Trajectory Planning Method, Device, Equipment and Storage Medium for Autonomous Driving Vehicle

Abstract

The present application discloses a trajectory planning method, device, equipment and storage medium for an autonomous driving vehicle, relating to the field of computer technology. The method includes: obtaining the driving state information of the autonomous driving vehicle, the planned path information, and the driving state information of dynamic obstacles; wherein, the autonomous driving vehicle has a steering intention; based on the planned path information of the autonomous driving vehicle and the driving state information of the dynamic obstacles, using a risk dynamic obstacle recognition strategy, determining risk dynamic obstacles from the dynamic obstacles; based on the planned path information of the autonomous driving vehicle and the driving state information of the risk dynamic obstacles, using a steering collision boundary recognition strategy, obtaining a steering collision boundary; based on the steering collision boundary, the driving state information of the autonomous driving vehicle, the planned path information, and the driving state information of the risk dynamic obstacles, performing trajectory planning processing on the autonomous driving vehicle to obtain the planned trajectory of the autonomous driving vehicle.

Description

Technical Field

[0001] This application relates to the field of computer technology, specifically to the fields of intelligent transportation and autonomous driving, etc., and particularly relates to a trajectory planning method, device, equipment and storage medium for an autonomous driving vehicle. Background Art

[0002] In the distribution scenario of autonomous driving vehicles, when the vehicle turns at an intersection, a rubbing accident with other turning vehicles may occur due to the inner wheel difference phenomenon. Therefore, when planning the driving trajectory of an autonomous driving vehicle at an intersection, it is necessary to consider the influence caused by the inner wheel difference of other turning dynamic obstacles.

[0003] Currently, most of the methods in the related art analyze how to avoid rubbing from the perspective of ordinary vehicles with drivers, and no effective solutions are given for such risk scenarios encountered by autonomous driving vehicles. Summary of the Invention

[0004] This application provides a trajectory planning method, device, equipment and storage medium for an autonomous driving vehicle, which realizes the risk assessment of the steering of an autonomous driving vehicle and improves the safety and reliability of the steering of an autonomous driving vehicle. The technical solutions are as follows:

[0005] In a first aspect, a trajectory planning method for an autonomous driving vehicle is provided. The method includes:

[0006] Obtain the driving state information of the autonomous driving vehicle, the planned path information, and the driving state information of the dynamic obstacle; wherein, the autonomous driving vehicle has a steering intention;

[0007] Based on the planned path information of the autonomous driving vehicle and the driving state information of the dynamic obstacle, use the risk dynamic obstacle recognition strategy to determine the risk dynamic obstacle from the dynamic obstacles;

[0008] Based on the planned path information of the autonomous driving vehicle and the driving state information of the risk dynamic obstacle, use the steering collision boundary recognition strategy to obtain the steering collision boundary;

[0009] Based on the steering collision boundary, the driving state information of the autonomous driving vehicle, the planned path information, and the driving state information of the risk dynamic obstacle, perform trajectory planning processing on the autonomous driving vehicle to obtain the planned trajectory of the autonomous driving vehicle.

[0010] In a possible implementation manner, the risk dynamic obstacle recognition strategy includes a steering risk area recognition algorithm and a risk dynamic obstacle recognition algorithm. Determining a risk dynamic obstacle from the dynamic obstacles by using the risk dynamic obstacle recognition strategy based on the planned path information of the autonomous driving vehicle and the driving state information of the dynamic obstacles includes:

[0011] Based on the planned path information of the autonomous driving vehicle, using the steering risk area recognition algorithm to obtain a steering risk area;

[0012] Based on the driving state information of the dynamic obstacle and the steering risk area, using the risk dynamic obstacle recognition algorithm to determine a risk dynamic obstacle from the dynamic obstacles.

[0013] In a possible implementation manner, obtaining a steering collision boundary by using a steering collision boundary recognition strategy based on the planned path information of the autonomous driving vehicle and the driving state information of the risk dynamic obstacle includes:

[0014] In response to the driving state information of the risk dynamic obstacle including the front wheel speed of the risk dynamic obstacle, based on the driving state information of the risk dynamic obstacle, using a preset Ackermann steering model to determine the rear wheel driving trajectory of the risk dynamic obstacle;

[0015] Based on the rear wheel driving trajectory of the risk dynamic obstacle, obtaining the steering collision boundary.

[0016] In a possible implementation manner, obtaining a steering collision boundary by using a steering collision boundary recognition strategy based on the planned path information of the autonomous driving vehicle and the driving state information of the risk dynamic obstacle includes:

[0017] In response to the driving state information of the risk dynamic obstacle not including the front wheel speed of the risk dynamic obstacle, based on the driving state information of the risk dynamic obstacle, using a preset decomposition algorithm to perform a decomposition process on the risk dynamic obstacle to obtain a decomposition unit corresponding to the front part of the risk dynamic obstacle;

[0018] Based on the planned path information of the autonomous driving vehicle and the decomposition unit corresponding to the front part, performing a speculation process on the front wheel driving trajectory of the risk dynamic obstacle to obtain the front wheel driving trajectory of the risk dynamic obstacle;

[0019] Based on the planned path information of the autonomous driving vehicle and the front wheel driving trajectory of the risk dynamic obstacle, using a preset offset algorithm to calculate a lateral offset;

[0020] Obtain the rear-wheel travel trajectory of the risk dynamic obstacle based on the lateral offset and the front-wheel travel trajectory;

[0021] Obtain the steering collision boundary based on the rear-wheel travel trajectory of the risk dynamic obstacle.

[0022] In a possible implementation, the travel state information of the risk dynamic obstacle includes the shape information of the risk dynamic obstacle. Based on the travel state information of the risk dynamic obstacle, use a preset decomposition algorithm to perform a decomposition process on the risk dynamic obstacle to obtain the decomposition unit corresponding to the front part of the risk dynamic obstacle, including:

[0023] Based on the shape information of the risk dynamic obstacle, use a preset decomposition algorithm to calculate the radius of the decomposition unit corresponding to the risk dynamic obstacle and the first distance between two adjacent decomposition units;

[0024] Based on the radius of the decomposition unit corresponding to the risk dynamic obstacle and the first distance between two adjacent decomposition units, perform a decomposition process on the risk dynamic obstacle to obtain multiple decomposition units corresponding to the risk dynamic obstacle;

[0025] Obtain the decomposition unit corresponding to the front part of the risk dynamic obstacle from the multiple decomposition units corresponding to the risk dynamic obstacle.

[0026] In a possible implementation, the speculation process for the front-wheel travel trajectory of the risk dynamic obstacle based on the planned path information of the autonomous vehicle and the decomposition unit corresponding to the front part to obtain the front-wheel travel trajectory of the risk dynamic obstacle includes:

[0027] Obtain each position point in the planned path information of the autonomous vehicle and the orientation angle corresponding to each position point;

[0028] Based on each position point and the decomposition unit corresponding to the front part, use a binary search algorithm to obtain the matching position point that matches the decomposition unit corresponding to the front part;

[0029] Obtain the second distance between the decomposition unit corresponding to the front part and the matching position point;

[0030] Based on the orientation angle corresponding to each position point, use a preset vector algorithm to determine the normal vector corresponding to each position point;

[0031] Based on each position point in the planned path information of the autonomous vehicle, the normal vector corresponding to each position point, and the second distance, perform a speculation process on the front-wheel travel trajectory of the risk dynamic obstacle to obtain the front-wheel travel trajectory of the risk dynamic obstacle.

[0032] In a possible implementation, based on the planned path information of the autonomous driving vehicle and the front-wheel driving trajectory of the risk dynamic obstacle, using a preset offset algorithm, a lateral offset is calculated, including:

[0033] Obtain the curvature value of each position point in the planned path information of the autonomous driving vehicle;

[0034] Based on the curvature value of each position point in the planned path information of the autonomous driving vehicle and the second spacing, determine the curvature value of each trajectory point in the front-wheel driving trajectory of the risk dynamic obstacle;

[0035] Obtain the first spacing between two adjacent decomposition units;

[0036] Based on the curvature value of each trajectory point and the first spacing, calculate the lateral offset.

[0037] In a possible implementation, the driving state information of the autonomous driving vehicle includes the shape information, speed information, acceleration information, and jerk information of the autonomous driving vehicle. Based on the steering collision boundary, the driving state information of the autonomous driving vehicle, the planned path information, and the driving state information of the risk dynamic obstacle, performing trajectory planning processing on the autonomous driving vehicle to obtain the planned trajectory of the autonomous driving vehicle, including:

[0038] Based on the shape information of the autonomous driving vehicle, the planned path information, and the driving state information of the risk dynamic obstacle, obtain the position relationship information between the autonomous driving vehicle and the risk dynamic obstacle;

[0039] Based on the speed information of the autonomous driving vehicle, the planned path information, and the steering collision boundary, obtain the first time when the autonomous driving vehicle reaches the steering collision boundary;

[0040] Based on the driving state information of the risk dynamic obstacle and the steering collision boundary, obtain the second time when the risk dynamic obstacle reaches the steering collision boundary;

[0041] In response to the position relationship information satisfying a preset first constraint condition, or the speed information, the first time, and the second time of the autonomous driving vehicle satisfying a preset second constraint condition, based on the speed information, acceleration information, and jerk information of the autonomous driving vehicle, using a preset objective function, perform trajectory planning processing on the autonomous driving vehicle to obtain the planned trajectory of the autonomous driving vehicle.

[0042] In a second aspect, a trajectory planning device for an autonomous driving vehicle is provided. The device includes:

[0043] An acquisition unit, configured to acquire the driving state information of an autonomous driving vehicle, the planned path information, and the driving state information of dynamic obstacles; wherein, the autonomous driving vehicle has a steering intention;

[0044] A determination unit, configured to determine a risk dynamic obstacle from the dynamic obstacles based on the planned path information of the autonomous driving vehicle and the driving state information of the dynamic obstacles, by using a risk dynamic obstacle recognition strategy;

[0045] An obtaining unit, configured to obtain a steering collision boundary based on the planned path information of the autonomous driving vehicle and the driving state information of the risk dynamic obstacle, by using a steering collision boundary recognition strategy;

[0046] A planning unit, configured to perform trajectory planning processing on the autonomous driving vehicle based on the steering collision boundary, the driving state information of the autonomous driving vehicle, the planned path information, and the driving state information of the risk dynamic obstacle, so as to obtain the planned trajectory of the autonomous driving vehicle.

[0047] In a third aspect, a computer-readable storage medium is provided, in which at least one instruction is stored, and the at least one instruction is loaded and executed by a processor to implement the method in the above-mentioned aspect and any possible implementation manner.

[0048] In a fourth aspect, an electronic device is provided, including:

[0049] At least one processor; and

[0050] A memory communicatively connected to the at least one processor; wherein,

[0051] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor, so that the at least one processor can execute the method in the above-mentioned aspect and any possible implementation manner.

[0052] In a fifth aspect, a computer program product is provided, including a computer program, and the computer program implements the method in the above-mentioned aspect and any possible implementation manner when being executed by a processor.

[0053] In a sixth aspect, an autonomous driving vehicle is provided, including the above-mentioned electronic device.

[0054] The beneficial effects of the technical solution provided by this application at least include:

[0055] As can be seen from the above technical solutions, embodiments of the present application can obtain the driving state information of an autonomous vehicle, the planned path information, and the driving state information of dynamic obstacles. The autonomous vehicle has a steering intention. Furthermore, based on the planned path information of the autonomous vehicle and the driving state information of the dynamic obstacles, using a risk dynamic obstacle recognition strategy, risk dynamic obstacles can be determined from the dynamic obstacles. Based on the planned path information of the autonomous vehicle and the driving state information of the risk dynamic obstacles, using a steering collision boundary recognition strategy, a steering collision boundary can be obtained, enabling trajectory planning processing of the autonomous vehicle based on the steering collision boundary, the driving state information of the autonomous vehicle, the planned path information, and the driving state information of the risk dynamic obstacles to obtain the planned trajectory of the autonomous vehicle. Since the steering collision boundary between the autonomous vehicle and the dynamic obstacles can be effectively estimated according to the driving state information of the autonomous vehicle, the planned path information, and the driving state information of the dynamic obstacles, and this steering collision boundary is the boundary that poses a danger to the steering of the host vehicle when the dynamic obstacle steers, risk estimation of the steering scenario of the autonomous vehicle is realized. Moreover, based on this steering collision boundary, planning the driving trajectory during the steering of the autonomous vehicle can effectively reduce the scraping risk, thereby improving the safety and reliability of the steering of the autonomous vehicle.

[0056] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] To more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0058] Figure 1 is a schematic flowchart of a trajectory planning method for an autonomous vehicle provided by an embodiment of the present application;

[0059] Figure 2 is a schematic flowchart of a trajectory planning method for an autonomous vehicle provided by another embodiment of the present application;

[0060] Figure 3 is a schematic diagram of a steering risk area of a trajectory planning method for an autonomous vehicle provided by another embodiment of the present application;

[0061] Figure 4Schematic diagram of risk dynamic obstacle steering for the trajectory planning method of an autonomous vehicle provided by another embodiment of this application;

[0062] Figure 5 Schematic diagram of another risk dynamic obstacle steering for the trajectory planning method of an autonomous vehicle provided by another embodiment of this application;

[0063] Figure 6 Schematic diagram of multiple decomposition units of risk dynamic obstacles for the trajectory planning method of an autonomous vehicle provided by another embodiment of this application;

[0064] Figure 7 Schematic diagram of the front-wheel driving trajectory of a risk dynamic obstacle for the trajectory planning method of an autonomous vehicle provided by another embodiment of this application;

[0065] Figure 8 Schematic diagram of the rear-wheel driving trajectory of a risk dynamic obstacle for the trajectory planning method of an autonomous vehicle provided by another embodiment of this application;

[0066] Figures 9a to 9c Schematic diagram of the positional relationship between an autonomous vehicle and a risk dynamic obstacle for the trajectory planning method of an autonomous vehicle provided by another embodiment of this application;

[0067] Figure 10 Structural block diagram of a trajectory planning device for an autonomous vehicle provided by yet another embodiment of this application;

[0068] Figure 11 Block diagram of an electronic device for implementing the trajectory planning method of an autonomous vehicle in an embodiment of this application. Detailed implementation manners

[0069] The following describes exemplary embodiments of this application with reference to the accompanying drawings. Various details of the embodiments of this application are included to facilitate understanding, and they should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this application. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following.

[0070] Obviously, the described embodiments are some, but not all, of the embodiments of this application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this application without creative efforts fall within the scope of protection of this application.

[0071] It should be noted that the terminal devices involved in the embodiments of the present application may include, but are not limited to, intelligent devices such as mobile phones, personal digital assistants (PDAs), wireless handheld devices, and tablet computers; the display devices may include, but are not limited to, devices with display functions such as personal computers and televisions.

[0072] In addition, the term "and / or" in this article is merely a description of the association relationship between associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after.

[0073] The inner wheel difference generally refers to the phenomenon that when a vehicle is turning, the rear wheels are closer to the center of rotation than the front wheels. The greater the steering angle of the steering wheel and the longer the vehicle body, the more obvious the inner wheel difference. Specifically, the inner wheel difference is mainly caused by the different driving paths of the front and rear wheels when the vehicle is turning. Since the front wheels are usually responsible for steering, while the rear wheels follow the trajectory of the front wheels. The front wheels can change the driving direction according to the steering angle of the steering wheel, while the rear wheels, due to being unable to steer, can only follow the tilt of the vehicle body, which results in a wider driving path for the rear wheels than the front wheels during the turning process. The area formed by the inner wheel difference phenomenon is usually a blind spot in the driver's field of view and is difficult to directly observe through the front windshield or side mirrors, which is also the reason why autonomous vehicles are often scratched in unmanned delivery scenarios (such as at park intersections, etc.).

[0074] Currently, most of the methods in the related art analyze how to avoid scratches from the perspective of human drivers of trucks, but no effective solutions are given for autonomous vehicles encountering such risk scenarios.

[0075] Please refer to Figure 1 , which shows a schematic flowchart of a trajectory planning method for an autonomous vehicle provided by an embodiment of the present application. The trajectory planning method for the autonomous vehicle may specifically include:

[0076] Step 101, obtain the driving state information of the autonomous vehicle, the planned path information, and the driving state information of dynamic obstacles; wherein, the autonomous vehicle has a steering intention.

[0077] Step 102, based on the planned path information of the autonomous vehicle and the driving state information of the dynamic obstacles, use a risk dynamic obstacle recognition strategy to determine risk dynamic obstacles from the dynamic obstacles.

[0078] Step 103: Based on the planned path information of the autonomous vehicle and the driving state information of the risk dynamic obstacle, use the steering collision boundary recognition strategy to obtain the steering collision boundary.

[0079] Step 104: Based on the steering collision boundary, the driving state information of the autonomous vehicle, the planned path information, and the driving state information of the risk dynamic obstacle, perform trajectory planning processing on the autonomous vehicle to obtain the planned trajectory of the autonomous vehicle.

[0080] It should be noted that the dynamic obstacle can be a vehicle in the driving environment of the autonomous vehicle. The dynamic obstacle can include large vehicles such as trucks and trailers with steering intentions.

[0081] Exemplarily, based on the planned path information of the autonomous vehicle, the steering intention of the autonomous vehicle can be obtained. If the steering intention of the autonomous vehicle is to turn left, the dynamic obstacle can be a vehicle driving on the right side of the autonomous vehicle. If the steering intention of the autonomous vehicle is to turn right, the dynamic obstacle can be a vehicle driving on the left side of the autonomous vehicle.

[0082] It should be noted that the steering collision boundary can be located between the autonomous vehicle and the dynamic obstacle, and the steering collision boundary can be used to represent the boundary line where a collision occurs between the autonomous vehicle and the dynamic obstacle.

[0083] It should be noted that the driving state information of the autonomous vehicle can include but is not limited to the speed information, acceleration information, jerk information, and shape information of the autonomous vehicle. The planned path information of the autonomous vehicle can include but is not limited to multiple position points on the path, the forward distance corresponding to each position point, the orientation angle corresponding to each position point, the curvature value corresponding to each position point, etc. The shape information of the autonomous vehicle can include but is not limited to the length, width, and position of the top corners of the autonomous vehicle. The driving state information of the dynamic obstacle can include but is not limited to pose information, speed information, shape information, and front wheel speed. The shape information of the dynamic obstacle can also include but is not limited to the length, width, wheelbase, and rear wheel track of the dynamic obstacle. The front wheel speed of the dynamic obstacle can include the left front wheel speed and the right front wheel speed of the dynamic obstacle.

[0084] It should be noted that part or all of the execution entities of steps 101 to 104 can be an application located on a local terminal, or can also be a functional unit such as a plug-in or a software development kit (SDK) set in the application located on the local terminal, or can also be a processing engine in a network-side server, or can also be a distributed system on the network side. For example, the processing engine or the distributed system in the data processing platform on the network side, etc. This embodiment does not make special limitations on this.

[0085] It can be understood that the application can be a native app installed on the local terminal, or can also be a web app of a browser on the local terminal. This embodiment does not make limitations on this.

[0086] In this way, according to the driving state information of the autonomous vehicle, the planned path information, and the driving state information of the dynamic obstacle, the steering collision boundary between the autonomous vehicle and the dynamic obstacle can be effectively estimated. The steering collision boundary is the boundary that poses a danger to the steering of the host vehicle when the dynamic obstacle steers, realizing the risk estimation of the steering scenario of the autonomous vehicle. Moreover, based on this steering collision boundary, the driving trajectory during the steering process of the autonomous vehicle can be planned, effectively reducing the scraping risk, thereby improving the safety and reliability of the steering of the autonomous vehicle.

[0087] Optionally, in a possible implementation manner of this embodiment, the risk dynamic obstacle recognition strategy includes a steering risk area recognition algorithm and a risk dynamic obstacle recognition algorithm. In step 102, specifically, based on the planned path information of the autonomous vehicle, the steering risk area recognition algorithm can be used to obtain the steering risk area. Furthermore, based on the driving state information of the dynamic obstacle and the steering risk area, the risk dynamic obstacle recognition algorithm can be used to determine the risk dynamic obstacle from the dynamic obstacles.

[0088] In this implementation manner, the planned path information of the autonomous vehicle can include, but is not limited to, multiple position points on the path, the forward distance corresponding to each position point, the orientation angle corresponding to each position point, the curvature value corresponding to each position point, etc.

[0089] In a specific implementation process of this implementation manner, first, based on the planned path information of the autonomous vehicle, the steering risk side of the autonomous vehicle can be determined. Secondly, based on a preset adjustment coefficient, the forward distance corresponding to each position point, and the curvature value corresponding to each position point, the exploration distance of the steering risk side corresponding to each position point can be calculated. Thirdly, based on the exploration distance of the steering risk side corresponding to each position point, the steering risk area can be obtained.

[0090] In this implementation manner, the driving state information of the dynamic obstacle may include the pose information and shape information of the dynamic obstacle.

[0091] In another specific implementation process of this implementation manner, first, based on the steering risk area, the pose information and shape information of the dynamic obstacle, it can be determined whether there is an overlapping relationship between the dynamic obstacle and the steering risk area. If there is an overlapping relationship, it can be determined that the dynamic obstacle is a risk dynamic obstacle, and then based on the judgment result of whether there is an overlapping relationship, the risk dynamic obstacle can be determined from the dynamic obstacles.

[0092] In this way, the risk dynamic obstacles with steering collision risks can be accurately and effectively selected from the dynamic obstacles, so that the steering collision boundary can be calculated for the risk dynamic obstacles subsequently, improving the efficiency and reliability of the trajectory planning process.

[0093] Optionally, in a possible implementation manner of this embodiment, in step 103, specifically, in response to the driving state information of the risk dynamic obstacle including the front wheel speed of the risk dynamic obstacle, based on the driving state information of the risk dynamic obstacle, using a preset Ackermann steering model, the rear wheel driving trajectory of the risk dynamic obstacle can be determined, and then based on the rear wheel driving trajectory of the risk dynamic obstacle, the steering collision boundary can be obtained.

[0094] In this implementation manner, the risk dynamic obstacle can be an ordinary vehicle or an ordinary large vehicle, or a semi-trailer.

[0095] In a specific implementation process of this implementation manner, if the risk dynamic obstacle is an ordinary vehicle or an ordinary large vehicle, etc., here, the driving state information of the risk dynamic obstacle may include the left front wheel speed, right front wheel speed, wheelbase, and rear wheel track of the vehicle. Using a preset Ackermann steering model, based on the left front wheel speed, right front wheel speed, wheelbase, and rear wheel track of the risk dynamic obstacle, the steering radius of the rear wheel of the risk dynamic obstacle can be deduced, and then based on the steering radius of the rear wheel, the rear wheel driving trajectory of the risk dynamic obstacle can be determined.

[0096] In another specific implementation process of this implementation manner, if the risk dynamic obstacle is a semi-trailer truck, the semi-trailer truck can be composed of a tractor and a trailer. Here, the driving state information of the risk dynamic obstacle can include the left front wheel speed of the tractor, the right front wheel speed of the tractor, the wheelbase of the tractor, the rear wheel track of the tractor, the wheelbase of the trailer, the rear wheel track of the trailer, and the distance from the connection between the trailer and the tractor to the rear axle of the tractor. The Ackermann steering model can be used to derive the steering radius of the rear wheels based on the left front wheel speed of the tractor, the right front wheel speed of the tractor, the wheelbase of the tractor, the rear wheel track of the tractor, the wheelbase of the trailer, the rear wheel track of the trailer, and the distance from the connection between the trailer and the tractor to the rear axle of the tractor. Furthermore, based on the steering radius of the rear wheels, the driving track of the rear wheels of the risk dynamic obstacle can be determined.

[0097] It can be understood that the driving track of the rear wheels in the steering direction can be calculated according to the steering direction of the risk dynamic obstacle. For example, if the risk dynamic obstacle turns left, the driving track of the left rear wheel can be calculated; if the risk dynamic obstacle turns right, the driving track of the right rear wheel can be calculated. Here, regardless of whether the steering direction of the risk dynamic obstacle is left or right, the specific implementation process in this implementation manner can be used to achieve it, and the detailed process will not be elaborated here.

[0098] In this way, when the front wheel speed of the risk dynamic obstacle can be obtained, the driving track of the rear wheels of the risk dynamic obstacle as the steering collision boundary can be directly determined by using the preset Ackermann steering model, which improves the accuracy and reliability of the steering collision boundary.

[0099] It should be noted that the various specific implementation processes provided in this implementation manner can be combined with each other to implement the steps of the trajectory planning method for the autonomous driving vehicle in this embodiment. For a detailed description, reference can be made to the relevant content in this implementation manner, which will not be elaborated here.

[0100] Optionally, in a possible implementation manner of this embodiment, in step 103, first, in response to the driving state information of the risk dynamic obstacle not including the front wheel speed of the risk dynamic obstacle, based on the driving state information of the risk dynamic obstacle, a preset decomposition algorithm is used to decompose the risk dynamic obstacle to obtain a decomposition unit corresponding to the front part of the risk dynamic obstacle. Second, based on the planned path information of the autonomous vehicle and the decomposition unit corresponding to the front part, speculation processing is performed on the front wheel driving trajectory of the risk dynamic obstacle to obtain the front wheel driving trajectory of the risk dynamic obstacle. Third, based on the planned path information of the autonomous vehicle and the front wheel driving trajectory of the risk dynamic obstacle, a preset offset algorithm is used to calculate a lateral offset. Fourth, based on the lateral offset and the front wheel driving trajectory, the rear wheel driving trajectory of the risk dynamic obstacle is obtained. Fifth, based on the rear wheel driving trajectory of the risk dynamic obstacle, the steering collision boundary is obtained.

[0101] In this implementation manner, the driving state information of the risk dynamic obstacle includes the shape information of the risk dynamic obstacle.

[0102] In a specific implementation process of this implementation manner, first, based on the shape information of the risk dynamic obstacle, a preset decomposition algorithm is used to calculate the radius of the decomposition unit corresponding to the risk dynamic obstacle and the first distance between two adjacent decomposition units. Second, based on the radius of the decomposition unit corresponding to the risk dynamic obstacle and the first distance between two adjacent decomposition units, the risk dynamic obstacle is decomposed to obtain a plurality of decomposition units corresponding to the risk dynamic obstacle. Third, the decomposition unit corresponding to the front part of the risk dynamic obstacle is obtained from the plurality of decomposition units corresponding to the risk dynamic obstacle.

[0103] In this specific implementation process, the shape information may include the length and width of the risk dynamic obstacle.

[0104] One case of this specific implementation process is to calculate the radius of the decomposition unit corresponding to the risk dynamic obstacle and the first distance between two adjacent decomposition units based on the preset number of decomposition units, the length and width of the risk dynamic obstacle.

[0105] In this implementation manner, the decomposition unit may be a circle that can cover a part of the contour of the risk dynamic obstacle. The first distance between two adjacent decomposition units may be the distance between the centers of two adjacent circles.

[0106] In another specific implementation process of this implementation manner, first, each position point in the planned path information of the autonomous vehicle and the orientation angle corresponding to each position point can be obtained. Secondly, based on each position point and the decomposition unit corresponding to the front part, the binary search algorithm can be used to obtain the matching position point that matches the decomposition unit corresponding to the front part. Thirdly, the second distance between the decomposition unit corresponding to the front part and the matching position point can be obtained. Thirdly, based on the orientation angle corresponding to each position point, the normal vector corresponding to each position point can be determined by using a preset vector algorithm. Thirdly, based on each position point in the planned path information of the autonomous vehicle, the normal vector corresponding to each position point, and the second distance, speculation processing can be performed on the front-wheel driving trajectory of the risk dynamic obstacle to obtain the front-wheel driving trajectory of the risk dynamic obstacle.

[0107] In this implementation manner, the matching position point that matches the decomposition unit corresponding to the front part can be the position point in the planned path information that is closest to the center of the circle of the decomposition unit corresponding to the front part, that is, the matching position point is the position point with the smallest distance from the center of the circle of the decomposition unit corresponding to the front part among all the position points of the planned path information.

[0108] One situation of this specific implementation process is that, first, the curvature value of each position point in the planned path information of the autonomous vehicle can be obtained. Secondly, based on the curvature value of each position point in the planned path information of the autonomous vehicle and the second distance, the curvature value of each trajectory point in the front-wheel driving trajectory of the risk dynamic obstacle can be determined. Thirdly, the first distance between two adjacent decomposition units can be obtained. Thirdly, based on the curvature value of each trajectory point and the first distance, the lateral offset can be calculated.

[0109] In this way, in the case where the front-wheel speed of the risk dynamic obstacle cannot be obtained, the risk dynamic obstacle can be decomposed into multiple decomposition units by using a preset decomposition algorithm, and then the rear-wheel driving trajectory of the risk dynamic obstacle can be deduced based on the multiple decomposition units, so as to facilitate the rear-wheel driving trajectory of the risk dynamic obstacle, obtain a more accurate steering collision boundary in this case, and improve the accuracy and reliability of the steering collision boundary.

[0110] It should be noted that the multiple specific implementation processes provided in this implementation manner can be combined with each other to implement the steps of the trajectory planning method for the autonomous vehicle in this embodiment. For a detailed description, reference can be made to the relevant content in this implementation manner, which will not be elaborated here.

[0111] Optionally, in a possible implementation of this embodiment, the driving state information of the autonomous vehicle includes the shape information, speed information, acceleration information, and jerk information of the autonomous vehicle. In step 104, first, based on the shape information of the autonomous vehicle, the planned path information, and the driving state information of the risk dynamic obstacle, the position relationship information between the autonomous vehicle and the risk dynamic obstacle can be obtained. Second, based on the speed information of the autonomous vehicle, the planned path information, and the steering collision boundary, the first time for the autonomous vehicle to reach the steering collision boundary can be obtained. Third, based on the driving state information of the risk dynamic obstacle and the steering collision boundary, the second time for the risk dynamic obstacle to reach the steering collision boundary can be obtained. Third, in response to the position relationship information satisfying a preset first constraint condition, or the speed information, the first time, and the second time of the autonomous vehicle satisfying a preset second constraint condition, based on the speed information, acceleration information, and jerk information of the autonomous vehicle, using a preset objective function, trajectory planning processing is performed on the autonomous vehicle to obtain the planned trajectory of the autonomous vehicle.

[0112] In this implementation, the position relationship information between the autonomous vehicle and the risk dynamic obstacle may include that the autonomous vehicle is in front of the risk dynamic obstacle, the autonomous vehicle is on the side of the risk dynamic obstacle, the autonomous vehicle is behind the risk dynamic obstacle, the first vector between the apex angle of the risk dynamic obstacle and the autonomous vehicle, and the second vector in the driving direction of the risk dynamic obstacle.

[0113] Here, the preset first constraint condition may be that the autonomous vehicle is on the side of the risk dynamic obstacle and the product of the first vector and the second vector is greater than zero. The preset second constraint condition may be that the difference between the first time and the second time is greater than zero, and the speed information of the autonomous vehicle is greater than zero and less than a preset speed threshold.

[0114] In a specific implementation process of this implementation, in response to the position relationship information satisfying the preset first constraint condition, based on the speed information, acceleration information, and jerk information of the autonomous vehicle, using a preset objective function, trajectory planning processing is performed on the autonomous vehicle to obtain the planned trajectory of the autonomous vehicle.

[0115] It can be understood that this specific implementation process may be a process for the autonomous vehicle to perform a cutting-in process.

[0116] In another specific implementation process of this implementation manner, in response to the speed information of the autonomous vehicle, the first time, and the second time satisfying a preset second constraint condition, based on the speed information, acceleration information, and jerk information of the autonomous vehicle, using a preset objective function, trajectory planning processing is performed on the autonomous vehicle to obtain the planned trajectory of the autonomous vehicle.

[0117] Here, the preset second constraint condition may be that the difference between the first time and the second time is greater than zero, and the speed information of the autonomous vehicle is greater than zero and less than a preset speed threshold.

[0118] It can be understood that this specific implementation process may be a process in which the autonomous vehicle performs a yielding process.

[0119] In this way, by according to the steering collision boundary and the position relationship information between the autonomous vehicle and the risk dynamic obstacle, the driving trajectory of the autonomous vehicle's steering can be planned, which can further improve the reliability of the driving trajectory, thereby further improving the safety of the autonomous vehicle's steering.

[0120] It should be noted that the multiple specific implementation processes provided in this implementation manner can be combined with each other to implement the steps of the trajectory planning method of the autonomous vehicle in this embodiment. For a detailed description, reference can be made to the relevant content in this implementation manner, which will not be elaborated here.

[0121] To better understand the method of the embodiment of the present application, the method of the embodiment of the present application will be described below in conjunction with the accompanying drawings and specific application scenarios.

[0122] Figure 2 is a schematic flowchart of a trajectory planning method for an autonomous vehicle provided by another embodiment of the present application, as Figure 2 shown.

[0123] Step 201, obtain the driving state information, planned path information, and driving state information of dynamic obstacles of the autonomous vehicle.

[0124] In this embodiment, the driving state information of the autonomous vehicle may include but is not limited to the speed information, acceleration information, jerk information, and shape information of the autonomous vehicle. The planned path information of the autonomous vehicle may include but is not limited to multiple position points on the path, the forward distance corresponding to each position point, the orientation angle corresponding to each position point, the curvature value corresponding to each position point, etc.

[0125] In this embodiment, the driving state information of the dynamic obstacle may include but is not limited to pose information, speed information, shape information, front wheel speed, etc.

[0126] Here, the shape information of the autonomous vehicle may include, but is not limited to, the length, width, positions of the top corners, etc. of the autonomous vehicle. The shape information of the dynamic obstacle may also include, but is not limited to, the length, width, wheelbase, rear wheel track, etc. of the dynamic obstacle. The front wheel speeds of the dynamic obstacle may include the left front wheel speed and the right front wheel speed of the dynamic obstacle.

[0127] It can be understood that the front wheel speeds of the dynamic obstacle can be obtained in some scenarios and cannot be obtained in other scenarios. For example, the autonomous vehicle and the dynamic obstacle can interact in real time, and the dynamic obstacle can sense the accurate front wheel speeds, while generally the accurate front wheel speeds of the dynamic obstacle cannot be obtained. After determining the risk dynamic obstacle, it can be first determined whether the front wheel speeds of the dynamic obstacle as the risk dynamic obstacle can be obtained.

[0128] Step 202: Based on the planned path information of the autonomous vehicle, use the steering risk area recognition algorithm to obtain the steering risk area.

[0129] In this embodiment, the planned path information of the autonomous vehicle may include, but is not limited to, multiple position points on the path, the forward distance corresponding to each position point, the orientation angle corresponding to each position point, the curvature value corresponding to each position point, etc.

[0130] Optionally, first, the steering risk side of the autonomous vehicle can be determined based on the planned path information of the autonomous vehicle. First, based on the preset adjustment coefficient, the forward distance corresponding to each position point, and the curvature value corresponding to each position point, the exploration distance of the steering risk side corresponding to each position point can be calculated. Second, based on the exploration distance of the steering risk side corresponding to each position point, the steering risk area can be obtained.

[0131] Specifically, based on the preset adjustment coefficient, the forward distance corresponding to each position point, and the curvature value corresponding to each position point, the exploration distance of the steering risk side corresponding to each position point can be calculated as shown in formula (1):

[0132]

[0133] Where, is the exploration distance of the steering risk side corresponding to position point i, and λ i can be the preset adjustment coefficient for adjusting the relationship between the forward distance and the exploration distance, s i can be the forward distance corresponding to position point i, and k i is the curvature value corresponding to this position point i.

[0134] Specifically, based on the exploration distance corresponding to each position point, the planned path of the autonomous vehicle is explored towards the risk side of the autonomous vehicle to obtain an exploration area, and then the exploration area can be used as the steering risk area.

[0135] Here, based on the planned path information of the autonomous vehicle, the steering intention of the autonomous vehicle can be obtained, and then based on the steering intention of the autonomous vehicle, the steering risk side of the autonomous vehicle can be determined. If the steering intention of the autonomous vehicle is to turn left, the right side of the autonomous vehicle is the steering risk side. If the steering intention of the autonomous vehicle is to turn right, the left side of the autonomous vehicle is the steering risk side.

[0136] In addition, it can be understood that when determining the steering risk area, two aspects can be mainly considered. For the path risk side of the autonomous vehicle (the right side is the risk side when turning left, and vice versa when turning right), on the one hand, the intention of the dynamic obstacle at the proximal end of the path is clear, and the randomness at the distal end is greater; on the other hand, based on the path curvature of the autonomous vehicle, the possibility of the dynamic obstacle turning at the place with a larger curvature is greater. Thus, the delineation of the steering risk area can be realized.

[0137] Step 203: Based on the driving state information of the dynamic obstacle and the steering risk area, use the risk dynamic obstacle recognition algorithm to determine the risk dynamic obstacle from the dynamic obstacles.

[0138] In this embodiment, the driving state information of the dynamic obstacle may include the pose information and shape information of the dynamic obstacle.

[0139] Optionally, based on the steering risk area, the pose information and shape information of the dynamic obstacle, it can be determined whether there is an overlapping relationship between the dynamic obstacle and the steering risk area. If there is an overlapping relationship, it can be determined that the dynamic obstacle is a risk dynamic obstacle, and then based on the judgment result of whether there is an overlapping relationship, the risk dynamic obstacle can be determined from the dynamic obstacles.

[0140] Specifically, the following formula (2) can be used to determine the risk dynamic obstacle in the dynamic obstacles:

[0141]

[0142] Among them, can be the risk dynamic obstacle, that is, the collision risk vehicle, V is the dynamic obstacle, that is, the vehicle, is the shape of the dynamic obstacle, that is, the two-dimensional space occupied by the dynamic obstacle, is the steering risk area, can represent an empty set.

[0143] In this embodiment,Figure 3 This is a schematic diagram of the steering risk area of the trajectory planning method for an autonomous vehicle provided by another embodiment of the present application. As Figure 3 shown, the autonomous vehicle (ego) turns left, and its right side is the steering risk side. Among the dynamic obstacles (agents) on the right side, there is a dynamic obstacle that overlaps with the steering risk area (M_risk), and this dynamic obstacle can be used as a risk dynamic obstacle.

[0144] In addition, it should be noted that the principles for judging risk dynamic obstacles may include that a dynamic obstacle vehicle turning in the same direction forms a risk of rubbing against the autonomous vehicle; a dynamic obstacle vehicle located behind the driving direction of the autonomous vehicle does not pose a risk to the vehicle's steering; along the driving direction of the autonomous vehicle, the consideration range of the dynamic obstacle vehicle is proportional to the forward distance.

[0145] Step 204: If the driving state information of the risk dynamic obstacle includes the front wheel speed of the risk dynamic obstacle, then based on the driving state information of the risk dynamic obstacle, use a preset Ackermann steering model to determine the rear wheel driving trajectory of the risk dynamic obstacle, and use the rear wheel driving trajectory of the risk dynamic obstacle as the steering collision boundary.

[0146] In this embodiment, the driving state information of the risk dynamic obstacle may include the front wheel speed and the shape information. The front wheel speed can be regarded as the left front wheel speed and the right front wheel speed. The shape information may include the wheelbase, the rear wheel track, etc. The wheelbase can be the distance between the front and rear axles.

[0147] Optionally, if the risk dynamic obstacle is an ordinary vehicle or an ordinary large vehicle, etc., a preset Ackermann steering model can be used to derive the steering radius of the rear wheel based on the left front wheel speed, the right front wheel speed, the wheelbase, and the rear wheel track of the risk dynamic obstacle, as shown in formulas (3) to (6):

[0148] First, the left front wheel steering angular velocity ω l can be expressed as:

[0149]

[0150] Second, the right front wheel steering angular velocity ω r can be expressed as:

[0151]

[0152] Third, from ω l = ω r , then there is:

[0153]

[0154] Third, derive the steering radius H of the rear wheel as:

[0155]

[0156] Among them, W can be the rear wheel track, and v fl can be the left front wheel speed, and v fr can be the right front wheel speed, and L can be the wheelbase.

[0157] Further, after obtaining the turning radius of the rear wheels, based on the turning radius of the rear wheels, the rear wheel driving trajectory of the risk dynamic obstacle can be determined.

[0158] In addition, after obtaining the turning radius of the rear wheels, the formula (3) can also be used to calculate and return the turning angular velocity of the left front wheel. Furthermore, based on the turning angular velocity of the left front wheel and the left front wheel speed, the turning radius of the left front wheel can be calculated. Based on the turning radius of the left front wheel, the front wheel driving trajectory of the risk dynamic obstacle can be determined.

[0159] It can be understood that here, the area enclosed by the front wheel driving trajectory and the rear wheel driving trajectory of the risk dynamic obstacle can be the area that will collide with the rear wheels of the risk dynamic obstacle, that is, the scraping danger area. The autonomous vehicle should avoid entering this scraping danger area when steering.

[0160] Exemplarily, Figure 4 is a schematic diagram of the steering of a risk dynamic obstacle in the trajectory planning method of an autonomous vehicle provided by another embodiment of the present application. As Figure 4 shown, when the risk dynamic obstacle steers, O is the steering center, R is the overall turning radius, W can be the rear wheel track, and v fl can be the left front wheel speed, and v fr can be the right front wheel speed, L can be the wheelbase, v can be the vehicle speed, risk_line can represent the rear wheel driving trajectory, that is, the steering collision boundary, and H is the turning radius of the rear wheels, that is, the turning radius of the left rear wheel.

[0161] In this embodiment, if the risk dynamic obstacle is a semi-trailer truck, the semi-trailer truck can be composed of a tractor and a trailer. Here, the turning radius of the rear wheels can be derived based on the left front wheel speed of the tractor, the right front wheel speed of the tractor, the wheelbase of the tractor, the rear wheel track of the tractor, the wheelbase of the trailer, the rear wheel track of the trailer, and the distance from the connection between the trailer and the tractor to the rear axle of the tractor by using a preset Ackermann steering model, as shown in formulas (7) to (9):

[0162] First, from the equality of the turning angular velocities of the left / right front wheels of the tractor, we can get:

[0163]

[0164] Secondly, derive the turning radius R of the rear wheels of the tractor 2 It is:

[0165]

[0166] Thirdly, from the geometric relationship, the turning radius R of the rear wheels of the trailer can be obtained 1 It is:

[0167]

[0168] Wherein, v fl can be the wheel speed of the left front wheel of the tractor, v fr can be the wheel speed of the right front wheel of the tractor, W 1 can be the rear wheel track of the trailer, W 2 can be the rear wheel track of the tractor, L 1 can be the wheelbase of the trailer, L 2 can be the wheelbase of the tractor, and b can be the distance from the connection between the trailer and the tractor to the rear axle of the tractor.

[0169] Furthermore, after obtaining the turning radius of the rear wheels of the trailer, based on the turning radius of the rear wheels of the trailer, the rear wheel travel trajectory of the risk dynamic obstacle can be determined.

[0170] Exemplarily, Figure 5 is a schematic diagram of the steering of another risk dynamic obstacle in the trajectory planning method of an autonomous vehicle provided by another embodiment of the present application. As Figure 5 shown, the risk dynamic obstacle can be a semi-trailer. When the risk dynamic obstacle steers, O is the steering center, R is the turning radius of the connection, R 1 is the turning radius of the rear wheels of the trailer, R 2 is the turning radius of the rear wheels of the tractor, R 3 is the turning radius of the left front wheel of the tractor, R 4 is the turning radius of the right front wheel of the tractor, v fl can be the wheel speed of the left front wheel of the tractor, v fr can be the wheel speed of the right front wheel of the tractor, W 1 can be the rear wheel track of the trailer, W 2 can be the rear wheel track of the tractor, L 1 can be the wheelbase of the trailer, L 2 can be the wheelbase of the tractor, b can be the distance from the connection between the trailer and the tractor to the rear axle of the tractor, and Risk line can represent the rear wheel travel trajectory, that is, the steering collision boundary.

[0171] In addition, after obtaining the turning radius of the rear wheels, the angular velocity of the left front wheel can be calculated and returned. Furthermore, based on the angular velocity of the left front wheel and the wheel speed of the left front wheel, the turning radius of the left front wheel can be calculated. Based on the turning radius of the left front wheel, the front wheel driving trajectory of the risk dynamic obstacle can be determined.

[0172] Here, by using the Ackermann steering model, the turning radius of the rear wheels of a general freight truck or a semi-trailer can be well inferred to derive an accurate turning collision boundary and obtain a scraping danger zone.

[0173] It should be noted that in actual application scenarios, when large freight trucks and semi-trailers, as dynamic obstacles, turn at the intersection of the park or at a crossroads, due to their large vehicle weight and inertia, their turning speed is relatively slow. Therefore, it can be assumed that the wheels of large freight trucks and semi-trailers roll completely during the turning process. According to the preset Ackermann steering model, that is, the Ackermann steering geometry model, the normal direction of the wheel center intersects at the instantaneous turning center O on the extension line of the rear axle, as Figure 4 and 5 shown.

[0174] Step 205: If the driving state information of the risk dynamic obstacle does not include the wheel speed of the front wheels of the risk dynamic obstacle, then based on the driving state information of the risk dynamic obstacle, use the preset decomposition algorithm to decompose the risk dynamic obstacle, so as to infer the rear wheel driving trajectory of the risk dynamic obstacle based on the result of the decomposition process and the planned path information of the autonomous vehicle, and use the rear wheel driving trajectory of the risk dynamic obstacle as the turning collision boundary.

[0175] In this embodiment, the driving state information of the risk dynamic obstacle may further include the shape information of the risk dynamic obstacle. The shape information may include the length and width of the risk dynamic obstacle.

[0176] In this embodiment, the result of the decomposition process may include multiple decomposition units corresponding to the risk dynamic obstacle. From the multiple decomposition units corresponding to the risk dynamic obstacle, the decomposition unit corresponding to the front part of the risk dynamic obstacle, that is, the decomposition unit corresponding to the front wheel part of the risk dynamic obstacle, can be obtained.

[0177] First, it can be determined whether the driving state information of the obtained risk dynamic obstacle includes the wheel speed of the front wheels of the risk dynamic obstacle.

[0178] Secondly, when the driving state information of the risk dynamic obstacle does not include the front wheel speed of the risk dynamic obstacle, based on the length and width of the risk dynamic obstacle, the radius of the decomposition unit corresponding to the risk dynamic obstacle and the first distance between two adjacent decomposition units can be calculated by using a preset decomposition algorithm. Furthermore, based on the radius of the decomposition unit corresponding to the risk dynamic obstacle and the first distance between two adjacent decomposition units, the risk dynamic obstacle can be decomposed to obtain multiple decomposition units corresponding to the risk dynamic obstacle, and the decomposition unit corresponding to the front part of the risk dynamic obstacle can be obtained from the multiple decomposition units corresponding to the risk dynamic obstacle.

[0179] Here, based on the number of preset decomposition units, the length and width of the risk dynamic obstacle, the radius r of the decomposition unit corresponding to the risk dynamic obstacle and the first distance d between two adjacent decomposition units can be calculated. The preset decomposition algorithm can be as shown in formulas (10) and (11):

[0180]

[0181]

[0182] where r is the radius of the decomposition unit, n is the number of preset decomposition units, l is the length of the risk dynamic obstacle, w is the width of the risk dynamic obstacle, and d is the first distance between two adjacent decomposition units.

[0183] Exemplarily, Figure 6 is a schematic diagram of multiple decomposition units of a risk dynamic obstacle in a trajectory planning method for an autonomous vehicle provided in another embodiment of the present application. As Figure 6 shown, after performing rotation-invariant primitive decomposition processing on the risk dynamic obstacle, the risk dynamic obstacle can be decomposed into multiple decomposition units. Each decomposition unit can be a circle covering the risk dynamic obstacle. r is the radius of the decomposition unit, l is the length of the risk dynamic obstacle, w is the width of the risk dynamic obstacle, and d is the first distance between two adjacent decomposition units, that is, the distance between the centers of two adjacent circles.

[0184] Thirdly, each position point in the planned path information of the autonomous vehicle and the orientation angle corresponding to each position point can be obtained. Based on each position point and the decomposition unit corresponding to the front part, the matching position point that matches the decomposition unit corresponding to the front part can be obtained by using the binary search algorithm. The second distance between the decomposition unit corresponding to the front part and the matching position point is obtained. Based on the orientation angle corresponding to each position point, the normal vector corresponding to each position point is determined by using a preset vector algorithm. Based on each position point, the normal vector corresponding to each position point, and the second distance in the planned path information of the autonomous vehicle, speculation processing is performed on the front wheel driving trajectory of the risk dynamic obstacle to obtain the front wheel driving trajectory of the risk dynamic obstacle.

[0185] Optionally, based on each position point and the corresponding decomposition unit of the front part, using the binary search algorithm, search for the position point whose distance from the decomposition unit corresponding to the front part satisfies the preset matching condition, and use this position point as the matching position point that matches the decomposition unit corresponding to the front part. Furthermore, the second distance between the decomposition unit corresponding to the front part and the matching position point can be obtained.

[0186] Here, the preset matching condition can be that the distance between the position point and the decomposition unit corresponding to the front part is the minimum distance. The decomposition unit corresponding to the front part can represent the front wheel part of the risk dynamic obstacle.

[0187] In this way, the optimal matching point of the risk dynamic obstacle on the planned path of the autonomous vehicle can be found through binary search, and furthermore, the minimum distance r between the risk dynamic obstacle and the autonomous vehicle can be obtained.

[0188] It can be understood that here, it is assumed that the risk dynamic obstacle on the risk side of the autonomous vehicle is rational during the turning process and will not intentionally deviate from the driving path of the autonomous vehicle. Based on the above assumption, the running trajectory of the decomposition unit corresponding to the front part of the risk dynamic obstacle, that is, the front wheel driving trajectory, can be expressed as

[0189] Optionally, based on the orientation angle corresponding to each position point of the autonomous vehicle, the steering information of the autonomous vehicle can be determined. Furthermore, using the preset vector algorithm corresponding to the steering information of the autonomous vehicle, based on the orientation angle corresponding to each position point, the normal vector corresponding to each position point can be calculated.

[0190] Exemplarily, when the autonomous vehicle is turning left, based on the orientation angle corresponding to each position point, using the preset vector algorithm corresponding to the left turn of the autonomous vehicle, the normal vector corresponding to each position point can be calculated as shown in formula (12):

[0191]

[0192] where can be the normal vector corresponding to position point i, that is, the normal vector to the right of position point i, θ can be the orientation angle corresponding to position point i, and T can be the planned path information of the autonomous vehicle.

[0193] When the autonomous vehicle is turning right, based on the orientation angle corresponding to each position point, using the preset vector algorithm corresponding to the right turn of the autonomous vehicle, the normal vector corresponding to each position point can be calculated as shown in formula (13):

[0194]

[0195] where It can be the normal vector corresponding to position point i, that is, the normal vector to the left of position point i. θ can be the orientation angle corresponding to position point i, and T can be the path planning information of the autonomous vehicle.

[0196] Optionally, after determining the normal vector corresponding to each position point, based on each position point, the normal vector corresponding to each position point, and the second spacing in the path planning information of the autonomous vehicle, speculation processing can be performed on the front-wheel driving trajectory of the risk dynamic obstacle to obtain the front-wheel driving trajectory of the risk dynamic obstacle.

[0197] Specifically, speculation processing on the front-wheel driving trajectory of the risk dynamic obstacle based on each position point, the normal vector corresponding to each position point, and the second spacing in the path planning information of the autonomous vehicle can be as shown in formula (14):

[0198]

[0199] where is position point i in the path planning information of the autonomous vehicle, Nv i is the normal vector corresponding to this position point i, q is the second spacing, that is, the minimum distance between the risk dynamic obstacle and the autonomous vehicle, is trajectory point i in the speculated front-wheel driving trajectory of the risk dynamic obstacle. Based on each speculated trajectory point, the front-wheel driving trajectory of the risk dynamic obstacle can be obtained.

[0200] Exemplarily, Figure 7 is a schematic diagram of the front-wheel driving trajectory of the risk dynamic obstacle in the trajectory planning method of the autonomous vehicle provided by another embodiment of this application. As Figure 7 shown, the front-wheel driving trajectory of the risk dynamic obstacle is speculated through the path planning information of the autonomous vehicle and the second spacing q. Among them, the circle can be the decomposition unit corresponding to the front part of the risk dynamic obstacle. The trajectory in front of the front decomposition unit of the risk dynamic obstacle can be the inferred trajectory (black dashed line), and the trajectory behind the front decomposition unit of the risk dynamic obstacle can be the spliced trajectory (red dashed line). The inferred trajectory and the spliced trajectory can form the speculated front-wheel driving trajectory of the risk dynamic obstacle.

[0201] Again, based on the planned path information of the autonomous vehicle and the front-wheel driving trajectory of the risk dynamic obstacle, the lateral offset can be calculated using a preset offset algorithm. Specifically, the curvature value of each position point in the planned path information of the autonomous vehicle can be obtained first, and based on the curvature value of each position point in the planned path information of the autonomous vehicle and the second spacing, the curvature value of each trajectory point in the front-wheel driving trajectory of the risk dynamic obstacle can be determined. The first spacing between two adjacent decomposition units can be obtained, and based on the curvature value of each trajectory point and the first spacing, the lateral offset can be calculated. Based on the lateral offset and the front-wheel driving trajectory, the rear-wheel driving trajectory of the risk dynamic obstacle can be obtained to use the rear-wheel driving trajectory of the risk dynamic obstacle as the steering collision boundary.

[0202] Optionally, here, based on the curvature value of each position point in the planned path information of the autonomous vehicle and the second spacing, the curvature value of each trajectory point in the front-wheel driving trajectory of the risk dynamic obstacle is determined It can be as shown in formula (15):

[0203]

[0204] Where, is the curvature value of the corresponding position point i of the autonomous vehicle, and q is the second spacing.

[0205] Using the preset offset algorithm, the lateral offset δ between the subsequent circle j movement trajectory point of the circle (i = 0) corresponding to the front part and the circle movement trajectory point of the decomposition unit corresponding to the front part can be as shown in formula (16):

[0206]

[0207] Where, can be the curvature value of each trajectory point in the front-wheel driving trajectory, that is, the speculated movement trajectory of the circle of the decomposition unit corresponding to the front part, d i,j can be the first spacing, that is, the distance between the center of the i decomposition unit and the center of the adjacent j decomposition unit.

[0208] Here, based on the lateral offset and the front-wheel driving trajectory, the rear-wheel driving trajectory of the risk dynamic obstacle can be calculated.

[0209] Exemplarily, Figure 8 is a schematic diagram of the rear-wheel driving trajectory of the risk dynamic obstacle in the trajectory planning method of the autonomous vehicle provided by another embodiment of the present application. As Figure 8As shown, the red trajectory point (1) on the left can represent the rear-wheel driving trajectory of the risk dynamic obstacle, and the red trajectory point (2) on the right can represent the front-wheel driving trajectory of the risk dynamic obstacle. The area formed between the two trajectories can be the area where a collision with the rear wheel of the risk dynamic obstacle may occur, that is, the scraping danger area. The autonomous vehicle should avoid entering this scraping danger area when steering.

[0210] In addition, it can be understood that the method of step 205 can be applied to different types of risk dynamic obstacles. For example, the risk dynamic obstacle can include, but is not limited to, ordinary vehicles, large trucks, semi-trailers, etc.

[0211] Step 206: Based on the steering collision boundary, the driving state information of the autonomous vehicle, the planned path information, and the driving state information of the risk dynamic obstacle, perform trajectory planning processing on the autonomous vehicle to obtain the planned trajectory of the autonomous vehicle.

[0212] In this embodiment, the driving state information of the autonomous vehicle includes the shape information, speed information, acceleration information, and jerk information of the autonomous vehicle.

[0213] Optionally, first, based on the shape information of the autonomous vehicle, the position points in the planned path information of the autonomous vehicle, and the pose information in the driving state information of the risk dynamic obstacle, obtain the position relationship information between the autonomous vehicle and the risk dynamic obstacle. Second, based on the speed information of the autonomous vehicle, the planned path information, and the steering collision boundary, obtain the first time when the autonomous vehicle reaches the steering collision boundary. Third, based on the driving state information of the risk dynamic obstacle and the steering collision boundary, obtain the second time when the risk dynamic obstacle reaches the steering collision boundary. Fourth, in response to the position relationship information satisfying the preset first constraint condition, or the speed information, the first time, and the second time of the autonomous vehicle satisfying the preset second constraint condition, based on the speed information, acceleration information, and jerk information of the autonomous vehicle, use the preset objective function to perform trajectory planning processing on the autonomous vehicle to obtain the planned trajectory of the autonomous vehicle.

[0214] Here, based on the shape information of the autonomous vehicle, the position points in the planned path information of the autonomous vehicle, and the pose information in the driving state information of the risk dynamic obstacle, the position relationship information between the autonomous vehicle and the risk dynamic obstacle can be obtained. The position relationship information between the autonomous vehicle and the risk dynamic obstacle can include that the autonomous vehicle is in front of the risk dynamic obstacle, the autonomous vehicle is on the side of the risk dynamic obstacle, the autonomous vehicle is behind the risk dynamic obstacle, and the first vector between the apex angle of the risk dynamic obstacle and the autonomous vehicle, and the second vector in the driving direction of the risk dynamic obstacle.

[0215] Optionally, the preset first constraint condition may be that the autonomous vehicle is located on the side of the risk dynamic obstacle, and the product of the first vector and the second vector is greater than zero. The preset second constraint condition may be that the difference between the first time and the second time is greater than zero, and the speed information of the autonomous vehicle is greater than zero and less than the preset speed threshold. The preset speed threshold may be preset according to the design situation of the autonomous vehicle or the road environment situation.

[0216] In this embodiment, in one case, when the autonomous vehicle is located on the side of the risk dynamic obstacle and the product of the first vector and the second vector is greater than zero, trajectory planning processing may be performed on the autonomous vehicle based on the speed information, acceleration information, and jerk information of the autonomous vehicle using the preset objective function to obtain the planned trajectory of the autonomous vehicle.

[0217] It can be understood that here, in this case, the autonomous vehicle can perform a cutting-in-ahead operation.

[0218] Exemplarily, when the autonomous vehicle is located on the side of the risk dynamic obstacle and the product of the first vector and the second vector is greater than zero, the preset objective function y may be as shown in formula (17):

[0219]

[0220] Wherein, may be the speed information of the autonomous vehicle, may be the acceleration information of the autonomous vehicle, may be the jerk information of the autonomous vehicle, λ 1 、λ 2 、λ 3 are all preset adjustment coefficients, may be the first vector, may be the second vector, and i may be the trajectory point of the autonomous vehicle.

[0221] In another case, when the difference between the first time and the second time is greater than zero and the speed information of the autonomous vehicle is greater than zero and less than the preset speed threshold, trajectory planning processing may be performed on the autonomous vehicle based on the speed information, acceleration information, and jerk information of the autonomous vehicle using the preset objective function to obtain the planned trajectory of the autonomous vehicle.

[0222] It can be understood that here, in this case, the autonomous vehicle can perform a yielding operation.

[0223] Exemplarily, when the difference between the first time and the second time is greater than zero and the speed information of the autonomous vehicle is greater than zero and less than the preset speed threshold, the preset objective function y may be as shown in formula (18):

[0224]

[0225]

[0226] 0 < v i < v max

[0227] i = 0, 1,...., T

[0228] It can be the speed information of the autonomous vehicle, It can be the acceleration information of the autonomous vehicle, It can be the jerk information of the autonomous vehicle, λ 1 、λ 2 、λ 3 All are preset adjustment coefficients, It can be the first time, It can be the second time, v i It can be the speed information of the autonomous vehicle at trajectory point i, v max It can be a preset speed threshold, and i can be the trajectory point of the autonomous vehicle.

[0229] Another situation can be that when the entire autonomous vehicle is in front of the risk dynamic obstacle, the product of the first vector and the second vector is greater than zero, and the process of the autonomous vehicle planning a steering trajectory can not consider the risk dynamic obstacle.

[0230] Another situation can be that when the autonomous vehicle is located at the side rear of the risk dynamic obstacle and the product of the first vector and the second vector is less than zero, it can be determined whether the speed information, the first time, and the second time of the autonomous vehicle satisfy the preset second constraint condition. If the speed information, the first time, and the second time of the autonomous vehicle satisfy the preset second constraint condition, then based on the speed information, acceleration information, and jerk information of the autonomous vehicle, using the preset objective function, trajectory planning processing can be performed on the autonomous vehicle to obtain the planned trajectory of the autonomous vehicle. Or, the vehicle speed of the autonomous vehicle can also be directly reduced based on the steering collision boundary to obtain the corresponding planned trajectory of the autonomous vehicle, so as to control the autonomous vehicle to follow and drive outside the steering collision boundary based on the planned trajectory of the autonomous vehicle. In this way, the autonomous vehicle can be prevented from cutting in line.

[0231] In addition, another situation can be that when the autonomous vehicle is already inside the steering collision boundary risk line of the risk dynamic obstacle and the above-mentioned preset objective function has no solution, an alarm message can be reported to request takeover.

[0232] Exemplarily, Figures 9a to 9c is a schematic diagram of the relationship between an autonomous vehicle and the position of a risk dynamic obstacle in the trajectory planning method of an autonomous vehicle provided by another embodiment of the present application. As Figure 9a shown, the autonomous vehicle is located in front of the risk dynamic obstacle. Here, "in front" may mean that the entire autonomous vehicle is in front of the risk dynamic obstacle, and it may not be directly in front. may be the first vector, may be the second vector, and the product of the first vector and the second vector is greater than zero. The process of the autonomous vehicle planning the steering trajectory may not need to consider the risk dynamic obstacle; as Figure 9b shown, the autonomous vehicle is located on the side of the risk dynamic obstacle, may be the vector from the upper left corner of the risk dynamic obstacle to the upper right corner of the autonomous vehicle, that is, the first vector, may be the vector in the driving direction of the risk dynamic obstacle, that is, the second vector. The product of the first vector and the second vector is greater than zero. The method based on the preset first constraint condition or the method based on the preset second constraint condition may be used to plan the steering trajectory; as Figure 9c shown, the autonomous vehicle is located behind the risk dynamic obstacle. Here, "behind" may mean that the entire autonomous vehicle is behind the risk dynamic obstacle, and it may not be directly behind. may be the first vector, may be the second vector, and the product of the first vector and the second vector is less than zero. The method based on the preset second constraint condition may be used to plan the steering trajectory. Alternatively, the vehicle speed of the autonomous vehicle may be directly reduced based on the steering collision boundary to obtain a planned trajectory of the autonomous vehicle that follows outside the steering collision boundary.

[0233] It can be understood that other positions of the autonomous vehicle and the risk dynamic obstacle may be selected to determine the first vector. For example, the lower vertex angle on the side of the autonomous vehicle close to the risk dynamic obstacle and the lower vertex angle on the side of the risk dynamic obstacle close to the autonomous vehicle may be selected to determine the first vector, or the lower vertex angle on the side of the autonomous vehicle close to the risk dynamic obstacle and the upper vertex angle on the side of the risk dynamic obstacle close to the autonomous vehicle may be selected to determine the first vector. It can be selected according to actual business needs and will not be specifically limited here.

[0234] In this way, by adopting the technical solution in this embodiment, the steering danger zone caused by the inner wheel difference phenomenon can be effectively estimated based on limited observation data such as the driving state information of the autonomous vehicle, the planned path information, and the driving state information of the dynamic obstacle, and the risk estimation of this type of risk scenario is realized.

[0235] Moreover, according to the steering danger area, the driving trajectory during the steering process of the autonomous vehicle can be planned, effectively reducing the scraping risk, thereby improving the safety and reliability of the autonomous vehicle steering.

[0236] It should be noted that, for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that this application is not limited by the described action sequence, because according to this application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0237] In the above embodiments, the descriptions of the various embodiments have their own emphases. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0238] Figure 10 The structural block diagram of the trajectory planning device of the autonomous vehicle provided by an embodiment of the present application is shown as Figure 10 shown. The trajectory planning device 800 of the autonomous vehicle in this embodiment may include an acquisition unit 1001, a determination unit 1002, an obtaining unit 1003, and a planning unit 1004. Among them, the acquisition unit 1001 is used to acquire the driving state information of the autonomous vehicle, the planned path information, and the driving state information of the dynamic obstacle; wherein, the autonomous vehicle has a steering intention; the determination unit 1002 is used to determine a risk dynamic obstacle from the dynamic obstacles based on the planned path information of the autonomous vehicle and the driving state information of the dynamic obstacle by using a risk dynamic obstacle recognition strategy; the obtaining unit 1003 is used to obtain a steering collision boundary based on the planned path information of the autonomous vehicle and the driving state information of the risk dynamic obstacle by using a steering collision boundary recognition strategy; the planning unit 1004 is used to perform trajectory planning processing on the autonomous vehicle based on the steering collision boundary, the driving state information of the autonomous vehicle, the planned path information, and the driving state information of the risk dynamic obstacle to obtain the planned trajectory of the autonomous vehicle.

[0239] It should be noted that part or all of the trajectory planning device of the autonomous vehicle in this embodiment can be an application located on the local terminal, or can also be a plug-in or software development kit (SDK) and other functional units set in the application located on the local terminal, or can also be a processing engine in the network-side server, or can also be a distributed system located on the network side. For example, the processing engine or distributed system in the data processing platform on the network side, etc. This embodiment does not make special limitations on this.

[0240] It can be understood that the application can be a native app installed on the local terminal, or can also be a web app of a browser on the local terminal. This embodiment does not make limitations on this.

[0241] Optionally, in a possible implementation manner of this embodiment, the risk dynamic obstacle recognition strategy includes a steering risk area recognition algorithm and a risk dynamic obstacle recognition algorithm. The determination unit 1002 can be used to obtain a steering risk area based on the planned path information of the autonomous vehicle by using the steering risk area recognition algorithm; and determine a risk dynamic obstacle from the dynamic obstacles based on the driving state information of the dynamic obstacles and the steering risk area by using the risk dynamic obstacle recognition algorithm.

[0242] Optionally, in a possible implementation manner of this embodiment, the obtaining unit 1003 can be used to, in response to the driving state information of the risk dynamic obstacle including the front wheel speed of the risk dynamic obstacle, determine the rear wheel driving trajectory of the risk dynamic obstacle based on the driving state information of the risk dynamic obstacle by using a preset Ackermann steering model; and obtain the steering collision boundary based on the rear wheel driving trajectory of the risk dynamic obstacle.

[0243] Optionally, in a possible implementation of this embodiment, the obtaining unit 1003 may be configured to, in response to the driving state information of the risk dynamic obstacle not including the front wheel speed of the risk dynamic obstacle, based on the driving state information of the risk dynamic obstacle, use a preset decomposition algorithm to perform a decomposition process on the risk dynamic obstacle to obtain a decomposition unit corresponding to the front part of the risk dynamic obstacle; based on the planned path information of the autonomous vehicle and the decomposition unit corresponding to the front part, perform a speculation process on the front wheel driving trajectory of the risk dynamic obstacle to obtain the front wheel driving trajectory of the risk dynamic obstacle; based on the planned path information of the autonomous vehicle and the front wheel driving trajectory of the risk dynamic obstacle, use a preset offset algorithm to calculate a lateral offset; based on the lateral offset and the front wheel driving trajectory, obtain the rear wheel driving trajectory of the risk dynamic obstacle; and based on the rear wheel driving trajectory of the risk dynamic obstacle, obtain the steering collision boundary.

[0244] Optionally, in a possible implementation of this embodiment, the driving state information of the risk dynamic obstacle includes the shape information of the risk dynamic obstacle. The obtaining unit 1003 may be configured to, based on the shape information of the risk dynamic obstacle, use a preset decomposition algorithm to calculate the radius of the decomposition unit corresponding to the risk dynamic obstacle and the first distance between two adjacent decomposition units; based on the radius of the decomposition unit corresponding to the risk dynamic obstacle and the first distance between two adjacent decomposition units, perform a decomposition process on the risk dynamic obstacle to obtain a plurality of decomposition units corresponding to the risk dynamic obstacle; and obtain the decomposition unit corresponding to the front part of the risk dynamic obstacle from the plurality of decomposition units corresponding to the risk dynamic obstacle.

[0245] Optionally, in a possible implementation of this embodiment, the obtaining unit 1003 may be configured to obtain each position point in the planned path information of the autonomous vehicle and the orientation angle corresponding to each position point; based on each position point and the decomposition unit corresponding to the front part, use a binary search algorithm to obtain a matching position point that matches the decomposition unit corresponding to the front part; obtain the second distance between the decomposition unit corresponding to the front part and the matching position point; based on the orientation angle corresponding to each position point, use a preset vector algorithm to determine the normal vector corresponding to each position point; and based on each position point in the planned path information of the autonomous vehicle, the normal vector corresponding to each position point, and the second distance, perform a speculation process on the front wheel driving trajectory of the risk dynamic obstacle to obtain the front wheel driving trajectory of the risk dynamic obstacle.

[0246] Optionally, in a possible implementation of this embodiment, the obtaining unit 1003 may be configured to obtain the curvature value of each position point in the planned path information of the autonomous vehicle; determine the curvature value of each trajectory point in the front-wheel driving trajectory of the risk dynamic obstacle based on the curvature value of each position point in the planned path information of the autonomous vehicle and the second distance; obtain the first distance between two adjacent decomposition units; and calculate the lateral offset based on the curvature value of each trajectory point and the first distance.

[0247] Optionally, in a possible implementation of this embodiment, the driving state information of the autonomous vehicle includes the shape information, speed information, acceleration information, and jerk information of the autonomous vehicle. The planning unit 1004 may be configured to obtain the position relationship information between the autonomous vehicle and the risk dynamic obstacle based on the shape information of the autonomous vehicle, the planned path information, and the driving state information of the risk dynamic obstacle; obtain the first time for the autonomous vehicle to reach the steering collision boundary based on the speed information of the autonomous vehicle, the planned path information, and the steering collision boundary; obtain the second time for the risk dynamic obstacle to reach the steering collision boundary based on the driving state information of the risk dynamic obstacle and the steering collision boundary; and in response to the position relationship information satisfying a preset first constraint condition, or the speed information, the first time, and the second time of the autonomous vehicle satisfying a preset second constraint condition, perform trajectory planning processing on the autonomous vehicle based on the speed information, acceleration information, and jerk information of the autonomous vehicle by using a preset objective function to obtain the planned trajectory of the autonomous vehicle.

[0248] In this embodiment, the driving state information of the autonomous vehicle, the planned path information, and the driving state information of the dynamic obstacle can be obtained by an acquisition unit; wherein, the autonomous vehicle has a steering intention. Further, a determination unit can determine a risk dynamic obstacle from the dynamic obstacles based on the planned path information of the autonomous vehicle and the driving state information of the dynamic obstacle by using a risk dynamic obstacle recognition strategy. An obtaining unit can obtain a steering collision boundary based on the planned path information of the autonomous vehicle and the driving state information of the risk dynamic obstacle by using a steering collision boundary recognition strategy, so that a planning unit can perform trajectory planning processing on the autonomous vehicle based on the steering collision boundary, the driving state information of the autonomous vehicle, the planned path information, and the driving state information of the risk dynamic obstacle to obtain the planned trajectory of the autonomous vehicle. Since the steering collision boundary between the autonomous vehicle and the dynamic obstacle can be effectively estimated according to the driving state information of the autonomous vehicle, the planned path information, and the driving state information of the dynamic obstacle, and the steering collision boundary is the boundary that is dangerous to the steering of the host vehicle formed when the dynamic obstacle steers, the risk estimation of the steering scenario of the autonomous vehicle is realized. Moreover, based on the steering collision boundary, the driving trajectory during the steering of the autonomous vehicle can be planned, which can effectively reduce the scraping risk, thereby improving the safety and reliability of the steering of the autonomous vehicle.

[0249] In the technical solution of the present application, the collection, storage, use, processing, transmission, provision, and disclosure of the user's personal information involved, such as the user's image and attribute data, etc., all comply with the provisions of relevant laws and regulations and do not violate public order and good customs.

[0250] According to an embodiment of the present application, the present application further provides an electronic device, a readable storage medium, and a computer program product.

[0251] According to an embodiment of the present application, further, an autonomous vehicle including the provided electronic device is provided. The autonomous vehicle can include a driverless vehicle at L2 level or above. For example, the driverless vehicle can include an unmanned logistics vehicle, an unmanned patrol vehicle, an unmanned delivery vehicle, etc.

[0252] Figure 11FIG. shows a schematic block diagram of an exemplary electronic device 1100 that can be used to implement embodiments of the present application. The electronic device is intended to represent various forms of digital computers, such as, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, personal digital processors, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely exemplary and are not intended to limit the implementation of the present application described and / or claimed herein.

[0253] As Figure 11 shown, the electronic device 1100 includes a computing unit 1101 that can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 1102 or a computer program loaded from a storage unit 1108 into a random access memory (RAM) 1103. In the RAM 1103, various programs and data required for the operation of the electronic device 1100 can also be stored. The computing unit 1101, the ROM 1102, and the RAM 1103 are connected to each other via a bus 1104. An input / output (I / O) interface 1106 is also connected to the bus 1104.

[0254] A plurality of components in the electronic device 1100 are connected to the I / O interface 1105, including: an input unit 1106, such as a keyboard, a mouse, etc.; an output unit 1107, such as various types of displays, speakers, etc.; a storage unit 1108, such as a magnetic disk, an optical disk, etc.; and a communication unit 1109, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1109 allows the electronic device 1100 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0255] The computing unit 1101 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 1101 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 1101 executes the various methods and processes described above, such as the trajectory planning method for an autonomous vehicle. For example, in some embodiments, the trajectory planning method for an autonomous vehicle can be implemented as a computer software program that is tangibly included in a machine-readable medium, such as the storage unit 1108. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 1100 via the ROM 1102 and / or the communication unit 1109. When the computer program is loaded into the RAM 1103 and executed by the computing unit 1101, one or more steps of the trajectory planning method for the autonomous vehicle described above can be executed. Alternatively, in other embodiments, the computing unit 1101 can be configured to execute the trajectory planning method for the autonomous vehicle in any other suitable manner (e.g., by means of firmware).

[0256] The various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-chip systems (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special or general programmable processor, receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0257] The program code for implementing the methods of this application can be written in any combination of one or more programming languages. These program codes can be provided to the processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing devices, such that when the program codes are executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The program codes can be executed entirely on the machine, partially on the machine, as an independent software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0258] In the context of this application, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0259] To provide for interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can also be used to provide for interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic, speech, or tactile input).

[0260] The systems and techniques described herein can be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of a communication network include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0261] A computer system may include a client and a server. The client and the server are generally far away from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the respective computers and having a client-server relationship with each other. The server may be a cloud server, or a server of a distributed system, or a server combined with a blockchain.

[0262] It should be understood that various forms of processes shown above can be used, steps can be reordered, added or deleted. For example, the steps described in the present disclosure can be executed in parallel, sequentially or in a different order, as long as the desired results of the technical solutions disclosed in the present application can be achieved, and no limitations are imposed herein.

[0263] The above specific embodiments do not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. A trajectory planning method for an autonomous driving vehicle, characterized in that: The method comprises: Acquiring driving state information, planned path information, and driving state information of dynamic obstacles of the autonomous driving vehicle; wherein the autonomous driving vehicle has a turning intention; Based on the planned path information of the autonomous driving vehicle and the driving state information of the dynamic obstacles, a risky dynamic obstacle is determined from the dynamic obstacles using a risky dynamic obstacle identification strategy; Based on the planned path information of the autonomous driving vehicle and the driving state information of the risky dynamic obstacle, a steering collision boundary is obtained by using a steering collision boundary identification strategy; Based on the steering collision boundary, the driving state information of the autonomous driving vehicle, the planned path information, and the driving state information of the risk dynamic obstacle, performing trajectory planning processing on the autonomous driving vehicle to obtain a planned trajectory of the autonomous driving vehicle; The method of obtaining a steering collision boundary by using a steering collision boundary identification strategy based on the planned path information of the autonomous driving vehicle and the driving state information of the risky dynamic obstacle includes: In response to the driving state information of the risky dynamic obstacle not including the front wheel speed of the risky dynamic obstacle, based on the driving state information of the risky dynamic obstacle, using a preset decomposition algorithm, decomposing the risky dynamic obstacle to obtain a decomposition unit corresponding to the front of the risky dynamic obstacle; Based on the planned path information of the autonomous driving vehicle and the decomposition unit corresponding to the front portion, the front wheel driving trajectory of the risky dynamic obstacle is inferred to obtain the front wheel driving trajectory of the risky dynamic obstacle; Based on the planned path information of the autonomous driving vehicle and the front wheel driving trajectory of the risk dynamic obstacle, a lateral offset is calculated using a preset offset algorithm; Based on the lateral offset and the front wheel driving trajectory, the rear wheel driving trajectory of the risk dynamic obstacle is obtained, so as to use the rear wheel driving trajectory as a steering collision boundary.

2. The method according to claim 1, characterized in that The risk dynamic obstacle identification strategy includes a steering risk area identification algorithm and a risk dynamic obstacle identification algorithm. The risk dynamic obstacle identification strategy is used based on the planned path information of the autonomous driving vehicle and the driving state information of the dynamic obstacle to determine the risk dynamic obstacle from the dynamic obstacles, including: Based on the planned path information of the autonomous driving vehicle, using the turning risk area identification algorithm, obtaining a turning risk area; Based on the driving state information of the dynamic obstacle and the turning risk area, the risky dynamic obstacle is determined from the dynamic obstacles using the risky dynamic obstacle identification algorithm.

3. The method according to claim 1, characterized in that The method of obtaining a steering collision boundary by using a steering collision boundary identification strategy based on the planned path information of the autonomous driving vehicle and the driving state information of the risky dynamic obstacle includes: In response to the driving state information of the risky dynamic obstacle including the front wheel speed of the risky dynamic obstacle, based on the driving state information of the risky dynamic obstacle, a rear wheel driving trajectory of the risky dynamic obstacle is determined by using a preset Ackerman steering model; The steering collision boundary is obtained based on the rear wheel driving trajectory of the risk dynamic obstacle.

4. The method according to claim 1, characterized in that: The driving state information of the risky dynamic obstacle includes shape information of the risky dynamic obstacle. Based on the driving state information of the risky dynamic obstacle, the risky dynamic obstacle is decomposed by using a preset decomposition algorithm to obtain a decomposition unit corresponding to the front of the risky dynamic obstacle, including: Based on the shape information of the risk dynamic obstacle, a preset decomposition algorithm is used to calculate the radius of the decomposition unit corresponding to the risk dynamic obstacle and the first distance between two adjacent decomposition units; Based on the radius of the decomposition unit corresponding to the risk dynamic obstacle and the first distance between two adjacent decomposition units, the risk dynamic obstacle is decomposed to obtain a plurality of decomposition units corresponding to the risk dynamic obstacle; The decomposition unit corresponding to the front of the risky dynamic obstacle is obtained from the multiple decomposition units corresponding to the risky dynamic obstacle.

5. The method according to claim 1, characterized in that The method of performing inference processing on the front wheel driving trajectory of the risk dynamic obstacle based on the planned path information of the autonomous driving vehicle and the decomposition unit corresponding to the front portion to obtain the front wheel driving trajectory of the risk dynamic obstacle includes: Obtain each position point and the orientation angle corresponding to each position point in the planned path information of the autonomous driving vehicle; Based on each position point and the decomposition unit corresponding to the front portion, a matching position point matching the decomposition unit corresponding to the front portion is obtained by using a binary search algorithm; Obtaining a second distance between the decomposition unit corresponding to the front portion and the matching position point; Based on the orientation angle corresponding to each position point, a normal vector corresponding to each position point is determined using a preset vector algorithm; Based on each position point in the planned path information of the autonomous driving vehicle, the normal vector corresponding to each position point, and the second distance, the front wheel driving trajectory of the risky dynamic obstacle is inferred to obtain the front wheel driving trajectory of the risky dynamic obstacle.

6. The method according to claim 5, characterized in that The lateral offset is calculated based on the planned path information of the autonomous driving vehicle and the front wheel driving trajectory of the risk dynamic obstacle using a preset offset algorithm, including: Obtaining a curvature value of each position point in the planned path information of the autonomous driving vehicle; Determine the curvature value of each track point in the front wheel driving track of the risk dynamic obstacle based on the curvature value of each position point in the planned path information of the autonomous driving vehicle and the second spacing; Get the first distance between two adjacent decomposition units; The lateral offset is calculated based on the curvature value of each trajectory point and the first spacing.

7. The method according to claim 1, characterized in that The driving state information of the autonomous driving vehicle includes shape information, speed information, acceleration information, and jerk information of the autonomous driving vehicle. Based on the steering collision boundary, the driving state information of the autonomous driving vehicle, the planned path information, and the driving state information of the risk dynamic obstacle, trajectory planning processing is performed on the autonomous driving vehicle to obtain the planned trajectory of the autonomous driving vehicle, including: Based on the shape information, the planned path information and the driving state information of the risky dynamic obstacle of the autonomous driving vehicle, obtaining the positional relationship information between the autonomous driving vehicle and the risky dynamic obstacle; Based on the speed information, the planned path information, and the steering collision boundary of the autonomous driving vehicle, obtaining a first time when the autonomous driving vehicle reaches the steering collision boundary; Based on the driving state information of the risky dynamic obstacle and the turning collision boundary, obtaining a second time when the risky dynamic obstacle reaches the turning collision boundary; In response to the position relationship information satisfying a preset first constraint condition, or the speed information, the first time, and the second time of the autonomous driving vehicle satisfying a preset second constraint condition, based on the speed information, acceleration information, and jerk information of the autonomous driving vehicle, a trajectory planning process is performed on the autonomous driving vehicle using a preset objective function to obtain a planned trajectory of the autonomous driving vehicle, wherein: The preset first constraint condition is that the autonomous driving vehicle is located on the side of the risky dynamic obstacle, and the product of the first vector and the second vector is greater than zero. The preset second constraint condition is that the difference between the first time and the second time is greater than zero, and the speed information of the autonomous driving vehicle is greater than zero and less than a preset speed threshold.

8. A trajectory planning device for an autonomous driving vehicle, characterized in that: The device comprises: An acquisition unit, used to acquire driving state information, planned path information, and driving state information of a dynamic obstacle of the autonomous driving vehicle; wherein the autonomous driving vehicle has a turning intention; a determination unit, configured to determine a risky dynamic obstacle from the dynamic obstacles by using a risky dynamic obstacle identification strategy based on the planned path information of the autonomous driving vehicle and the driving state information of the dynamic obstacle; an obtaining unit for, in response to the driving state information of the risky dynamic obstacle not including the front wheel speed of the risky dynamic obstacle, decomposing the risky dynamic obstacle based on the driving state information of the risky dynamic obstacle by using a preset decomposition algorithm to obtain a decomposition unit corresponding to the front of the risky dynamic obstacle; inferring the front wheel driving trajectory of the risky dynamic obstacle based on the planned path information of the autonomous driving vehicle and the decomposition unit corresponding to the front to obtain the front wheel driving trajectory of the risky dynamic obstacle; calculating a lateral offset based on the planned path information of the autonomous driving vehicle and the front wheel driving trajectory of the risky dynamic obstacle by using a preset offset algorithm; obtaining a rear wheel driving trajectory of the risky dynamic obstacle based on the lateral offset and the front wheel driving trajectory, so as to use the rear wheel driving trajectory as a steering collision boundary; A planning unit is used to perform trajectory planning processing on the autonomous driving vehicle based on the steering collision boundary, the driving state information of the autonomous driving vehicle, the planned path information, and the driving state information of the risky dynamic obstacle to obtain the planned trajectory of the autonomous driving vehicle.

9. An electronic device, comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-7.

11. A computer program product, comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 7.

12. An autonomous driving vehicle comprising the electronic device as claimed in claim 9.

Images