the incremental feature points with the first feature points to obtain incremental feature point pairs, where each incremental feature point pair includes a matched incremental feature point and a first feature point; A registration module, configured to register the reference trajectory and the travel trajectory based on a first feature point sequence formed by first feature points in a plurality of groups of feature point pairs in chronological order, and a second feature point sequence formed by second feature points in the plurality of groups of feature point pairs in chronological order; wherein, the plurality of groups of feature point pairs at least include some of the feature point pairs obtained by matching the first feature points of the reference trajectory with the second feature points of the travel trajectory, and the plurality of groups of feature point pairs at least include some of the incremental feature point pairs; A positioning module, configured to determine a current positioning result of the vehicle based on a reference trajectory point corresponding to an end travel trajectory point in the registered travel trajectory in response to the matching of the registered reference trajectory and travel trajectory.

15. An electronic device, comprising: A memory, configured to store a computer program product; A processor, configured to execute the computer program product stored in the memory, and when the computer program product is executed, implement the method according to any one of claims 1-13 above.

16. A computer-readable storage medium, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, implement the method according to any one of claims 1-13 above.

Images