CN117170377A

CN117170377A - Automatic driving method and device and vehicle

Info

Publication number: CN117170377A
Application number: CN202311248263.XA
Authority: CN
Inventors: 白天; 田涛涛; 孙杰
Original assignee: Hangzhou Hikvision Digital Technology Co Ltd
Current assignee: Hangzhou Hikvision Digital Technology Co Ltd
Priority date: 2023-09-25
Filing date: 2023-09-25
Publication date: 2023-12-05

Abstract

The application provides an automatic driving method, an automatic driving device and a vehicle, wherein the method comprises the following steps: selecting a plurality of teaching path points from the teaching paths; determining expected path points corresponding to each teaching path point; generating a target path based on a plurality of expected path points corresponding to the plurality of taught path points; controlling a target vehicle to travel based on the target path; and when the target path is positioned at the center position, the control target vehicle runs based on a first speed, and when the target path is positioned at the right position, the control target vehicle runs based on a second speed, wherein the first speed is greater than the second speed. According to the technical scheme, the method and the device for generating the central path or the right-leaning path can be used for generating the central path or the right-leaning path in a self-adaptive mode, and when a vehicle runs on the central path, the running speed of the vehicle can be increased, and the running efficiency is improved. When the vehicle runs on a right path, the running speed of the vehicle can be reduced, and the running safety of the vehicle is ensured.

Description

Automatic driving method and device and vehicle

Technical Field

The present application relates to the field of automatic driving technologies, and in particular, to an automatic driving method and apparatus, and a vehicle.

Background

Autopilot is a technology that enables a vehicle to navigate and operate autonomously without human intervention using advanced sensors and computing technologies. The automatic driving can provide safer, more convenient and efficient travel modes for people, and simultaneously, the whole traffic system and city planning are also affected deeply.

The autonomous stay parking (AVP, automated Valet Parking) function is a technology of automatic driving, and a vehicle with the autonomous stay parking function does not need manual intervention, so that the automatic identification of a parking space can be realized through a vehicle-mounted sensor, a processor and a control system, and the parking process can be automatically completed. When the autonomous parking function is used, a driver can get off at a designated get-off point, a parking instruction is issued through the mobile phone APP, and the vehicle can automatically travel to a parking space of a parking lot after receiving the instruction without user operation.

However, during autonomous parking of the vehicle, a collision may occur with surrounding obstacles, such as the vehicle being unable to avoid the surrounding obstacles during traveling, thereby causing a safety hazard to the vehicle.

Disclosure of Invention

The application provides an automatic driving method, which is applied to a target vehicle, wherein the target vehicle stores a teaching path, and after the target vehicle starts automatic driving, the method comprises the following steps:

selecting a plurality of teaching path points from the teaching paths;

determining expected path points corresponding to each teaching path point; wherein, for each teaching path point, if it is determined that no obstacle exists on the left side and/or the right side of the teaching path point based on the perception information, the teaching path point is taken as a desired path point; if the first obstacle exists on the left side of the teaching path point and the second obstacle exists on the right side of the teaching path point based on the perception information, the teaching path point is taken as an expected path point when the distance between the first obstacle and the second obstacle is smaller than a threshold value; when the distance is not smaller than the threshold value, taking the center point between the first obstacle and the second obstacle as an expected path point;

Generating a target path based on a plurality of expected path points corresponding to the plurality of taught path points;

controlling a target vehicle to travel based on the target path; wherein the control target vehicle travels based on a first speed when the target path is located at the center position, and travels based on a second speed when the target path is located at the right position, the first speed being greater than the second speed.

The present application provides an automatic driving apparatus applied to a target vehicle, the apparatus comprising:

the determining module is used for selecting a plurality of teaching path points from the teaching path after the target vehicle starts automatic driving, and determining expected path points corresponding to each teaching path point; wherein the target vehicle stores a teaching path; wherein, for each taught path point, if it is determined that no obstacle exists on the left side and/or the right side of the taught path point based on the perception information, the taught path point is taken as a desired path point; if the first obstacle exists on the left side of the teaching path point and the second obstacle exists on the right side of the teaching path point based on the perception information, the teaching path point is taken as an expected path point when the distance between the first obstacle and the second obstacle is smaller than a threshold value; when the distance is not smaller than the threshold value, taking the center point between the first obstacle and the second obstacle as an expected path point;

The generation module is used for generating a target path based on a plurality of expected path points corresponding to the plurality of teaching path points;

a control module for controlling the target vehicle to travel based on the target path; wherein the control target vehicle travels based on a first speed when the target path is located at the center position, and travels based on a second speed when the target path is located at the right position, and the first speed is greater than the second speed.

The present application provides an autonomous vehicle comprising: a processor and a machine-readable storage medium storing machine-executable instructions executable by the processor; the processor is configured to execute machine-executable instructions to implement the autopilot method of the present application as exemplified above.

As can be seen from the above technical solutions, in the embodiments of the present application, when the distance between the first obstacle on the left side and the second obstacle on the right side is large, a centering path (i.e., the center positions of all lanes can cross the marking in the middle of the road) is adaptively generated, so that the vehicle is controlled to travel on the centering path, and the traveling speed of the vehicle can be improved, the vehicle is controlled to travel quickly, and the traveling efficiency is improved. When the speed of the vehicle is increased, the foot reaction time (namely, the interval between the centering path and the right side is larger, and the foot reaction time can be reserved) is reserved for pedestrians or vehicles which suddenly appear on the right side, so that no collision is ensured. When no obstacle exists on the left side and/or the right side, or the distance between the first obstacle on the left side and the second obstacle on the right side is smaller, a right-leaning path (namely a right lane which cannot cross a marking in the middle of a road) is adaptively generated, the vehicle is controlled to run on the right-leaning path, the running speed of the vehicle can be reduced, the running safety of the vehicle is ensured, and no collision is ensured by reducing the running speed of the vehicle.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following description will briefly describe the drawings required to be used in the embodiments of the present application or the description in the prior art, and it is obvious that the drawings in the following description are only some embodiments described in the present application, and other drawings may be obtained according to these drawings of the embodiments of the present application for a person having ordinary skill in the art.

FIG. 1 is a flow chart of an autopilot method in one embodiment of the present application;

FIG. 2 is a flow chart of an autopilot method in one embodiment of the present application;

3A-3C are schematic diagrams of determining a target lane model in one embodiment of the application;

FIGS. 4A-4C are schematic illustrations of determining a target gradient in one embodiment of the application;

FIG. 5 is a schematic illustration of an obstruction constraint in one embodiment of the application;

FIG. 6 is a schematic representation of the travel of a target vehicle in one embodiment of the application;

FIG. 7 is a schematic view of an autopilot in one embodiment of the present application;

fig. 8 is a hardware configuration diagram of a vehicle in an embodiment of the present application.

Detailed Description

The terminology used in the embodiments of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to any or all possible combinations including one or more of the associated listed items.

It should be understood that although the terms first, second, third, etc. may be used in embodiments of the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the application. Depending on the context, furthermore, the word "if" used may be interpreted as "at … …" or "at … …" or "in response to a determination".

The embodiment of the application provides an automatic driving method which can be applied to a target vehicle (namely, a vehicle supporting an automatic driving function), and the target vehicle stores a teaching path. Referring to fig. 1, which is a schematic flow chart of the method, after the target vehicle starts automatic driving, the method may include:

Step 101, selecting a plurality of teaching path points from the teaching paths.

102, determining expected path points corresponding to each teaching path point; wherein, for each taught path point, if it is determined that no obstacle exists on the left side and/or the right side of the taught path point based on the perception information, the taught path point is taken as a desired path point; if the first obstacle exists on the left side of the teaching path point and the second obstacle exists on the right side of the teaching path point based on the perception information, the teaching path point is taken as an expected path point when the distance between the first obstacle and the second obstacle is smaller than a threshold value; when the distance is not less than the threshold value, a center point between the first obstacle and the second obstacle is set as a desired path point.

And 103, generating a target path based on a plurality of expected path points corresponding to the plurality of teaching path points.

104, controlling the target vehicle to run based on the target path; wherein the control target vehicle travels based on a first speed when the target path is located at the center position, and travels based on a second speed when the target path is located at the right position, the first speed may be greater than the second speed.

For example, if the target path corresponds to at least two lanes, the center position may be the center position of all lanes (the vehicle may span the middle of the road marking when in the center position), and the right side position may be the right side lane position (e.g., the rightmost one lane, the vehicle may not span the middle of the road marking when in the right side position).

Illustratively, for each taught path point, determining a desired path point corresponding to the taught path point may include, but is not limited to: determining a target lane model corresponding to the teaching path point; if the right side of the teaching path point is determined to have no obstacle based on the perception information, determining that the target lane model is a lane-free model; if the left side of the teaching path point is determined to have no obstacle based on the perception information, determining that the target lane model is a single lane model; if the first obstacle exists on the left side and the second obstacle exists on the right side of the teaching path point based on the perception information, and the distance between the first obstacle and the second obstacle is smaller than a threshold value, determining that the target lane model is a single lane model; and if the first obstacle exists on the left side and the second obstacle exists on the right side of the teaching path point based on the perception information, and the distance between the first obstacle and the second obstacle is not smaller than a threshold value, determining that the target lane model is a double lane model.

Determining a desired path point based on the target lane model; if the target lane model is a lane-free model or a single lane model, the teaching path point is taken as an expected path point; if the target lane model is a two-lane model, a center point between the first obstacle and the second obstacle is taken as a desired path point.

Illustratively, generating the target path based on a plurality of desired path points corresponding to a plurality of taught path points may include, but is not limited to: and taking a path formed by a plurality of expected path points as a target path. Or,

determining target gradients corresponding to the plurality of expected path points; wherein, for each desired path point, the target gradient corresponding to the desired path point may be determined based on at least one of a smooth term gradient of the desired path point, a curvature term gradient of the desired path point, and an obstacle term gradient of the desired path point. And optimizing the plurality of expected path points based on the target gradients corresponding to the plurality of expected path points to obtain a plurality of candidate path points corresponding to the plurality of expected path points. If the plurality of candidate path points meet the convergence condition, a reference path may be generated based on the plurality of candidate path points, and a target path may be generated based on the reference path; if the plurality of candidate path points do not satisfy the convergence condition, the plurality of candidate path points may be used as a plurality of desired path points, the operation of determining the target gradient corresponding to the plurality of desired path points may be performed in a return manner, and the above-described process may be repeated.

For example, for each desired path point, the determination of the smoothed term gradient for that desired path point may include, but is not limited to: the smoothed term gradient for the desired path point is determined using the following formula:

wherein Deltax is _i+2 A vector representing a distance between a first desired path point subsequent to the desired path point and a second desired path point subsequent to the desired path point; Δx _i+1 A vector representing a distance between the desired path point and a first desired path point subsequent to the desired path point; Δx _i A vector representing a first desired path point preceding the desired path point and the desired path point; Δx _i-1 For representing a second desired path point preceding the desired path point and the periodA vector between a first desired path point preceding the look-ahead path point; x is x _i For representing the desired path point; grad _smooth A smoothed term gradient representing the desired path point.

For example, for each desired path point, the determination of the curvature item gradient for that desired path point may include, but is not limited to: the curvature term gradient for the desired path point is determined using the following formula:

wherein Deltax is _i A vector representing a first desired path point preceding the desired path point and the desired path point; ΔΦ of _i The method comprises the steps of representing an included angle between a first straight line and a second straight line, wherein the first straight line is a connecting line of a first expected path point in front of the expected path point and the expected path point, and the second straight line is a connecting line of a first expected path point behind the expected path point and the expected path point; x is x _i For representing the desired path point; x is x _i-1 For representing a first desired waypoint preceding the desired waypoint; x is x _i+1 For representing a first desired path point subsequent to the desired path point; grad _curvature A gradient of curvature terms representing the desired path point.

For example, for each desired waypoint, the determination of the obstacle item gradient for that desired waypoint may include, but is not limited to: the obstacle term gradient for the desired path point is determined using the following formula:

wherein x represents the abscissa of the expected path point corresponding to the grid graph, and y represents the expected path point in the gridThe ordinate corresponding to the figure; epsilon represents the configured value; f (x, y) represents the grid offset of (x, y) from the nearest obstacle in the grid map; grad _obs An obstacle item gradient representing the desired path point.

Wherein d _obs Representing the closest distance of the desired path point to the obstacle; x represents the abscissa of the desired path point and y represents the ordinate of the desired path point; const represents the configured furthest distance value; d, d _{min_consider} Representing const and d _obs A difference between them; grad _obs An obstacle item gradient representing the desired path point.

Illustratively, generating the target path based on the reference path may include, but is not limited to: the reference path is taken as the target path. Or selecting a plurality of reference path points from the reference path, and acquiring a plurality of path segments based on the plurality of reference path points, wherein each path segment can comprise a first reference path point and a second reference path point which are adjacent to each other; for each path segment, an iterative optimization parameter may be determined based on a first reference path point and a second reference path point for the path segment; and determining a candidate polynomial curve corresponding to the path segment based on the iterative optimization parameter and the configured numerical interval, wherein the numerical interval can comprise a plurality of numerical values between 0 and 1. On this basis, the target path may be determined based on the candidate polynomial curve.

Illustratively, determining the target path based on the candidate polynomial curve may include, but is not limited to: a first curvature constraint parameter and a second curvature constraint parameter are determined based on the candidate polynomial curve, and a target curvature is determined based on the first curvature constraint parameter and the second curvature constraint parameter. If the target curvature is smaller than a preset threshold, the candidate polynomial curve can be used as a target path. If the target curvature is not smaller than the preset threshold value, the first point and/or the last point of the candidate polynomial curve can be adjusted to obtain an adjusted curve, the first point of the adjusted curve is used as a first reference path point, the last point of the adjusted curve is used as a second reference path point, and the operation of determining the iterative optimization parameters based on the first reference path point and the second reference path point of the path section is carried out back on the basis.

Illustratively, determining a first curvature constraint parameter and a second curvature constraint parameter based on the candidate polynomial curve, determining a target curvature based on the first curvature constraint parameter and the second curvature constraint parameter, including but not limited to: determining a first curvature constraint parameter, a second curvature constraint parameter, and a target curvature based on the following formula:

wherein x 'represents the first lateral derivative of the target point on the candidate polynomial curve, and the target point may be any point on the candidate polynomial curve, y' represents the first longitudinal derivative of the target point, x "represents the second lateral derivative of the target point, y" represents the second longitudinal derivative of the target point, C ₁ Representing the first curvature constraint parameter, C ₂ Representing the second curvature constraint parameter, K representing the target curvature.

The following describes an automatic driving method according to an embodiment of the present application with reference to specific embodiments.

For a vehicle supporting an autopilot function (hereinafter referred to as a target vehicle), in order to implement an autonomous stay function (also referred to as a memory stay function), the target vehicle may deploy sensors such as a camera (such as a look-around camera), an ultrasonic radar (which may be replaced with a millimeter wave radar or a laser radar), and the like.

In order to realize the autonomous stay parking function, a differential GPS base station can be set up at a proper position, a target vehicle receives differential GPS signals by using a vehicle-mounted antenna to obtain longitude and latitude height information of the target vehicle, and the base station is used as an origin to calculate transverse coordinates and longitudinal coordinates of the target vehicle, such as transverse coordinates and longitudinal coordinates in a northeast coordinate system. Wherein, differential GPS means: a GPS receiver is arranged on the reference station for observation, the distance correction of the reference station to the satellite is calculated according to the known precise coordinates of the reference station, and the reference station sends the data in real time. The user receiver receives the correction sent by the reference station while performing GPS observation, and corrects the positioning result, thereby improving the positioning precision.

In order to realize the autonomous stay function, the target vehicle can utilize vehicle-mounted inertial navigation equipment (such as an IMU (Inertial Measurement Unit, inertial measurement unit) and the like) to measure real-time attitude angle information of the target vehicle, and combine a vehicle-mounted wheel speed encoder to perform data fusion so as to obtain accurate speed information of the target vehicle.

In order to realize the autonomous stay function, the throttle, brake, steering wheel, etc. of the target vehicle are connected to the controller through a CAN (Controller Area Network ) bus to be controlled respectively.

To implement the autonomous stay-in-park function, a cruise phase and a park phase may be involved. The cruising stage refers to the cruising stage of the autonomous passenger-taking parking function after the target vehicle starts automatic driving and before the target vehicle reaches a specified parking space to start reversing. The parking stage refers to a process of reversing and warehousing after the target vehicle reaches a specified parking space, and the process is called a parking stage of an autonomous passenger parking function. The automatic driving method in this embodiment may be a cruising stage for the autonomous stay function.

In order to realize the autonomous stay parking function, the autonomous stay parking method relates to a teaching process of a target vehicle and an autonomous stay parking process of the target vehicle. In the teaching process of the target vehicle, the user is required to drive the target vehicle to complete the teaching process. For example, when the mode of the target vehicle is adjusted to the teaching mode, the user learns the teaching path of the user while driving the target vehicle (from the teaching start point to the teaching end point, i.e., designating the parking space), and the target vehicle constructs a high-precision map on the teaching path in real time. After the teaching process of the target vehicle is completed, the target vehicle stores the taught path and stores a high-precision map on the taught path.

For example, the teaching process of the target vehicle may be performed only once, or may be performed multiple times (for example, the teaching process of the target vehicle is performed again every month), so that the target vehicle stores the latest teaching path and the latest high-precision map, and the teaching process of the target vehicle is not limited in this embodiment.

After the target vehicle stores the taught path and the high-precision map, an autonomous stay process of the target vehicle may be performed. For example, the mode of the target vehicle is adjusted to an automatic driving mode, and the user drives the target vehicle to reach the teaching starting point, so that the target vehicle can complete the autonomous stay process based on the teaching path and the high-precision map, namely, the target vehicle runs from the teaching starting point to the teaching end point and reaches the designated parking space.

For example, the automatic valet parking function is generally applied in a parking lot environment or a closed park environment, and the uncertainty of ground traffic participants in the scene is high, pedestrians or obstacles around a target vehicle can suddenly appear, so that the target vehicle can collide with the surrounding obstacles, and potential safety hazards exist.

For the above findings, in the embodiment of the present application, a centered path (i.e., the center position of all lanes, which can cross the marking in the middle of the road) or a right-hand path (i.e., the right-hand lane, which cannot cross the marking in the middle of the road) may be adaptively generated. For the centering path, the running speed of the vehicle can be increased, and even if pedestrians or vehicles suddenly appear, the reaction time can be reserved due to the fact that the distance between the centering path and the right side is larger, so that collision is avoided. For the right-leaning path, the running speed of the vehicle can be reduced, and even if a pedestrian or the vehicle suddenly appears, the running speed of the target vehicle is slower, the sufficient reaction time can be reserved, so that no collision is ensured.

Illustratively, at the application implementation level, the user may set the driving style mode of the target vehicle. If the driving style mode is a speed priority mode, the target vehicle generates a central path and increases the running speed of the vehicle, i.e. the target vehicle can run at a higher speed across the marked line in the middle of the road. Alternatively, if the driving style mode is the safety priority mode, the target vehicle generates a right-hand path and reduces the running speed of the vehicle, that is, the target vehicle runs at a lower speed in the rightmost lane. Alternatively, if the driving style mode is the intelligent mode, the target vehicle may adaptively generate a centered path or a right path, that is, adaptively adjust the speed priority or the safety priority according to the front obstacle distribution, and the intelligent mode is taken as an example in the embodiment.

In an embodiment of the present application, an automatic driving method is provided, after a target vehicle travels to a teaching origin and the target vehicle starts automatic driving (i.e., an automatic driving mode), as shown in fig. 2, the method includes:

step 201, the target vehicle selects a plurality of taught path points from the taught path.

For example, the target vehicle stores a taught path, and a plurality of taught path points may be selected from the taught path, for example, a path point may be selected from the taught path as a taught path point at a predetermined distance. For example, the teaching path may be sampled, and a plurality of discrete points at equal intervals may be obtained as teaching path points.

Step 202, for each taught path point, the target vehicle determines a target lane model corresponding to the taught path point, where the target lane model may be a lane-free model, a single lane model, or a two lane model.

Referring to fig. 3A, the following steps may be employed to determine a target lane model corresponding to the taught path point:

step 2021, the target vehicle acquires the sensing information around the own target vehicle.

For example, the perceptual information may also be referred to as environmental information, which may include, but is not limited to, at least one of: wall information, column information, road edge information, lane line information (such as a dotted lane line and/or a solid lane line of a road center, and the like), parking space information, road surface parking vehicle information, and the like.

The target vehicle can acquire sensing information around the target vehicle based on a camera (such as a look-around camera), an ultrasonic radar and other sensors, and the position of the target vehicle is determined in a high-precision map based on the sensing information.

Step 2022, the target vehicle determines whether an obstacle exists on the right side of the taught path point based on the perception information. If not, step 2023 may be performed, and if so, step 2024 may be performed.

For example, for each taught path point, based on the perception information and the high-precision map, it may be determined whether an obstacle exists on the right side of the taught path point and whether an obstacle exists on the left side of the taught path point, and the determination manner is not limited. The barriers can be walls, upright posts, road edges, lane lines, parking spaces, vehicles parked on the road surface, pedestrians and the like, and the types of the barriers are not limited.

In step 2023, the target vehicle determines that the target lane model corresponding to the taught path point is a lane-free model, that is, when no obstacle exists on the right side of the taught path point, the target lane model is a lane-free model.

Step 2024, the target vehicle determines whether an obstacle exists on the left side of the taught path point based on the perception information. If not, step 2025 may be performed, and if so, step 2026 may be performed.

In step 2025, the target vehicle determines that the target lane model corresponding to the taught path point is a single-lane model, that is, the target lane model is a single-lane model when no obstacle exists on the left side of the taught path point.

Step 2026 may refer to an obstacle on the left side of the taught path point as a first obstacle (the first obstacle is one of all left side obstacles closest to the taught path point), an obstacle on the right side of the taught path point as a second obstacle (the second obstacle is one of all right side obstacles closest to the taught path point), the target vehicle determines whether the distance between the first obstacle and the second obstacle is smaller than a threshold (may be empirically configured, such as the threshold is larger than two lane widths, etc.), if yes, step 2027 may be performed, and if no, step 2028 may be performed.

In step 2027, the target vehicle determines that the target lane model corresponding to the taught path point is a single-lane model, that is, the target lane model is a single-lane model when the distance between the first obstacle and the second obstacle is less than the threshold value.

In step 2028, the target vehicle determines that the target lane model corresponding to the taught path point is a two-lane model, that is, the target lane model is a two-lane model when the distance between the first obstacle and the second obstacle is not less than the threshold value.

To this end, step 202 is completed, and a target lane model corresponding to each taught path point may be determined.

Step 203, for each teaching path point, the target vehicle determines an expected path point of the teaching path point based on the target lane model corresponding to the teaching path point, so as to obtain the expected path point of each teaching path point.

For example, if the target lane model corresponding to the taught path point is a lane-free model, that is, it is determined that there is no obstacle on the right side of the taught path point based on the perception information, then the perception information is not reliable because there is usually an obstacle on the right side of the lane, and in order to ensure driving safety, when the target lane model is a lane-free model, the taught path point may be regarded as a desired path point, that is, right-to-right driving is required at the desired path point (usually driving on the right side lane in the teaching process, that is, the taught path point is located on the right side lane).

If the target lane model corresponding to the teaching path point is a single-lane model, if it is determined that no obstacle exists on the left side of the teaching path point based on the perception information, then the perception information is not reliable because the obstacle exists on the left side of the lane, and in order to ensure the driving safety, the teaching path point can be used as an expected path point, namely, the right-to-right driving is required at the expected path point when the target lane model is the single-lane model.

If the target lane model corresponding to the teaching path point is a single lane model, if it is determined that a first obstacle exists on the left side of the teaching path point and a second obstacle exists on the right side of the teaching path point based on the perception information, and the distance between the first obstacle and the second obstacle is smaller than a threshold value, then, in order to ensure driving safety, the teaching path point can be used as an expected path point when the target lane model is a single lane model, that is, the target lane model needs to be right to drive at the expected path point.

If the target lane model corresponding to the teaching path point is a two-lane model, if it is determined that a first obstacle exists on the left side of the teaching path point and a second obstacle exists on the right side of the teaching path point based on the perception information, and the distance between the first obstacle and the second obstacle is not smaller than a threshold value, then when the target lane model is a two-lane model due to the fact that the distance between the first obstacle and the second obstacle is relatively large, a center point between the first obstacle and the second obstacle can be used as an expected path point, namely, the target lane model can be driven in the middle of the expected path point.

In summary, the desired path point of each taught path point may be obtained, and the desired path point may be the taught path point or a center point between the first obstacle and the second obstacle. If the expected path point is a teaching path point, the vehicle needs to travel right at the expected path point, and if the expected path point is a center point between the first obstacle and the second obstacle, the vehicle needs to travel centered at the expected path point.

For example, referring to fig. 3B and 3C, a schematic diagram is shown in which the desired path point is a center point, that is, a schematic diagram in which the target vehicle needs to travel centrally. In fig. 3B, continuous parking space information and wall information are provided on both sides, and the parking space information and the wall information can be derived from a high-precision map or from sensing information, and form an obstacle in the current environment for determining an expected path point of a taught path point. In fig. 3C, although there is no parking space information as a reference, the right side and the parked vehicle as a reference, that is, constitute an obstacle in the current environment, are used to determine a desired path point of the taught path point.

After the above reference information is obtained, discrete sampling nodes are generated at intervals of a preset distance (e.g., 1 m). In the case of a two-lane situation, after the generation of the sampling nodes on both sides, the centrally centered sampling node may be generated together, as shown in fig. 3B. Under the single-lane condition or the no-lane condition, the sampling node centered in the center has no calculated reference basis, and the adjacent sampling points on the teaching track are directly taken. The sampling points of the target vehicle can generate a position in the middle of the lane in an open scene. And under a non-open scene, the sampling point of the target vehicle can be generated at a position on the right of the lane. Because open scenes are difficult to judge, a method of judging describing non-open scenes may include, but is not limited to: the scene does not have the conditions of dynamic motor vehicles, the dynamic motor vehicles in the parking spaces (considering the perception of false detection), the target vehicles in front of the gates or on the ascending and descending slopes and the like.

Step 204, after obtaining a plurality of expected path points corresponding to the plurality of teaching path points, the target vehicle determines a target gradient corresponding to the plurality of expected path points. Wherein, for each desired path point, the target gradient corresponding to the desired path point may be determined based on at least one of a smooth term gradient of the desired path point, a curvature term gradient of the desired path point, and an obstacle term gradient of the desired path point.

For example, in order to obtain the driving track of the target vehicle, the driving track of the target vehicle may be determined according to an optimization algorithm, where the objective of the optimization algorithm is to obtain a smooth curve, and the curvature is smaller (suitable for the vehicle to follow), and is further away from a fixed obstacle such as a wall, based on which the smooth term gradient, the curvature term gradient, and the obstacle term gradient of the curve may be considered in the optimization. In summary, for each desired path point, the target gradient corresponding to the desired path point may be determined based on the smooth term gradient of the desired path point, the curvature term gradient of the desired path point, and the obstacle term gradient of the desired path point. For example, the sum of the smooth term gradient, the curvature term gradient, and the obstacle term gradient may be taken as the target gradient.

For the smoothing term gradient of the desired path point, for the curve smoothing term, the current point, the first two points and the second two points on the curve can be considered, and five points in total form the smoothing term, for example, the smoothing term gradient of the desired path point can be determined by the following formula, which is, of course, merely an example and is not limited thereto.

Referring to FIG. 4A, a schematic diagram of a determination of a smooth term gradient for a current point x _i Is obtained from four sets of variation calculated from five points back and forth (each set of variation is composed of (x, y) gradients).

Referring to FIG. 4A, x _i Representing the desired path point, x _i-1 Representing a first desired path point, x, preceding the desired path point _i-2 Representing a second desired path point, x, preceding the desired path point _i+1 Representing a first desired path point, x, following the desired path point _i+2 Representing a second desired path following the desired path pointAnd (5) a diameter point.

See FIG. 4A, deltax _i+2 Representing a vector between a first desired path point subsequent to the desired path point and a second desired path point subsequent to the desired path point. Δx _i+1 Representing a vector between the desired path point and a first desired path point subsequent to the desired path point. Δx _i Representing a vector between a first desired path point preceding the desired path point and the desired path point. Δx _i-1 Representing a vector between a second desired path point preceding the desired path point and a first desired path point preceding the desired path point.

In the above formula, grad _smooth A smoothed term gradient representing the desired path point.

Referring to the above formula, since the coordinates of each desired path point, i.e., the known desired path point x, can be obtained _i 、x _i-1 、x _i-2 、x _i+1 、x _i+2 Thus, Δx can be obtained _i+2 、Δx _i+1 、Δx _i 、Δx _i-1 Equal vectors, and then the expected path point x can be obtained _i Is a smoothed term gradient grad of (2) _smooth The determination process is not described in detail.

For the curvature item gradient of the expected path point, for the curvature item of the curve, the current point, the leading point and the trailing point on the curve can be considered, and the curvature item can be formed by three points in total, for example, the curvature item gradient of the expected path point can be determined by adopting the following formula, which is, of course, only an example and is not limited thereto.

Referring to FIG. 4A, a schematic diagram of determining the gradient of the curvature term, after converting the curvature term into the above formula, ΔΦ _i Concerning x _i-1 ,x _i ,x _i+1 The partial derivative term of (2) may be calculated from a triangle formed by the current point, the preceding point, and the following point, and in fig. 4A, the triangle formed by these three points is shown by the dashed line.

See FIG. 4A, deltax _i Representing a vector between a first desired path point preceding the desired path point and the desired path point. ΔΦ of _i And the included angle between the first straight line and the second straight line is represented, the first straight line is the connection line between the first expected path point in front of the expected path point and the expected path point, and the second straight line is the connection line between the first expected path point behind the expected path point and the expected path point. X is x _i Representing the desired path point. X is x _i-1 Representing a first desired waypoint preceding the desired waypoint. X is x _i+1 Representing a first desired path point following the desired path point. In the above formula, grad _curvature A gradient of curvature terms representing the desired path point.

Referring to the above formula, since the coordinates of each desired path point, i.e., the known desired path point x, can be obtained _i 、x _i-1 、x _i+1 Thus, Δx can be obtained _i Equal vectors and obtain DeltaPhi _i Equal angles, then the desired path point x can be obtained _i Curvature term gradient grad of (2) _curvature The determination process is not described in detail.

For the obstacle item gradient of the desired path point, in a first calculation mode of the obstacle item gradient, the mode of description of the obstacle uses an occupancy-value grid map, and the occupancy-value grid map is defined as follows: when the grid is an obstacle, the occupancy value is 0, and when the grid is a non-obstacle, the occupancy value is the grid offset from the nearest obstacle. In this expression, the obstacle term gradient of the desired path point may be determined using the following formula, which is, of course, merely an example, and is not limited thereto.

In the above formula, x represents an abscissa corresponding to the desired path point in the raster pattern, and y represents an ordinate corresponding to the desired path point in the raster pattern; epsilon represents the configured value; f (x, y) represents the grid offset of (x, y) from the nearest obstacle in the grid map; grad _obs An obstacle item gradient representing the desired path point.

Referring to FIG. 4B, a diagram of an obstacle gradient is shown in the description of an occupancy numerical grid map for the desired path point x _i X represents the desired path point x _i On the abscissa corresponding to the raster pattern, y represents the desired path point x _i On the ordinate corresponding to the grid pattern. Epsilon represents a configured value and can be configured empirically.

Wherein a grid of increasing epsilon size along the X-axis, i.e. f (x+epsilon, y), is considered, this grid representing the desired path point X _i Grid offset from the nearest obstacle in the grid map (i.e., nearest obstacle to the right). A grid of decreasing epsilon size along the X-axis, i.e. f (X-epsilon, y), can be considered, this grid representing the desired path point X _i Grid offset from the nearest obstacle in the grid map (i.e., nearest obstacle to the left).

Wherein a grid of increasing epsilon size along the Y-axis, i.e. f (x, Y + epsilon), is considered, this grid representing the desired path point x _i Grid offset from the nearest obstacle in the grid map (i.e., the upper nearest obstacle). A grid of decreasing epsilon size along the Y-axis, i.e. f (x, Y-epsilon), can be considered, this grid representing the desired path point x _i Grid offset from the nearest obstacle in the grid map (i.e., the nearest obstacle on the lower side).

Based on the above-mentioned value magnitude variation of the 4 grid offsets, the gradient in the X-axis direction and the gradient in the Y-axis direction, which are the obstacle item gradient grad, can be calculated _obs 。

For the obstacle item gradient of the desired path point, in a second calculation mode of the obstacle item gradient, the mode of obstacle description uses a vector form. In this expression, the obstacle term gradient for the desired path point may be determined using the following formula, which is, of course, merely an example and is not limiting.

In the above formula, d _obs Representing the closest distance of the desired path point to the obstacle; x represents the abscissa of the desired path point and y represents the ordinate of the desired path point; const represents the configured furthest distance value; d, d _{min_consider} Representing const and d _obs A difference between them; grad _obs An obstacle item gradient representing the desired path point.

Referring to FIG. 4C, a schematic diagram of the gradient of the obstacle for the desired path point x is shown in the vector diagram _i X represents the abscissa of the desired path point and y represents the ordinate of the desired path point. In this way, the obstacle is described using a polygon (closed) or a multi-segment line (non-closed), d _obs Representing the desired path point x _i The closest distance to the obstacle, const, represents the furthest distance to be considered, and may be a constant, e.g., the constant value takes 1m, etc. On the basis, if the path point x is expected _i D when the distance between the barrier and the vehicle is greater than 1m _{min_consider} The value of (2) is negative and the max term in the gradient calculation function will not work if the desired path point x _i D when the distance between the barrier and the barrier is less than 1m _{min_consider} The value of (2) is positive and the gradient term is valid.

In summary, for each desired path point, a smooth term gradient of the desired path point, a curvature term gradient of the desired path point, and an obstacle term gradient of the desired path point may be calculated, and then a target gradient corresponding to the desired path point may be determined based on the smooth term gradient, the curvature term gradient, and the obstacle term gradient.

Step 205, the target vehicle optimizes the plurality of expected path points based on the target gradients corresponding to the plurality of expected path points, and obtains a plurality of candidate path points corresponding to the plurality of expected path points. For example, for each expected path point, the expected path point is optimized based on the target gradient corresponding to the expected path point, so as to obtain the candidate path point corresponding to the expected path point, and the optimization mode of the expected path point is not limited.

For example, based on the target gradient corresponding to the plurality of expected path points, the plurality of expected path points can be optimized by a calculation geometry method to obtain a plurality of candidate path points corresponding to the plurality of expected path points, and the connecting line of the plurality of candidate path points is a smooth curve. The calculation geometry method may include, but is not limited to, polynomial curves, B-spline curves, etc., and the optimization method of the calculation geometry method is not limited thereto.

Step 206, the target vehicle determines whether the plurality of candidate route points satisfy the convergence condition.

If so, step 207 is performed. If not, the plurality of candidate waypoints may be used as a plurality of desired waypoints, and step 204 is performed back, i.e., the target gradients corresponding to the plurality of desired waypoints are redetermined.

For example, whether the convergence condition is satisfied may be determined based on the target gradient corresponding to each expected path point, if the target gradients corresponding to all expected path points are smaller than a threshold, the convergence condition is determined to be satisfied, otherwise, it is determined that the convergence condition is not satisfied. Or, if the target gradient corresponding to a part of the expected path points (such as 80%, 90% of the total number) is smaller than the threshold value, determining that the convergence condition is met, otherwise, determining that the convergence condition is not met.

For another example, if the number of iterations of the desired path point (i.e., the number of repetitions of steps 204-206) reaches a preset threshold, then it is determined that the convergence condition is met, otherwise it is determined that the convergence condition is not met.

In step 207, the target vehicle generates a reference path based on the plurality of candidate path points, for example, a curve formed by the plurality of candidate path points is used as the reference path, and the generation mode of the reference path is not limited.

Step 208, the target vehicle may generate a target path based on the reference path.

For example, after obtaining the reference path, a target path may be generated based on the reference path, the target path being a travel locus of the target vehicle, that is, a real-time travel locus of the target vehicle. In generating the target path of the target vehicle, the factors of smoothness, curvature and obstacles may also be considered, the smoothness of the target path is considered by an optimization function, and the curvature and obstacles of the target path are represented by constraints.

In one possible implementation, the target path may be generated by:

step 2081, selecting a plurality of reference path points from the reference paths, and acquiring a plurality of path segments based on the plurality of reference path points, wherein each path segment comprises a first reference path point and a second reference path point which are adjacent to each other.

For example, the reference paths may be sampled discretely, such as sampling a path point as a reference path point every predetermined length (e.g., 0.5 m), so that a plurality of reference path points, such as a reference path point p1, a reference path point p2, a reference path point p3, a reference path point p4, etc., may be obtained, and so on. The method comprises the steps of acquiring a plurality of path segments based on a plurality of reference path points, wherein the path segment s1 comprises a reference path point p1 and a reference path point p2 which are adjacent, the reference path point p1 is used as a first reference path point, and the reference path point p2 is used as a second reference path point. The path segment s2 includes a reference path point p2 and a reference path point p3 adjacent to each other, the reference path point p2 being a first reference path point, the reference path point p3 being a second reference path point, and so on.

Step 2082, for each path segment, determining an iteration optimization parameter based on the first reference path point and the second reference path point of the path segment, that is, the iteration optimization parameter corresponding to the path segment.

For example, the first reference path point is used as a starting point of the path segment, the second reference path point is used as an ending point of the path segment, and the following formula can be used to determine the corresponding lateral parameter of the path segment. Of course, the following formula is merely an example of determining the lateral parameter, and the determination of the lateral parameter is not limited thereto.

In the above formula, x _s Represents the transverse coordinates, x, of the first reference path point _s ' first order lateral derivative, x, representing a first reference path point _s "means the second lateral derivative of the first reference path point. X is x _e Representing the transverse coordinates, x, of the second reference path point _e ' first order lateral derivative, x, representing the second reference path point _e "means the second lateral derivative of the second reference path point. a, a ₀ ，a ₁ ，a ₂ ，a ₃ Representing the corresponding lateral parameters of the path segment.

A of the above formula ₀ ，a ₁ ，a ₂ ，a ₃ Replaced by b ₀ ，b ₁ ，b ₂ ，b ₃ Will x _s Longitudinal coordinate y replaced by first reference path point _s Will x _s ' first longitudinal derivative y replaced by first reference path point _s ' x is _s "second-order longitudinal derivative y replaced by first reference path point _s ", x is _e Longitudinal coordinate y replaced by second reference path point _e Will x _e ' first longitudinal derivative y replaced by second reference path point _e ' x is _e "second-order longitudinal derivative y replaced by second reference path point _e ", obtaining the longitudinal parameter b corresponding to the path segment ₀ ，b ₁ ，b ₂ ，b ₃ 。

Obviously, the transverse parameter a ₀ ，a ₁ ，a ₂ ，a ₃ And longitudinal parameter b ₀ ，b ₁ ，b ₂ ，b ₃ The iterative optimization parameters corresponding to the path segment can be formed, so that the iterative optimization parameters corresponding to the path segment are obtained.

Step 2083, determining a candidate polynomial curve corresponding to the path segment based on the iterative optimization parameter and the configured value interval, wherein the value interval may include a plurality of values between 0 and 1.

For example, the calculation of the smoothing term may take into account the minimization of the first and second derivatives of the multi-segment cubic spline helical curve, based on which, for a real-time trajectory between every two discrete points (i.e., path segments between every two reference path points), a segmented cubic spline representation of the XY space may be used, i.e., the candidate polynomial curve corresponding to each path segment is represented by a segmented cubic spline of the XY space.

For example, for each path segment, the candidate polynomial curve corresponding to that path segment may be determined using the following formula: x= (1, s) ² ,s ³ )(a ₀ ,a ₁ ,a ₂ ,a ₃ ) ^T ，y＝(1,s,s ² ,s ³ )(b ₀ ,b ₁ ,b ₂ ,b ₃ ) ^T . Wherein a is ₀ ，a ₁ ，a ₂ ，a ₃ Representing the corresponding transverse parameter of the path segment, b ₀ ，b ₁ ，b ₂ ，b ₃ Representing the corresponding longitudinal parameter of the path segment.

When s is 0, x represents the abscissa of the starting point of the candidate polynomial curve, and y represents the ordinate of the starting point of the candidate polynomial curve, that is, the starting point coordinate after the first reference path point of the path segment is corrected.

When s is 1, x represents the end abscissa of the candidate polynomial curve, and y represents the end ordinate of the candidate polynomial curve, that is, the end coordinate after correcting the second reference path point of the path segment.

When s is a number between 0 and 1, such as 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, etc., a plurality of coordinate points may be obtained, and these coordinate points may constitute the candidate polynomial curve.

In summary, for each path segment, the iterative optimization parameters (a) corresponding to the path segment are based on ₀ ，a ₁ ，a ₂ ，a ₃ ，b ₀ ，b ₁ ，b ₂ ，b ₃ ) And a plurality of values in the value interval 0-1, so that a candidate polynomial curve corresponding to the path segment can be determined, and the candidate polynomial curve can be uniquely determined by the 0-2-order values of the two end points.

Step 2084, determining a first curvature constraint parameter and a second curvature constraint parameter based on the candidate polynomial curve, and determining a target curvature based on the first curvature constraint parameter and the second curvature constraint parameter.

For example, the first curvature constraint parameter, the second curvature constraint parameter, and the target curvature may be determined based on the following formula, which is, of course, merely an example and is not limited in this manner.

In the above formula, x 'represents the first lateral derivative of the target point on the candidate polynomial curve, and the target point may be any point on the candidate polynomial curve, y' represents the first longitudinal derivative of the target point, x "represents the second lateral derivative of the target point, y" represents the second longitudinal derivative of the target point, C ₁ Representing the first curvature constraint parameter, C ₂ Representing the second curvature constraint parameter, K representing the target curvature.

Obviously, any point can be selected from the candidate polynomial curve as the target point, x represents the abscissa of the target point, and y represents the ordinate of the target point. Then, the first lateral derivative x ' and the first longitudinal derivative y ' of the target point may be calculated, and the second lateral derivative x″ and the second longitudinal derivative y ' of the target point may be calculated. Based on the above information, a first curvature constraint parameter, a second curvature constraint parameter, and a target curvature can be determined.

For example, one extreme nonlinear constraint can be found in the following expression:the expression can be equivalently expressed as the three expressions (i.e., K, C ₁ ，C ₂ ). In the problem solving process, the constraint term can be considered as a term to be activated. If the path calculated by the current optimization problem does not exceed the limiting curvature, the term optimization may not be activated. When the curvature exceeds the limit, the optimization term needs to be started, the first derivative term of the optimization path with the curvature exceeding the limit is taken as a constant, and the second derivative is optimizedAn item. In this case, the first derivative term of the trajectory may be used as a constant, and the constraint problem conforming to the curvature may be calculated.

Step 2085, determining whether the target curvature is less than a preset threshold (which may be empirically configured). If so, then step 2086 may be performed; if not, step 2087 may be performed.

For example, if the target curvatures of the candidate polynomial curves corresponding to all the path segments are smaller than the preset threshold, step 2086 is performed. For each path segment, if the target curvature of the candidate polynomial curve corresponding to the path segment is not less than the preset threshold, step 2087 is executed for the path segment, otherwise, step 2087 is not executed for the path segment until the target curvature of all the candidate polynomial curves is less than the preset threshold.

Step 2086, taking the candidate polynomial curve as the target path. For example, candidate polynomial curves corresponding to all path segments may be combined to obtain the target path of the target vehicle.

Step 2087, for each path segment, if the target curvature of the candidate polynomial curve corresponding to the path segment is not less than the preset threshold, the first point (i.e., the starting point) and/or the last point (i.e., the ending point) of the candidate polynomial curve are adjusted to obtain an adjusted curve, the first point of the adjusted curve is taken as a first reference path point, the last point of the adjusted curve is taken as a second reference path point, and step 2082 is executed in a return manner, i.e., the adjusted curve is taken as a path segment, and the iterative optimization parameters corresponding to the path segment are recalculated.

In one possible implementation, the curvature and the obstacle of the target path may be represented by constraints, as regards the curvature constraints of the target path, see steps 2081-2087, as regards the obstacle constraints of the target path, the obstacle constraints may be described using an affine transformation expression as follows: ax is less than or equal to b. The affine transformation expression is used to represent that for each trajectory point to be optimized, it is constrained in one affine transformation space, as shown in fig. 5. The optimization target and constraint of the problem conform to the specification of the quadratic programming problem, and a quadratic programming algorithm can be used for optimization solution.

To this end, step 208 is completed, and a target path of the target vehicle may be obtained.

Step 209, the target vehicle controls the target vehicle to travel based on the target path.

For example, when the target path is at the center position, the control target vehicle travels based on the first speed, i.e., the control target vehicle travels centrally based on the first speed. When the target path is located at the right position, the control target vehicle travels based on the second speed, that is, the control target vehicle travels right to right based on the second speed.

Illustratively, the first speed may be greater than the second speed, e.g., the first speed may be 15km/h (e.g., the highest speed), and the second speed may be 8km/h (e.g., the system vigilance speed). Obviously, a central expected track or a right expected track can be calculated, when the central track is generated, the speed of the vehicle is increased to a first speed, and when the right track is generated, the speed of the vehicle is reduced to a second speed.

Centered travel means: in the cruising process of memory parking, the target vehicle increases the running speed in an open scene, and runs in the center of multiple lanes, and is driven automatically without considering the traffic rules of the lanes, and the target vehicle is far away from dynamic obstacles which may suddenly appear on two sides, so that sufficient reaction time is reserved. Right-to-travel means: in the cruising process of memory parking, the target vehicle is in order to avoid possible dynamic or static barriers in a complex scene, obeys the traffic rules indicated by ground marking, walks right to the right and reduces the speed of a vehicle, and improves the safety.

According to the technical scheme, in the embodiment of the application, the centering path can be adaptively generated, the vehicle is controlled to run on the centering path, the running speed of the vehicle can be improved, the vehicle is controlled to run fast, and the running efficiency is improved. When the speed of the vehicle is increased, the foot reaction time (namely, the interval between the centering path and the right side is larger, and the foot reaction time can be reserved) is reserved for pedestrians or vehicles which suddenly appear on the right side, so that no collision is ensured. The method can adaptively generate a right-leaning path, control the vehicle to run on the right-leaning path, reduce the running speed of the vehicle, and ensure the running safety of the vehicle, namely, ensure no collision by reducing the running speed of the vehicle. Referring to fig. 6, when there is no dynamic vehicle in front, the target vehicle can run in the middle, and when the vehicle speed is improved, the reaction time is reserved for the situations such as pedestrians suddenly appearing on the right side, so that no collision is ensured. When dynamic vehicles exist in front, the target vehicle is right to travel, the speed is reduced, and smooth passing of meeting vehicles is ensured.

Based on the same application concept as the above method, an automatic driving device is provided in the embodiment of the present application, which is applied to a target vehicle, and is shown in fig. 7, and is a schematic structural diagram of the device, where the device includes:

A determining module 71, configured to select a plurality of taught path points from a taught path after the target vehicle starts automatic driving, and determine an expected path point corresponding to each taught path point; wherein the target vehicle stores a teaching path; wherein, for each taught path point, if it is determined that no obstacle exists on the left side and/or the right side of the taught path point based on the perception information, the taught path point is taken as a desired path point; if the first obstacle exists on the left side of the teaching path point and the second obstacle exists on the right side of the teaching path point based on the perception information, the teaching path point is taken as an expected path point when the distance between the first obstacle and the second obstacle is smaller than a threshold value; when the distance is not smaller than the threshold value, taking the center point between the first obstacle and the second obstacle as an expected path point; a generating module 72, configured to generate a target path based on a plurality of expected path points corresponding to the plurality of taught path points; a control module 73 for controlling the target vehicle to travel based on the target path; wherein the control target vehicle travels based on a first speed when the target path is located at the center position, and travels based on a second speed when the target path is located at the right position, and the first speed is greater than the second speed.

For example, if the target path corresponds to at least two lanes, the center position is a center position of all lanes, and the right position is a right lane position; for each taught path point, the determining module 71 is specifically configured to, when determining a desired path point corresponding to the taught path point:

determining a target lane model corresponding to the teaching path point; if the right side of the teaching path point is determined to have no obstacle based on the perception information, determining that the target lane model is a lane-free model; if the left side of the teaching path point is determined to have no obstacle based on the perception information, determining that the target lane model is a single lane model; if the first obstacle exists on the left side and the second obstacle exists on the right side of the teaching path point based on the perception information, determining that the target lane model is a single lane model when the distance between the first obstacle and the second obstacle is smaller than a threshold value; when the distance is not smaller than a threshold value, determining that the target lane model is a double lane model;

determining a desired path point based on the target lane model; if the target lane model is a lane-free model or a single lane model, the teaching path point is used as an expected path point; if the target lane model is a two-lane model, a center point between the first obstacle and the second obstacle is taken as a desired path point.

Illustratively, the generating module 72 is specifically configured to, when generating the target path based on the plurality of desired path points corresponding to the plurality of taught path points: determining target gradients corresponding to the plurality of expected path points; for each desired path point, a target gradient corresponding to the desired path point is determined based on at least one of a smoothed term gradient of the desired path point, a curvature term gradient of the desired path point, and an obstacle term gradient of the desired path point; optimizing the plurality of expected path points based on target gradients corresponding to the plurality of expected path points to obtain a plurality of candidate path points corresponding to the plurality of expected path points; if the plurality of candidate path points meet the convergence condition, generating a reference path based on the plurality of candidate path points, and generating the target path based on the reference path; and if the plurality of candidate path points do not meet the convergence condition, taking the plurality of candidate path points as a plurality of expected path points, and returning to the operation of determining the target gradients corresponding to the plurality of expected path points.

Illustratively, the generating module 72 is specifically configured to, when generating the target path based on the reference path: selecting a plurality of reference path points from the reference path, and acquiring a plurality of path segments based on the plurality of reference path points, wherein each path segment comprises a first reference path point and a second reference path point which are adjacent to each other; for each path segment, determining an iterative optimization parameter based on a first reference path point and a second reference path point of the path segment; determining a candidate polynomial curve corresponding to the path segment based on the iterative optimization parameter and a configured numerical interval, wherein the numerical interval comprises a plurality of numerical values between 0 and 1; a target path is determined based on the candidate polynomial curve.

Illustratively, the generating module 72 is specifically configured to, when determining the target path based on the candidate polynomial curve: determining a first curvature constraint parameter and a second curvature constraint parameter based on the candidate polynomial curve, and determining a target curvature based on the first curvature constraint parameter and the second curvature constraint parameter; if the target curvature is smaller than a preset threshold value, the candidate polynomial curve is used as a target path; and if the target curvature is not smaller than a preset threshold value, adjusting the first point and/or the last point of the candidate polynomial curve to obtain an adjusted curve, taking the first point of the adjusted curve as a first reference path point, taking the last point of the adjusted curve as a second reference path point, and returning to execute the operation of determining the iterative optimization parameter based on the first reference path point and the second reference path point of the path section.

Based on the same application concept as the above method, an automatic driving vehicle is provided in an embodiment of the present application, and as shown in fig. 8, the automatic driving vehicle includes: a processor 81 and a machine-readable storage medium 82, the machine-readable storage medium 82 storing machine-executable instructions executable by the processor 81; the processor 81 is configured to execute machine-executable instructions to implement the autopilot method disclosed in the above examples of the application.

Based on the same application concept as the above method, the embodiment of the present application further provides a machine-readable storage medium, where a number of computer instructions are stored, where the computer instructions can implement the automatic driving method disclosed in the above example of the present application when the computer instructions are executed by a processor.

Wherein the machine-readable storage medium may be any electronic, magnetic, optical, or other physical storage device that can contain or store information, such as executable instructions, data, or the like. For example, a machine-readable storage medium may be: RAM (Radom Access Memory, random access memory), volatile memory, non-volatile memory, flash memory, a storage drive (e.g., hard drive), a solid state drive, any type of storage disk (e.g., optical disk, dvd, etc.), or a similar storage medium, or a combination thereof.

The system, apparatus, module or unit set forth in the above embodiments may be implemented in particular by a computer or an entity, or by an article of manufacture having a certain function. A typical implementation device is a computer, which may be in the form of a personal computer, laptop computer, cellular telephone, camera phone, smart phone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or a combination of any of these devices.

For convenience of description, the above devices are described as being functionally divided into various units, respectively. Of course, the functions of each element may be implemented in the same piece or pieces of software and/or hardware when implementing the present application.

It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the application may take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and variations of the present application will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. which come within the spirit and principles of the application are to be included in the scope of the claims of the present application.

Claims

1. An autopilot method for application to a target vehicle, the target vehicle storing a taught path, the method comprising, after the target vehicle initiates autopilot:

selecting a plurality of teaching path points from the teaching paths;

2. The method of claim 1, wherein if the target path corresponds to at least two lanes, the center position is a center position of all lanes and the right position is a right lane position;

for each teaching path point, determining the expected path point corresponding to the teaching path point comprises the following steps:

3. The method of claim 1, wherein the step of determining the position of the substrate comprises,

the generating a target path based on a plurality of expected path points corresponding to the plurality of taught path points includes:

determining target gradients corresponding to the plurality of expected path points; wherein, for each desired path point, a target gradient corresponding to the desired path point is determined based on at least one of a smoothed term gradient of the desired path point, a curvature term gradient of the desired path point, and an obstacle term gradient of the desired path point;

optimizing the plurality of expected path points based on target gradients corresponding to the plurality of expected path points to obtain a plurality of candidate path points corresponding to the plurality of expected path points;

if the plurality of candidate path points meet the convergence condition, generating a reference path based on the plurality of candidate path points, and generating the target path based on the reference path;

And if the plurality of candidate path points do not meet the convergence condition, taking the plurality of candidate path points as a plurality of expected path points, and returning to execute the operation of determining the target gradient corresponding to the plurality of expected path points.

4. The method of claim 3, wherein the step of,

for each desired path point, the determination of the smoothed term gradient for that desired path point includes:

the smoothed term gradient for the desired path point is determined using the following formula:

wherein Deltax is _i+2 A vector representing a distance between a first desired path point subsequent to the desired path point and a second desired path point subsequent to the desired path point; Δx _i+1 A vector representing a distance between the desired path point and a first desired path point subsequent to the desired path point; Δx _i A vector representing a first desired path point preceding the desired path point and the desired path point; Δx _i-1 A vector representing a distance between a second desired path point preceding the desired path point and a first desired path point preceding the desired path point; x is x _i For representing the desired path point; grad _smooth A smoothed term gradient representing the desired path point.

5. The method of claim 3, wherein the step of,

For each desired path point, the determination of the curvature item gradient for that desired path point includes:

the curvature term gradient for the desired path point is determined using the following formula:

wherein Deltax is _i A vector representing a first desired path point preceding the desired path point and the desired path point; ΔΦ of _i For indicating the angle between a first line, which is the line connecting the first desired path point before the desired path point, and a second line, which is the line following the desired path pointA first desired path point of the face is connected with the desired path point; x is x _i For representing the desired path point; x is x _i-1 For representing a first desired waypoint preceding the desired waypoint; x is x _i+1 For representing a first desired path point subsequent to the desired path point; grad _curvature A gradient of curvature terms representing the desired path point.

6. The method of claim 3, wherein the step of,

for each desired waypoint, the determination of the obstacle item gradient for that desired waypoint includes:

the obstacle term gradient for the desired path point is determined using the following formula:

wherein x represents an abscissa of the expected path point corresponding to the grid graph, and y represents an ordinate of the expected path point corresponding to the grid graph; epsilon represents the configured value; f (x, y) represents the grid offset of (x, y) from the nearest obstacle in the grid map; grad _obs An obstacle item gradient representing the desired path point.

7. The method of claim 3, wherein the step of,

wherein d _obs Representing the closest distance of the desired path point to the obstacle; x represents the abscissa of the desired path point and y represents the ordinate of the desired path point; const represents the configured furthest distanceSeparating value; d, d _{min_consider} Representing const and d _obs A difference between them; grad _obs An obstacle item gradient representing the desired path point.

8. The method of claim 3, wherein the step of,

the generating the target path based on the reference path includes:

selecting a plurality of reference path points from the reference path, and acquiring a plurality of path segments based on the plurality of reference path points, wherein each path segment comprises a first reference path point and a second reference path point which are adjacent;

for each path segment, determining an iterative optimization parameter based on a first reference path point and a second reference path point of the path segment; determining a candidate polynomial curve corresponding to the path segment based on the iterative optimization parameter and a configured numerical interval, wherein the numerical interval comprises a plurality of numerical values between 0 and 1;

The target path is determined based on the candidate polynomial curve.

9. The method of claim 8, wherein the step of determining the position of the first electrode is performed,

the determining the target path based on the candidate polynomial curve includes:

determining a first curvature constraint parameter and a second curvature constraint parameter based on the candidate polynomial curve, and determining a target curvature based on the first curvature constraint parameter and the second curvature constraint parameter;

if the target curvature is smaller than a preset threshold value, the candidate polynomial curve is used as a target path;

and if the target curvature is not smaller than a preset threshold value, adjusting the first point and/or the last point of the candidate polynomial curve to obtain an adjusted curve, taking the first point of the adjusted curve as a first reference path point, taking the last point of the adjusted curve as a second reference path point, and returning to execute the operation of determining the iterative optimization parameter based on the first reference path point and the second reference path point of the path segment.

10. The method of claim 9, wherein the step of determining the position of the substrate comprises,

the determining a first curvature constraint parameter and a second curvature constraint parameter based on the candidate polynomial curve determining a target curvature based on the first curvature constraint parameter and the second curvature constraint parameter comprises:

Determining a first curvature constraint parameter, a second curvature constraint parameter, and a target curvature based on the following formula:

wherein x 'represents a first lateral derivative of the target point on the candidate polynomial curve, the target point being any point on the candidate polynomial curve, y' represents a first longitudinal derivative of the target point, x "represents a second lateral derivative of the target point, y" represents a second longitudinal derivative of the target point, C ₁ Representing the first curvature constraint parameter, C ₂ Representing the second curvature constraint parameter, K representing the target curvature.

11. An autopilot device for application to a target vehicle, the device comprising:

12. An autonomous vehicle, comprising: a processor and a machine-readable storage medium storing machine-executable instructions executable by the processor; the processor is configured to execute machine executable instructions to implement the method of any of claims 1-10.