WO2023221537A1 - Path planning method and apparatus and autonomous vehicle - Google Patents

Abstract

A path planning method and apparatus, an autonomous vehicle, and a computer-readable storage medium. The path planning method comprises: according to a navigation path of an autonomous vehicle, determining a guide line used for assisting in driving the autonomous vehicle (S110); according to the guide line, determining a plurality of sampling lines for assisting in multi-layer position point sampling, each sampling line corresponding to sampling of one layer of position points (S120); on the basis of a current position point of the autonomous vehicle, sampling at least one candidate position point from each sampling line by using information related to vehicle dynamics (S130); from the plurality of candidate position points located on the plurality of sampling lines, screening out a plurality of target position points at which the autonomous vehicle successively arrives from the current position point (S140); and according to the current position point and the plurality of target position points, determining a current frame path of the autonomous vehicle (S150). The path planning method can improve the stability of a planned path.

Description

Path planning method and device, unmanned vehicle

Cross-references to related applications

This application is based on the application with CN application number 202210528039.5 and the filing date is May 16, 2022, and claims its priority. The disclosure content of the CN application is hereby incorporated into this application as a whole.

Technical field

The present disclosure relates to the field of computer technology, particularly to the field of unmanned driving, and in particular to path planning methods and devices, unmanned vehicles, and computer storage media.

Background technique

Currently, autonomous driving equipment is used to automatically transport people or objects from one location to another. The autonomous driving equipment collects environmental information through sensors on the equipment and completes automatic transportation. Logistics and transportation using unmanned delivery vehicles controlled by autonomous driving technology has greatly improved the convenience of production and life and saved labor costs. In the field of autonomous driving, path planning is a basic task.

In related technology, in the path planning process in the field of autonomous driving, candidate location points are obtained using uniform sampling, and multiple target location points are screened from the candidate location points, and then based on the current location point of the autonomous driving device and multiple targets Position points determine the current frame path of the autonomous driving device.

Contents of the invention

According to a first aspect of the present disclosure, a path planning method for autonomous driving is provided, including: determining a guide line for assisting the driving of the unmanned vehicle according to the navigation path of the unmanned vehicle; , determine multiple sampling lines used to assist multi-layer position point sampling, each sampling line corresponds to a layer of position point sampling; based on the current position point of the unmanned vehicle, using vehicle dynamics related information, from each sampling line Sampling on the line obtains at least one candidate position point; from multiple candidate position points located on multiple sampling lines, screen multiple target position points arriving sequentially starting from the current position point; according to the current position point and the The multiple target position points are used to determine the current frame path of the unmanned vehicle.

In some embodiments, based on the current position point, using vehicle dynamics related information, sampling from each sampling line to obtain at least one candidate position point includes: for each sampling line, using vehicle dynamics related information, determining The reference position point corresponding to each sampling line is at least one reference path starting from the starting point, wherein, when each sampling line is the sampling line closest to the current position point, the reference position The point is the current position point. When each sampling line is another sampling line, the reference position point is a sampling line located adjacent to each sampling line and close to the direction of the current position point. each candidate position point on the; determine at least one candidate position point located on each sampling line according to the intersection point of each sampling line and the at least one reference path.

In some embodiments, for the sampling line closest to the current position point, vehicle dynamics related information is used to determine at least one reference path starting from the current position point; according to the distance between the current position point and the sampling line The intersection point of the nearest sampling line and at least one reference path starting from the current position point determines at least one candidate position point located on the sampling line closest to the current position point; for each other sampling line , using the vehicle dynamics related information to determine at least one reference path starting from each candidate position point on the sampling line adjacent to each other sampling line and closest to the current position point ; According to each other sampling line and at least one reference path starting from each candidate position point on the sampling line adjacent to each other sampling line and closest to the current position point The intersection point determines at least one candidate position point located on each of the other sampling lines.

In some embodiments, the vehicle dynamics-related information includes vehicle dynamics attribute information or vehicle motion trajectory information. The vehicle dynamics-related information is used to determine at least one point starting from the reference position point corresponding to each sampling line. A reference path includes: using the vehicle dynamics attribute information or vehicle motion trajectory information to determine at least one reference path starting from the reference position point corresponding to each of the sampling lines.

In some embodiments, the path planning method further includes: sampling the angle range of the unmanned vehicle to obtain multiple reference angles, wherein the reference path includes a path under the multiple reference angles.

In some embodiments, based on the current position point of the unmanned vehicle, using vehicle dynamics related information, sampling from each sampling line to obtain at least one candidate position point further includes: for each sampling line, determining the unmanned vehicle. The intersection point of the previous frame path of the person and vehicle and each of the sampling lines is a candidate position point located on each of the sampling lines.

In some embodiments, the candidate position point on each sampling line is an intersection point that satisfies a first preset motion condition. The first preset motion condition includes a reference position point corresponding to each sampling line and the The unmanned vehicle orientation difference between the candidate location points on each sampling line is less than or equal to the unmanned vehicle orientation difference threshold.

In some embodiments, screening multiple target location points arriving sequentially from the current location point from multiple candidate location points located on multiple sampling lines includes: based on the current location point, located on the multiple Multiple candidate location points and preset driving costs on a sampling line are used to determine the multiple target location points using a path-finding algorithm, where the stopping conditions of the path-finding algorithm include starting from the current location point and arriving at the last The mileage of a target location point is greater than or equal to the mileage threshold.

In some embodiments, the preset driving cost includes at least one of a smoothness cost and a safety cost, wherein the smoothness cost is related to the location point currently processed by the path-finding algorithm and the candidate location point. There is a positive correlation between the direction difference of the unmanned vehicle between the two, and the safety cost is negatively correlated with the shortest distance from the line segment between the position point currently processed by the path-finding algorithm and the candidate position point to the obstacle. The position point currently processed by the road algorithm is the current position point or the candidate position point.

In some embodiments, when the candidate location point includes the intersection of the previous frame path of the unmanned vehicle and each sampling line, the preset driving cost also includes a path similarity cost, and the path The similarity cost represents the similarity between the path starting from the position point currently processed by the path-finding algorithm and the path in the previous frame.

In some embodiments, the path similarity cost is related to the shortest distance of the path segment between the position point currently processed by the path-finding algorithm and each adjacent position point on the path of the previous frame or the path-finding algorithm. The shortest distance of the path line segment between the previous position point of the position point currently processed by the algorithm and each adjacent position point on the path of the previous frame is positively correlated.

In some embodiments, screening multiple target location points arriving sequentially from the current location point from multiple candidate location points located on multiple sampling lines includes: selecting from multiple candidate location points located on multiple sampling lines. Among the position points, select a plurality of candidate position points that meet the second preset movement condition. The second preset movement condition includes at least one of no collision between the position points and the candidate position point being located within the preset map range. ; Select the plurality of target position points from a plurality of candidate position points that satisfy the second preset motion condition.

According to a second aspect of the present disclosure, a path planning device for autonomous driving is provided, including: a first determination module configured to determine guidance for assisting the driving of the autonomous vehicle according to the navigation path of the autonomous vehicle. line; the second determination module is configured to determine multiple sampling lines for assisting multi-layer sampling based on the guide line, with each sampling line corresponding to one layer of sampling; the sampling module is configured to determine based on the unmanned The current position point of the vehicle uses vehicle dynamics related information to sample at least one candidate position point from each sampling line; the screening module is configured to filter from multiple candidate position points located on multiple sampling lines. The current position point starts from a plurality of target position points that are arrived in sequence; a third determination module is configured to determine the current frame path of the unmanned vehicle based on the current position point and the plurality of target position points.

According to a third aspect of the present disclosure, a path planning device for autonomous driving is provided, including: a memory; and a processor coupled to the memory, the processor being configured to based on instructions stored in the memory , execute the path planning method described in any of the above embodiments.

According to a fourth aspect of the present disclosure, an unmanned vehicle is provided, including: the path planning device for automatic driving described in any of the above embodiments.

In some embodiments, the unmanned vehicle further includes: at least one of a positioning module, a navigation module, a sensing module and a map module, wherein the positioning module is configured to send the speed and current location of the unmanned vehicle. The position coordinates of the point and the direction of the unmanned vehicle to the path planning device; the navigation module is configured to send the navigation path of the unmanned vehicle to the path planning device; the sensing module is configured to sense the unmanned vehicle. obstacles around people and vehicles, and sends the perceived obstacle information to the path planning device; the map module is configured to provide map data to the path planning device.

In some embodiments, the unmanned vehicle further includes: a control module configured to receive the current frame path from the path planning device and control the unmanned vehicle to travel according to the current frame path.

According to a fifth aspect of the present disclosure, a computer-storable medium is provided, on which computer program instructions are stored. When the instructions are executed by a processor, the path planning method described in any of the above embodiments is implemented.

Description of the drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and, together with the description, serve to explain principles of the disclosure.

The present disclosure may be more clearly understood from the following detailed description with reference to the accompanying drawings, in which:

Figure 1 is a flowchart illustrating a path planning method according to some embodiments of the present disclosure;

Figure 2 is a schematic diagram illustrating determining a sampling line according to some embodiments of the present disclosure;

Figure 3A is a schematic diagram showing the results of path planning according to a path planning method based on uniform sampling;

Figure 3B is a schematic diagram showing the results of path planning according to the path planning method of some embodiments of the present disclosure;

Figure 4 is a block diagram illustrating a path planning device according to some embodiments of the present disclosure;

Figure 5 is a block diagram illustrating a path planning device according to other embodiments of the present disclosure;

Figure 6 is a block diagram illustrating an autonomous vehicle according to some embodiments of the present disclosure;

Figure 7 is a block diagram illustrating an autonomous vehicle according to other embodiments of the present disclosure;

Figure 8 is a side view illustrating an autonomous vehicle according to some embodiments of the present disclosure;

Figure 9 is a block diagram illustrating a computer system for implementing some embodiments of the present disclosure.

Detailed ways

Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. It should be noted that the relative arrangement of components and steps, numerical expressions, and numerical values set forth in these examples do not limit the scope of the disclosure unless otherwise specifically stated.

At the same time, it should be understood that, for convenience of description, the dimensions of various parts shown in the drawings are not drawn according to actual proportional relationships.

The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the disclosure, its application or uses.

Techniques, methods and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, such techniques, methods and devices should be considered a part of the specification.

In all examples shown and discussed herein, any specific values are to be construed as illustrative only and not as limiting. Accordingly, other examples of the exemplary embodiments may have different values.

It should be noted that similar reference numerals and letters refer to similar items in the following figures, so that once an item is defined in one figure, it does not need further discussion in subsequent figures.

In related technologies, uniform sampling is used to obtain candidate location points, but path accessibility is poor and may easily lead to path jitter. Based on this, the present disclosure proposes a path planning method that can improve the stability of the planned path.

FIG. 1 is a flowchart illustrating a path planning method according to some embodiments of the present disclosure.

As shown in Figure 1, the path planning method for autonomous driving includes: step S110, determining the guide line for assisting the driving of the unmanned vehicle according to the navigation path of the unmanned vehicle; step S120, determining the guide line for assisting the driving according to the guide line. Multiple sampling lines for multi-layer position point sampling, where each sampling line corresponds to one layer of sampling; step S130, based on the current position point of the unmanned vehicle, using vehicle dynamics related information, sample from each sampling line At least one candidate position point; Step S140, from multiple candidate position points located on multiple sampling lines, screen multiple target position points arriving sequentially starting from the current position point; Step S150, based on the current position point and multiple targets Position point to determine the current frame path of the unmanned vehicle. In some embodiments, the path planning method is executed by a path planning device for autonomous driving.

In the above embodiment, the guidance line matching the navigation path is used as a reference, and vehicle dynamics related information is used to perform multi-layer sampling of candidate location points from multiple sampling lines. By combining guide lines and vehicle dynamics-related information for multi-layer sampling, the accessibility of unmanned vehicles at candidate location points is considered, which can further improve the stability of the planned path.

In step S110, a guidance line for assisting the driving of the unmanned vehicle is determined based on the navigation path of the unmanned vehicle. The guide line is consistent with the direction of the navigation path and is used to guide the driving direction of the unmanned vehicle. In some embodiments, the navigation path is a lane-level navigation path, and the guideline matching the navigation path is determined based on the lane indicated by the navigation path. For example, the coordinate points of the center line of the lane indicated by the navigation path are smoothed to obtain the guide line. In some embodiments, the leader line may be constructed based on the Frenent coordinate system. The Frenet coordinate system describes the position of the unmanned vehicle relative to the road. In the Frenet coordinate system, s represents the distance along the road, which is called the ordinate, and d represents the displacement from the longitudinal line, which is called the abscissa.

In some embodiments, the navigation path can be obtained from the navigation module of the unmanned vehicle.

In step S120, multiple sampling lines used to assist in multi-layer location point sampling are determined based on the guide lines, where each sampling line corresponds to one layer of sampling. For example, multiple sampling lines start from the current position point of the unmanned vehicle and are arranged in sequence along the direction of the guide line. Multiple sampling lines are perpendicular to the leader line. In some embodiments, adjacent sampling lines are equally spaced in the direction of the leader line.

FIG. 2 is a schematic diagram illustrating determining a sampling line according to some embodiments of the present disclosure.

As shown in Figure 2, after the current position point c of the unmanned vehicle is projected onto the guidance line, the mileage value curr_s of the projection point is obtained, thereby determining the position of the projection point on the guidance line. According to the preset step size, the mileage value next_s of the sampling reference point of the preset step length from the current position point on the guide line to the projection point on the guide line can be determined, thereby determining the preset step length of the projection point on the guide line from the current position point. The position of the sampling reference point on the leader line. Figure 2 only shows one sampling reference point. Usually, multiple sampling reference points need to be determined, and the mileage values between adjacent sampling reference points differ by a preset step.

After determining the position of the reference sampling point on the guide line, the vertical line perpendicular to the guide line and intersecting the guide line at the reference sampling point is determined as the sampling line. Figure 2 only shows one sampling line. Usually, multiple sampling lines need to be determined to achieve multi-layer sampling.

In step S130, based on the current position point of the unmanned vehicle and using vehicle dynamics related information, at least one candidate position point is sampled from each sampling line. For example, the current position point of the unmanned vehicle can be obtained from the positioning module of the unmanned vehicle.

In some embodiments, the above step S130 can be implemented in the following manner.

First, for each sampling line, vehicle dynamics related information is used to determine at least one reference path starting from the reference position point corresponding to each sampling line. When each sampling line is the sampling line closest to the current position point, the reference position point is the current position point. In the case where each sampling line is another sampling line, the reference position point is each candidate position point located on the sampling line adjacent to each sampling line and close to the direction of the current position point.

Then, at least one candidate position point located on each sampling line is determined based on the intersection point of each sampling line and at least one reference path.

In the above embodiment, vehicle dynamics-related information is used to determine the reference path that the unmanned vehicle may travel, and the intersection of the reference path and the sampling line is used to implement the process of sampling candidate location nodes from the sampling line. In this way, the path accessibility of unmanned vehicles in the path planning process is further considered, thereby further improving the stability of the planned path. By improving path stability, the traffic capacity of unmanned vehicles or self-driving vehicles can be improved, making it easier for downstream control modules to track trajectories, avoiding the phenomenon of drawing a dragon, reducing misjudgments by other traffic participants, reducing traffic congestion or collisions, and improving automatic control. Driving safety.

In some embodiments, the above process of sampling candidate location points can be implemented in the following manner.

First, for the sampling line closest to the current position point, vehicle dynamics related information is used to determine at least one reference path starting from the current position point. Furthermore, at least one candidate position point located on the sampling line closest to the current position point is determined based on the intersection point of the sampling line closest to the current position point and at least one reference path starting from the current position point.

Then, for each other sampling line, vehicle dynamics related information is used to determine at least one candidate location point starting from each candidate location point on the sampling line adjacent to each other sampling line and closest to the current location point. Reference path. Furthermore, based on the intersection point of each other sampling line and at least one reference path starting from each candidate location point on the sampling line adjacent to each other sampling line and closest to the current location point, determine the location at each At least one candidate location point on other sampling lines.

In some embodiments, the candidate position points on each sampling line are intersection points that satisfy the first preset motion condition. The first preset movement condition includes that the unmanned vehicle orientation difference between the reference position point corresponding to each sampling line and the candidate position point on each sampling line is less than or equal to the unmanned vehicle orientation difference threshold. For example, the threshold for the orientation difference of an autonomous vehicle is 90 degrees. The unmanned vehicle orientation difference threshold can also be other values. By limiting the orientation difference of the unmanned vehicle between the reference position point and the candidate position point, candidate position points that do not conform to the movement patterns of the unmanned vehicle can be filtered, thereby further improving path accessibility and path stability.

In some embodiments, the vehicle dynamics related information includes vehicle dynamics attribute information or vehicle motion trajectory information. The vehicle dynamics attribute information or vehicle motion trajectory information can be used to determine the reference position point corresponding to each sampling line as the starting point. at least one reference path.

Taking the vehicle dynamics related information as an example, including the preset speed of the unmanned vehicle, the direction of the unmanned vehicle, the turning angle of the unmanned vehicle and the wheelbase of the unmanned vehicle, according to the preset speed of the unmanned vehicle, the unmanned vehicle's wheelbase, etc. Vehicle dynamic attribute information such as orientation, unmanned vehicle turning angle, and unmanned vehicle wheelbase can determine at least one reference path starting from the reference position point corresponding to each sampling line. For example, the vehicle motion trajectory information can be determined based on the vehicle dynamic attribute information such as the preset speed of the autonomous vehicle, the orientation of the autonomous vehicle, the turning angle of the autonomous vehicle, and the wheelbase of the autonomous vehicle. The reference position point corresponding to each sampling line is at least one reference path starting from the starting point.

In some embodiments, before determining the reference path, the rotation angle range of the unmanned vehicle may be sampled to obtain multiple reference rotation angles. Reference paths include paths under multiple reference corners. Through corner sampling, the stability of the planned path can be improved without excessive consumption of computing power, thus balancing the relationship between computing power consumption and path stability.

In some embodiments, the rotation angle range is determined by preset resolution

Perform sampling to obtain the unmanned vehicle corner set

in,

is the minimum turning angle of the unmanned vehicle,

is the maximum turning angle of the unmanned vehicle.

Gather around the corner with self-driving cars

For example, information related to vehicle dynamics can be used to determine the reference path of an unmanned vehicle at different unmanned vehicle turning angles.

In some embodiments, the vehicle dynamics related information includes the preset speed of the unmanned vehicle, the direction of the unmanned vehicle, the unmanned vehicle turning angle in the sampled unmanned vehicle turning angle set, the unmanned vehicle wheelbase and other vehicle dynamics attributes. information, and determine the vehicle movement trajectory information of the unmanned vehicle based on the relationship between the vehicle dynamics attribute information and the vehicle movement trajectory information.

For example, the relationship between vehicle dynamic attribute information and vehicle motion trajectory information can be represented by a bicycle model. The functional relationship represented by the bicycle model is

In the above functional relationship, θ is the direction of the unmanned vehicle at the current position, v is the preset speed of the unmanned vehicle,

is the sampled unmanned vehicle rotation angle set, and L is the unmanned vehicle wheelbase. Taking the coordinates of the current position point of the unmanned vehicle as (x, y, θ) as an example,

It represents the vehicle motion trajectory information of the unmanned vehicle, that is, the coordinate change of the unmanned vehicle starting from the current position point.

According to the current position point of the unmanned vehicle and the vehicle movement trajectory information of the unmanned vehicle, multiple coordinates that reflect the vehicle movement trajectory of the unmanned vehicle at different unmanned vehicle corners can be obtained in a discretized manner. Based on these discretized coordinates, it is possible to determine where the unmanned vehicle will gather at the corner of the unmanned vehicle

The reference path under the unmanned vehicle corner in . Self-driving car at the unmanned car corner

arrive

Refer to Figure 2 for the multiple reference paths below. The reference trajectories under different unmanned vehicle turning angles have different turning radii R, where,

In some embodiments, the vehicle dynamics related information includes vehicle trajectory description information, and the vehicle trajectory description information can be used to determine at least one reference path starting from the reference position point corresponding to each sampling line.

Still taking Figure 2 as an example, for a sampling line shown in Figure 2, candidate location points include P ₁ to P ₇ . A plurality of candidate position points on a sampling line similar to that shown in Figure 2 constitute a candidate position point set S={P ₁ , P ₂ ,…,P ₇ ,…}.

In some embodiments, for each sampling line, the intersection point of the previous frame path of the unmanned vehicle and each sampling line may also be determined as a candidate position point located on each sampling line. In this way, the relationship between the selection of the target position point on the path of the current frame and the path of the previous frame of the unmanned vehicle is considered, thereby further considering the similarity of the paths, reducing path jitter, and further improving the stability of the planned path. sex.

Taking Figure 2 as an example, the intersection point of the unmanned vehicle's path in the previous frame and the sampling line is P ₈ . P ₈ is also added to the candidate position point set as a candidate position point.

In step S140, multiple target location points arriving sequentially from the current location point are screened from multiple candidate location points located on multiple sampling lines.

In some embodiments, a pathfinding algorithm can be used to determine multiple target location points based on the current location point, multiple candidate location points located on multiple sampling lines, and preset driving costs. The stopping condition of the pathfinding algorithm includes that the mileage from the current position point to the last target position point is greater than or equal to the mileage threshold. For example, pathfinding algorithms include algorithms such as the A-star algorithm for finding the best path.

In some embodiments, multiple target location points can be determined by using the candidate location point set S and the current location point as inputs to the pathfinding algorithm, and presetting the driving cost as the target constraint.

In some embodiments, the preset driving cost includes at least one of a smoothness cost and a safety cost. The smoothness cost is positively related to the difference in the direction of the unmanned vehicle between the position point currently processed by the path-finding algorithm and the candidate position point. The safety cost is inversely related to the closest distance to the obstacle from the line segment between the position point currently processed by the pathfinding algorithm and the candidate position point. The location point currently processed by the pathfinding algorithm is the current location point or candidate location point.

In some embodiments, when the candidate location point includes the intersection of the previous frame path of the unmanned vehicle and each sampling line, the preset driving cost also includes a path similarity cost. The path similarity cost represents the similarity between the path starting from the position point currently processed by the pathfinding algorithm and the path in the previous frame. Path stability can be further improved by reflecting the similarity between the planned path of the current frame and the path of the previous frame through the path similarity cost.

In some embodiments, taking the preset driving cost including smoothness cost, safety cost and path similarity cost as an example, the preset driving cost is obtained by weighting the smoothness cost, safety cost and path similarity cost. For example, the weight value used for the weighting operation includes a first weight value, a second weight value, and a third weight value. The default driving cost = first weight value × smoothness cost + second weight value × safety cost + third weight value × path similarity cost. The first weight value, the second weight value and the third weight value are set according to the actual situation.

In some embodiments, the path similarity cost is related to the shortest distance of the path segment between the position point currently processed by the path-finding algorithm and each adjacent position point on the path of the previous frame or the previous position point currently processed by the path-finding algorithm. The shortest distance between a position point and each adjacent position point on the path of the previous frame is positively correlated.

In some embodiments, multiple candidate position points that satisfy the second preset motion condition may be first selected from multiple candidate position points located on multiple sampling lines, where the second preset motion condition includes one of the position points. There is at least one of the following: no collision between the motions and the candidate location point being located within the preset map range. Then, select a plurality of target position points from a plurality of candidate position points that satisfy the second preset motion condition. For the process of selecting multiple target location points from multiple candidate location points that satisfy the second preset motion condition, you may refer to the above-mentioned pathfinding algorithm.

For example, the obstacle information around the unmanned vehicle or self-driving vehicle can be obtained from the perception module of the unmanned vehicle to determine whether the movement between location points is collision-free. Taking Figure 2 as a reference, the path from the current position point c to P ₁ passes through obstacles, so the candidate position point P ₁ will be deleted from the candidate position point set S.

For example, the map data can be obtained from the map module of the unmanned vehicle, and based on the preset map range, it can be determined whether the position of the candidate location point in the map data is within the preset map range.

In step S150, the current frame path of the unmanned vehicle is determined based on the current position point and multiple target position points. In some embodiments, the initial path composed of the current position point and multiple target position points is smoothed to obtain the current frame path of the unmanned vehicle. The current frame path is also called a trajectory. The trajectory consists of a series of trajectory points arranged from first to last according to relative time. The trajectory point information includes but is not limited to coordinates, speed, acceleration, orientation, relative time, etc.

FIG. 3A is a schematic diagram showing the results of path planning according to the path planning method based on uniform sampling.

FIG. 3B is a schematic diagram showing the results of path planning according to the path planning method of some embodiments of the present disclosure.

Figure 3A shows the changes in the position coordinates of each path point obtained by path planning based on the path planning method based on uniform sampling. Figure 3B shows the changes in the position coordinates of each path point obtained by path planning according to the path planning method of some embodiments of the present disclosure. The x-axis and y-axis together constitute the position coordinates of the path point.

Comparing Figure 3A and Figure 3B, it can be found that in the path planning results based on the uniform sampling method in Figure 3A, the path frame jitter is obvious, especially within the area marked by the rectangular box in Figure 3A. The position coordinates of the planned path points change in a large range and are scattered. In the result of path planning according to the path planning method of some embodiments of the present disclosure in Figure 3B, the path frame is relatively more stable. With reference to the area marked by the rectangular box similar to Figure 3A, the path points mentioned in the path planning process are The range of position coordinate changes is small and concentrated.

FIG. 4 is a block diagram illustrating a path planning device according to some embodiments of the present disclosure.

As shown in FIG. 4 , the path planning device 41 for automatic driving includes a first determination module 411 , a second determination module 412 , a sampling module 413 , a filtering module 414 and a third determination module 415 .

The first determination module 411 is configured to determine the guidance line for assisting the driving of the unmanned vehicle according to the navigation path of the unmanned vehicle, for example, performing step S110 as shown in FIG. 1 .

The second determination module 412 is configured to determine multiple sampling lines for assisting multi-layer sampling according to the guide line, and each sampling line corresponds to one layer of sampling, for example, performing step S120 as shown in FIG. 1 .

The sampling module 413 is configured to sample at least one candidate position point from each sampling line based on the current position point of the unmanned vehicle using vehicle dynamics related information, for example, performing step S130 as shown in Figure 1 .

The screening module 414 is configured to screen multiple target location points arriving sequentially from the current location point from multiple candidate location points located on multiple sampling lines, for example, performing step S140 as shown in FIG. 1 .

The third determination module 415 is configured to determine the current frame path of the unmanned vehicle based on the current location point and multiple target location points, for example, performing step S150 as shown in Figure 1 .

In some embodiments, the third determination module 315 includes an initial path determination module and a path smoothing module. The initial path determination module is configured to determine an initial path composed of a current location point and a plurality of target location points. The path smoothing module is configured to smooth the initial path composed of the current position point and multiple target position points to obtain the current frame path of the unmanned vehicle.

FIG. 5 is a block diagram illustrating a path planning device according to other embodiments of the present disclosure.

As shown in FIG. 5 , the path planning device 51 for autonomous driving includes a memory 511; and a processor 512 coupled to the memory 511. The memory 511 is used to store instructions for executing corresponding embodiments of the path planning method. The processor 512 is configured to execute the path planning method in any embodiment of the present disclosure based on instructions stored in the memory 511 .

Figure 6 is a block diagram illustrating an autonomous vehicle according to some embodiments of the present disclosure.

As shown in FIG. 6 , the unmanned vehicle 6 includes a path planning device 61 . The path planning device 61 is configured to execute the path planning method in any embodiment of the present disclosure. For example, the path planning device 61 and the path planning device 41 or 51 have the same or similar structure or function.

In some embodiments, the unmanned vehicle 6 further includes a positioning module 62 . The positioning module 62 is configured to send the speed of the unmanned vehicle, the position coordinates of the current location point, and the direction of the unmanned vehicle to the path planning device 61 . For example, the speed of an autonomous vehicle is a preset speed.

In some embodiments, the unmanned vehicle 6 further includes a navigation module 63 . The navigation module 63 is configured to send the navigation path of the unmanned vehicle to the path planning device 61 .

In some embodiments, the unmanned vehicle 6 further includes a map module 64 . The map module 64 is configured to provide map data to the route planning device 61 .

In some embodiments, the unmanned vehicle 6 further includes a sensing module 65 . The sensing module 65 is configured to sense obstacles around the unmanned vehicle and send the perceived obstacle information to the path planning device 61 . The sensing module 65 is also configured to sense traffic light information and the like.

In some embodiments, the unmanned vehicle 6 further includes a control module 66 . The control module 66 is configured to receive the current frame path from the path planning device 61 and control the unmanned vehicle to travel according to the current frame path.

FIG. 7 is a block diagram illustrating an autonomous vehicle according to other embodiments of the present disclosure.

As shown in Figure 7, the unmanned vehicle 7 mainly includes four parts: an automatic driving module 71, a chassis module 72, a remote monitoring flow module 73, and a cargo box module 74.

The automatic driving module 71 includes an automatic driving sensor and core processing unit (Orin or Xavier module) component 711, a traffic light recognition camera 712, front, rear, left and right surround cameras 7131, 7132, 7133, 7134, a multi-line laser radar 714, and a positioning module 715 (such as Beidou, GPS, etc.), inertial navigation unit 716. The camera and the autonomous driving module can communicate. In order to increase the transmission speed and reduce the wiring harness, GMSL link communication can be used. The automatic driving sensor and core processing unit (Orin or Xavier module) component 711 includes the path planning device in any embodiment of the present disclosure and is configured to execute the path planning method in any embodiment of the present disclosure.

In some embodiments, the autonomous driving module 71 also includes a switch 717 and front, rear, left and right blinding radars 7181, 7182, 7183, and 7184.

The chassis module 72 mainly includes a battery 721, a power management device 722, a chassis controller 723, a motor driver 724, and a power motor 725. The battery 721 provides power for the entire unmanned vehicle system, and the power management device 722 converts the battery output into voltages of different levels that can be used by each functional module, and controls power on and off. The chassis controller 723 receives movement instructions issued by the automatic driving module and controls the steering, forward, backward, braking, etc. of the unmanned vehicle. For example, the motion instruction is determined by the control module 66 shown in FIG. 6 according to the current frame path. Chassis module 72 also includes main battery 726 .

The remote monitoring streaming module 73 is composed of a front surveillance camera 731, a rear surveillance camera 732, a left surveillance camera 733, a right surveillance camera 734 and a streaming module 734. This module transmits the video data collected by the surveillance cameras to the background server for background operations. Personnel viewing.

The cargo box module 74 is a cargo carrying device for the unmanned vehicle and includes a delivery box 741 . The cargo box module 74 is also provided with a display interaction module 742. The display interaction module 742 is used for the unmanned vehicle to interact with the user. The user can perform operations such as picking up, depositing, and purchasing goods through the display interaction module. The type of cargo box can be changed according to actual needs. For example, in a logistics scenario, a cargo box can include multiple sub-boxes of different sizes, and the sub-boxes can be used to load goods for distribution. In a retail scenario, the cargo box can be set up as a transparent box so that users can intuitively see the products for sale.

Carton module 74 also includes antenna 743 . The chassis module 72 also includes a wireless communication module 727. The wireless communication module 727 communicates with the backend server through the antenna 743, allowing the backend operator to remotely control the unmanned vehicle.

Figure 8 is a side view illustrating an autonomous vehicle according to some embodiments of the present disclosure.

As shown in Figure 8, the unmanned vehicle includes a display interaction module 81, a chassis 82, a left blind filling radar 83, a right blind filling radar 84, a rear blind filling radar 85, a laser radar 86, a right camera 87, and a cargo box. 88. The functions of the interactive module 81, chassis 82, left blind filling radar 83, right blind filling radar 84, rear blind filling radar 85, lidar 86, right camera 87, and cargo box 88 can be referred to the description in Figure 7. No further details will be given here.

Figure 9 is a block diagram illustrating a computer system for implementing some embodiments of the present disclosure.

As shown in Figure 9, computer system 90 may be embodied in the form of a general purpose computing device. Computer system 90 includes memory 910, a processor 920, and a bus 900 that connects various system components.

Memory 910 may include, for example, system memory, non-volatile storage media, and the like. System memory stores, for example, operating systems, applications, boot loaders, and other programs. System memory may include volatile storage media such as random access memory (RAM) and/or cache memory. The non-volatile storage medium stores, for example, instructions for executing corresponding embodiments of the path planning method. Non-volatile storage media includes but is not limited to disk storage, optical storage, flash memory, etc.

The processor 920 may be implemented as a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, discrete hardware components such as discrete gates or transistors. accomplish. Correspondingly, each module, such as the judgment module and the determination module, can be implemented by instructions in the central processing unit (CPU) running memory to perform the corresponding steps, or by dedicated circuits that perform the corresponding steps.

Bus 900 may use any of a variety of bus structures. For example, bus structures include, but are not limited to, Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MCA) bus, and Peripheral Component Interconnect (PCI) bus.

The computer system 90 may also include an input/output interface 930, a network interface 940, a storage interface 950, etc. These interfaces 930, 940, 950, the memory 910 and the processor 920 may be connected through a bus 900. The input and output interface 930 can provide a connection interface for input and output devices such as a monitor, mouse, and keyboard. Network interface 940 provides a connection interface for various networked devices. The storage interface 950 provides a connection interface for external storage devices such as floppy disks, USB disks, and SD cards.

Various aspects of the disclosure are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus, and computer program products according to embodiments of the disclosure. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks, can be implemented by computer-readable program instructions.

These computer readable program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable device to produce a machine, such that execution of the instructions by the processor produces implementations in one or more blocks of the flowcharts and/or block diagrams. A device with specified functions.

Computer-readable program instructions, which may also be stored in computer-readable memory, cause the computer to operate in a specific manner to produce an article of manufacture, including implementing the functions specified in one or more blocks of the flowcharts and/or block diagrams. instructions.

The disclosure may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment that combines software and hardware aspects.

Through the path planning method and device, unmanned vehicle, and computer storage medium in the above embodiments, the stability of the planned path can be improved.

So far, the path planning method and device, the unmanned vehicle, and the computer-storable medium according to the present disclosure have been described in detail. To avoid obscuring the concepts of the present disclosure, some details that are well known in the art have not been described. Based on the above description, those skilled in the art can completely understand how to implement the technical solution disclosed here.

Claims (19)

1. A path planning method for autonomous driving, including:

   According to the navigation path of the unmanned vehicle, determine the guidance line used to assist the driving of the unmanned vehicle;

   According to the guide line, multiple sampling lines used to assist in multi-layer position point sampling are determined, and each sampling line corresponds to one layer of position point sampling;

   Based on the current position point of the unmanned vehicle, use vehicle dynamics related information to sample at least one candidate position point from each sampling line;

   From multiple candidate location points located on multiple sampling lines, screen multiple target location points that arrive sequentially starting from the current location point;

   According to the current position point and the plurality of target position points, the current frame path of the unmanned vehicle is determined.
2. The path planning method according to claim 1, wherein based on the current position point, using vehicle dynamics related information to sample at least one candidate position point from each sampling line includes:

   For each sampling line, use vehicle dynamics related information to determine at least one reference path starting from the reference position point corresponding to each sampling line, wherein each sampling line is the same as the current position. In the case of the sampling line closest to the point, the reference position point is the current position point; in the case of each sampling line being other sampling lines, the reference position point is located close to each of the sampling lines. Each candidate position point on the sampling line adjacent to the sampling line and close to the direction of the current position point;

   At least one candidate position point located on each sampling line is determined based on the intersection point of each sampling line and the at least one reference path.
3. The path planning method according to claim 2, wherein,

   For the sampling line closest to the current position point, use vehicle dynamics related information to determine at least one reference path starting from the current position point;

   Determine at least one candidate position located on the sampling line closest to the current position point based on the intersection point of the sampling line closest to the current position point and at least one reference path starting from the current position point point;

   For each other sampling line, use the vehicle dynamics related information to determine each candidate position point located on the sampling line adjacent to each other sampling line and closest to the current position point. At least one reference path to the starting point;

   According to the intersection point of each other sampling line and at least one reference path starting from each candidate position point on the sampling line adjacent to each other sampling line and closest to the current position point , determine at least one candidate position point located on each of the other sampling lines.
4. The path planning method according to any one of claims 1 to 3, wherein the vehicle dynamics related information includes vehicle dynamics attribute information or vehicle motion trajectory information, and the vehicle dynamics related information is used to determine the relationship with each of the vehicle dynamics. At least one reference path starting from a reference position point corresponding to each sampling line includes:

   Using the vehicle dynamics attribute information or vehicle motion trajectory information, at least one reference path starting from the reference position point corresponding to each sampling line is determined.
5. The path planning method according to claim 2, further comprising:

   The rotation angle range of the unmanned vehicle is sampled to obtain multiple reference rotation angles, wherein the reference path includes a path under the multiple reference rotation angles.
6. The path planning method according to claim 2, wherein based on the current position point of the unmanned vehicle, using vehicle dynamics related information to sample at least one candidate position point from each sampling line further includes:

   For each sampling line, the intersection point of the previous frame path of the unmanned vehicle and each sampling line is determined to be a candidate position point located on each sampling line.
7. The path planning method according to claim 2, wherein the candidate position point on each sampling line is an intersection point that satisfies a first preset motion condition, and the first preset motion condition includes one corresponding to each sampling line. The unmanned vehicle orientation difference between the reference position point and the candidate position point on each sampling line is less than or equal to the unmanned vehicle orientation difference threshold.
8. The path planning method according to claim 1, wherein screening multiple target location points arriving sequentially from the current location point from multiple candidate location points located on multiple sampling lines includes:

   According to the current location point, multiple candidate location points located on the multiple sampling lines and preset driving costs, a path-finding algorithm is used to determine the multiple target location points, wherein the stop of the path-finding algorithm The condition includes that the mileage from the current location point to the last target location point is greater than or equal to the mileage threshold.
9. The path planning method according to claim 8, wherein the preset driving cost includes at least one of a smoothness cost and a safety cost, wherein the smoothness cost is related to the position currently processed by the path-finding algorithm. The difference in direction of the unmanned vehicle between the point and the candidate location point is positively correlated, and the safety cost is related to the shortest distance from the line segment between the location point currently processed by the path-finding algorithm and the candidate location point to the obstacle. There is a negative correlation, and the position point currently processed by the path-finding algorithm is the current position point or the candidate position point.
10. The path planning method according to claim 9, wherein when the candidate location point includes the intersection of the previous frame path of the unmanned vehicle and each sampling line, the preset driving cost also includes the path Similarity cost, the path similarity cost represents the similarity between the path starting from the position point currently processed by the path-finding algorithm and the path in the previous frame.
11. The path planning method according to claim 10, wherein the path similarity cost is related to the path line segment between the position point currently processed by the path finding algorithm and each adjacent position point on the path of the previous frame. The shortest distance or the shortest distance of the path line segment between the previous position point of the position point currently processed by the path-finding algorithm and each adjacent position point on the path of the previous frame is positively correlated.
12. The path planning method according to claim 1, wherein screening multiple target location points arriving sequentially from the current location point from multiple candidate location points located on multiple sampling lines includes:

    Select a plurality of candidate position points that satisfy a second preset motion condition from a plurality of candidate position points located on the plurality of sampling lines, the second preset motion condition includes no collision in motion between position points and candidate position points Located within at least one of the preset map ranges;

    Select the plurality of target position points from a plurality of candidate position points that satisfy the second preset motion condition.
13. A path planning device for autonomous driving, including:

    The first determination module is configured to determine the guidance line for assisting the driving of the unmanned vehicle according to the navigation path of the unmanned vehicle;

    The second determination module is configured to determine multiple sampling lines for assisting multi-layer sampling according to the guide line, with each sampling line corresponding to one layer of sampling;

    A sampling module configured to sample at least one candidate position point from each sampling line based on the current position point of the unmanned vehicle using vehicle dynamics related information;

    A screening module configured to screen multiple target location points arriving sequentially starting from the current location point from multiple candidate location points located on multiple sampling lines;

    A third determination module is configured to determine the current frame path of the unmanned vehicle according to the current position point and the plurality of target position points.
14. A path planning device for autonomous driving, including:

    memory; and

    A processor coupled to the memory, the processor being configured to execute the path planning method of any one of claims 1 to 12 based on instructions stored in the memory.
15. An unmanned vehicle, including:

    The path planning device for automatic driving according to claim 13 or 14.
16. The unmanned vehicle according to claim 15, further comprising:

    At least one of a positioning module, a navigation module, a perception module and a map module, wherein,

    The positioning module is configured to send the speed of the unmanned vehicle, the position coordinates of the current location point and the direction of the unmanned vehicle to the path planning device;

    The navigation module is configured to send the navigation path of the unmanned vehicle to the path planning device;

    The sensing module is configured to sense obstacles around the unmanned vehicle and send the perceived obstacle information to the path planning device;

    The map module is configured to provide map data to the route planning device.
17. The unmanned vehicle according to claim 15, further comprising:

    A control module configured to receive the current frame path from the path planning device and control the unmanned vehicle to travel according to the current frame path.
18. A computer storage medium on which computer program instructions are stored. When the instructions are executed by a processor, the path planning method according to any one of claims 1 to 12 is implemented.
19. A computer program consisting of:

    Instructions, which when executed by a processor cause the processor to execute the path planning method according to any one of claims 1 to 12.

Images