Track planning method and device, automatic driving equipment and readable storage medium

Abstract

The invention discloses a trajectory planning method and device, automatic driving equipment and a readable storage medium. The method comprises the following steps: acquiring perception information, a transverse track state and a longitudinal track state of the automatic driving equipment, wherein the perception information comprises pose information, speed information and environment information; in response to the fact that the transverse track state is a first state, a first transverse track is determined based on the pose information, and the first transverse track is used for keeping extension of the current course angle; in response to the fact that the longitudinal track state is a second state, a first longitudinal track is determined based on the speed information, and the first longitudinal track is used for representing a speed track of stable deceleration; and performing current frame trajectory planning on the automatic driving equipment based on the first transverse trajectory and the first longitudinal trajectory. According to the method, the discontinuity caused by the track state of the current frame can be greatly improved, the discomfort of passengers caused by strong braking is improved, and the driving experience of the passengers is improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent driving technology, and specifically to a trajectory planning method, device, automatic driving equipment and readable storage medium. Background Art

[0002] For vehicles equipped with autonomous driving technology, in relevant scenarios, such as highway pilot assisted driving and city pilot assisted driving, trajectory planning based on environmental perception is required to automatically calculate a collision-free, executable trajectory (including path and speed information) so that the vehicle can be safely driven from the starting point to the destination.

[0003] The trajectory planning of autonomous driving is usually carried out continuously in multiple planning cycles. After each frame of trajectory planning is completed, the trajectory planning module in the vehicle will send the frame trajectory to the controller for use. However, when special circumstances occur, such as: the front vehicle brakes suddenly with great acceleration, the current frame trajectory planning may fail. At this time, the trajectory planning module may send a forced braking command to the controller to directly execute forced braking. Therefore, passengers may feel uncomfortable, which in turn affects the driving experience. Summary of the invention

[0004] Embodiments of the present application provide a trajectory planning method, apparatus, autonomous driving equipment, and readable storage medium to improve the passenger's driving experience when the current frame trajectory planning fails.

[0005] In order to solve the above technical problems, the embodiments of the present application disclose the following technical solutions:

[0006] In a first aspect, a trajectory planning method is provided, which is applied to an autonomous driving device, and the method comprises:

[0007] Acquire perception information, lateral trajectory state, and longitudinal trajectory state of the autonomous driving device, wherein the perception information includes posture information, speed information, and environmental information;

[0008] In response to the lateral trajectory state being a first state, determining a first lateral trajectory based on the position information, the first lateral trajectory being used to maintain the current heading angle extension;

[0009] In response to the longitudinal trajectory state being the second state, determining a first longitudinal trajectory based on the speed information, the first longitudinal trajectory being used to characterize a speed trajectory of smooth deceleration;

[0010] Based on the first lateral trajectory and the first longitudinal trajectory, current frame trajectory planning is performed for the automatic driving device.

[0011] In a second aspect, a trajectory planning device is provided, which is configured in an automatic driving device, and the device includes:

[0012] an information acquisition unit, configured to acquire perception information, a lateral trajectory state, and a longitudinal trajectory state of the autonomous driving device, wherein the perception information includes position information, speed information, and environmental information;

[0013] A first planning unit, configured to determine, in response to the lateral trajectory state being a first state, a first lateral trajectory based on the posture information, wherein the first lateral trajectory is configured to maintain the extension of a current heading angle;

[0014] a second planning unit, configured to determine, in response to the longitudinal trajectory state being a second state, a first longitudinal trajectory based on the speed information, the first longitudinal trajectory being used to characterize a speed trajectory of smooth deceleration;

[0015] The third planning unit is used to plan the current frame trajectory of the automatic driving device based on the first lateral trajectory and the first longitudinal trajectory.

[0016] In a third aspect, a trajectory planning device for an autonomous driving device is provided, comprising:

[0017] Memory, used to store programs;

[0018] A processor, configured to execute a program stored in the memory;

[0019] When the program stored in the memory is executed, the processor executes the trajectory planning method described in any one of the first aspects.

[0020] In a fourth aspect, an autonomous driving device is provided, comprising the trajectory planning device described in the second aspect.

[0021] In a fifth aspect, a computer-readable storage medium is provided, wherein the computer-readable medium stores instructions for execution by a computing device, and when the computing device executes the instructions, the method as described in any one of the first aspects is implemented.

[0022] One of the above technical solutions has the following advantages or beneficial effects:

[0023] Compared with the prior art, a trajectory planning method of the present application includes: obtaining the perception information, lateral trajectory state and longitudinal trajectory state of the automatic driving device, the perception information includes posture information, speed information and environmental information; in response to the lateral trajectory state being the first state, determining the first lateral trajectory based on the posture information, the first lateral trajectory is used to maintain the extension of the current heading angle; in response to the longitudinal trajectory state being the second state, determining the first longitudinal trajectory based on the speed information, the first longitudinal trajectory is used to characterize the speed trajectory of smooth deceleration; based on the first lateral trajectory and the first longitudinal trajectory, the current frame trajectory planning of the automatic driving device is performed. The trajectory planning method provided by the present application can generate a lateral trajectory according to the current heading angle when the lateral trajectory state of the current frame is the first state, and select a comfortable braking trajectory when the longitudinal trajectory state is the second state, so that the trajectory discontinuity that may be caused by the trajectory state of the current frame can be greatly improved, and the discomfort caused by strong braking of passengers can be improved, thereby improving the driving experience of passengers.

[0024] The trajectory planning device provided in the present application can generate a lateral trajectory according to the current heading angle direction when the current frame trajectory planning fails, and select a comfortable braking trajectory according to the fact that the target driving device does not pose a danger to the automatic driving device. Therefore, it can greatly improve the discontinuity caused by the failure of the current frame trajectory planning, improve the discomfort of passengers caused by strong braking, and enhance the driving experience of passengers.

[0025] The trajectory planning device of the autonomous driving equipment provided in the present application can generate a lateral trajectory according to the current heading angle direction when the current frame trajectory planning fails, and select a comfortable braking trajectory according to the fact that the target driving device does not pose a danger to the autonomous driving equipment. Therefore, it can greatly improve the discontinuity caused by the failure of the current frame trajectory planning, improve the discomfort of passengers caused by strong braking, and enhance the driving experience of passengers.

[0026] The autonomous driving device provided in the present application can start the backup trajectory planning when the current frame trajectory planning fails, thereby greatly improving the discontinuity caused by the failure of the current frame trajectory planning, as well as improving the discomfort of passengers caused by strong braking, thereby enhancing the passengers' driving experience.

[0027] The computer-readable storage medium provided in the present application can implement backup trajectory planning when the current frame trajectory planning fails, which can greatly improve the discontinuity caused by the failure of the current frame trajectory planning, as well as improve the discomfort caused by strong braking to passengers, thereby enhancing the passengers' driving experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0029] Figure 1 A schematic diagram of the system architecture of the autonomous driving device provided in the embodiment of the present application;

[0030] Figure 2 A schematic diagram of the overall principle framework of trajectory planning for an autonomous driving device provided in an embodiment of the present application;

[0031] Figure 3 A schematic diagram of the overall process of the trajectory planning method according to an embodiment of the present application;

[0032] Figure 4 This is a schematic diagram of the shape of the first lateral track in the embodiment of the present application;

[0033] Figure 5 This is a schematic diagram of the shape of the second lateral track in the embodiment of the present application;

[0034] Figure 6 This is a schematic diagram of an example flow chart of a trajectory planning method according to an embodiment of the present application;

[0035] Figure 7 A simplified flowchart of a trajectory planning method according to an embodiment of the present application;

[0036] Figure 8 A schematic diagram of the structure of a trajectory planning device according to an embodiment of the present application;

[0037] Fig. 9 A schematic diagram of the structure of a trajectory planning device for an autonomous driving device according to an embodiment of the present application.

[0038] Reference numerals:

[0039] 100-perception system; 200-planning system; 300-control system; 110-map device; 120-path planning device; 130-positioning device; 140-perception fusion device; 150-local area network communication device; 210-state machine; 220-lane change decision device; 230-lateral available space determination device; 240-longitudinal available space determination device; 250-lateral planning device; 260-longitudinal planning device; 270-alternate trajectory planning device; 271-information acquisition unit; 272-first planning unit; 273-second planning unit; 274-third planning unit; 310-lateral control device; 320-longitudinal control device; 901-memory; 902-processor. DETAILED DESCRIPTION

[0040] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0041] In the description of the present application, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, which are only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present application. In addition, the terms "first" and "second" are used only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more features. In the description of the present application, "multiple" means two or more, and at least one means one, two or more, unless otherwise clearly and specifically defined.

[0042] See also Figure 1 , Figure 1The system architecture of the autonomous driving device provided in the embodiment of the present application is illustrated. The system architecture of the device equipped with autonomous driving technology, such as a vehicle with L2+ or L3 autonomous driving technology, may generally include NUC PC1 (first compact computing device), NUC PC2 (second compact computing device), CAR PC (on-board computer), ASDM (Automated System Development and Management), Autobox3 (Autoboxing), and Host Vehicle CAN (host vehicle LAN communication). Among them, NUC PC1 receives the positioning information transmitted by RTK, converts it and transmits it to NUC PC2 for self-vehicle positioning; ASDM receives four short-range millimeter waves (SRR), one long-range millimeter wave (MRR) and a front camera (Front Camera) information and fuses them, and gives the fusion result to NUC PC2 for perception of the surrounding environment. At the same time, ASDM gives the information to DVTOOL (visualization tool) in CAR PC (on-board computer) for recording; NUC PC2 receives the positioning and fusion information and performs trajectory planning, and finally gives the planned trajectory to Autobox for real vehicle control.

[0043] See also Figure 2 , Figure 2 The overall principle framework of trajectory planning of the autonomous driving equipment provided in the embodiment of the present application is illustrated. Based on the software level, a vehicle equipped with autonomous driving technology may generally include a perception system 100, a planning system 200 and a control system 300. The perception system 100 may include a map device (HD-Map) 110, a path planning device (Routing) 120, a positioning device (Localization) 130, a perception fusion device (Perception) 140 and a local area network communication device (VehicleCAN) 150, wherein the map device 110 is provided with a high-precision map (High Definition Map) for providing map information, the path planning device 120 is used to provide navigation information, and the positioning device 130 (for example: can be used) Figure 1 The NUC PC1 in the embodiment is used to provide the positioning information of the vehicle, and the perception fusion device 140 (for example, it can be used Figure 1 The ASDM in the vehicle is used to provide the perception fusion information of the vehicle surrounding information, and the local area network communication device 150 (for example: can be used Figure 1 The Host Vehicle CAN in the host vehicle is used to provide vehicle data information, such as vehicle speed, vehicle acceleration, etc. The planning system 200 (for example, can be used Figure 1The NUC PC2 in the planning system 200 may include a state machine 210, a lane change decision device 220, a lateral available space determination device 230, a longitudinal available space determination device 240, a lateral planning device 250 and a longitudinal planning device 260, wherein the various devices in the planning system 200 work together, and finally the lateral planning device 250 generates a lateral trajectory of the current frame and sends the frame lateral trajectory to the control system 300 (for example, it can be adopted Figure 1 The lateral control device 310 in the Autobox 3) in the control system 300 is used, and the longitudinal planning device 260 generates a longitudinal trajectory of the current frame and sends the longitudinal trajectory of the frame to the longitudinal control device 320 in the control system 300 for use.

[0044] However, when special circumstances occur, for example, the vehicle in front brakes suddenly with great acceleration, the current frame trajectory planning may fail. At this time, the planning system 200 will send a forced braking instruction to the control system 300 to make the control system 300 directly execute forced braking. Therefore, passengers may feel uncomfortable, which will affect the driving experience.

[0045] In view of this, an embodiment of the present application provides an autonomous driving device, which adds a backup trajectory planning device 270 in the planning system 200, generates a lateral trajectory based on the current heading angle when the trajectory state of the current frame is abnormal, and selects a comfortable braking trajectory, thereby improving the passenger's driving experience when the trajectory state of the current frame is abnormal, thereby solving at least part of the above-mentioned technical problems.

[0046] Please continue reading Figure 2 The autonomous driving device of the embodiment of the present application also includes a fallback trajectory planning device (FallBack Trajectory Judge) 270, which is connected to the lateral planning device 250 and the longitudinal planning device 260 respectively. The fallback trajectory planning device 270 is used to adopt the trajectory planning method of the embodiment of the present application, and when the lateral planning device 250 and / or the longitudinal planning device 260 fail to successfully generate the trajectory of the current frame, generate the trajectory of the current frame to send the trajectory of the current frame to the lateral control device 310 and the longitudinal control device 320 in the control system 300 for use.

[0047] The trajectory planning method of an embodiment of the present application is introduced below with reference to the accompanying drawings.

[0048] See also Figure 3 , Figure 3The overall process of the trajectory planning method of the embodiment of the present application is illustrated. The trajectory planning method is applied to an autonomous driving device and specifically includes the following steps:

[0049] Step 301: Acquire perception information, lateral trajectory status, and longitudinal trajectory status of the autonomous driving device, where the perception information includes position information, speed information, and environmental information.

[0050] In some examples, the posture information may include the current heading angle. The speed information may include first speed information of the autonomous driving device and second speed information of the target driving device, wherein the target driving device may be used to characterize the vehicle located in front of the autonomous driving device along the driving direction. The environmental information may include current obstacles, and the current obstacles may include information such as surrounding stationary obstacles, roadside edges, lane lines, and rear moving obstacles.

[0051] Step 302: In response to the lateral trajectory state being a first state, determining a first lateral trajectory based on the posture information, where the first lateral trajectory is used to maintain the current heading angle extension.

[0052] In some embodiments, the first state is configured to satisfy a first lateral trajectory information constraint.

[0053] In some examples, the lateral trajectory in the first lateral trajectory information constraint includes a current frame lateral trajectory and a previous frame lateral trajectory. The first lateral trajectory information constraint is configured such that the current frame lateral trajectory planning fails and the previous frame lateral trajectory is unavailable. The previous frame is a frame before the current frame.

[0054] For example, one application scenario of the first state is a scenario where lane line information is not detected from the environmental information. Then, when the lateral trajectory planning of the current frame fails and the lateral trajectory of the previous frame is unavailable, when the lane line information is not detected from the environmental information, the method of step 302 is used to determine the first lateral trajectory that maintains the current heading angle.

[0055] Specifically, when the lateral planning device 250 fails to plan the lateral trajectory of the current frame, the storage module for temporarily storing the lateral trajectory of the current frame will be empty. At this time, a first label indicating the failure of lateral trajectory planning, such as Lat_Plan_Failed, may be added. After the backup trajectory planning device 270 detects the first label, it can be known that the lateral trajectory planning of the current frame has failed.

[0056] In some embodiments, before executing step 302, the trajectory planning method of the embodiment of the present application may further include the following steps:

[0057] In response to a failure in planning the lateral trajectory of the current frame, a lateral trajectory situation of a previous frame is determined based on a first preset condition.

[0058] In some examples, the first preset condition can be configured as the intersection of the current obstacle and the lateral trajectory of the previous frame, and / or the number of historical failure frames corresponding to the lateral trajectory of the previous frame is less than a preset failure threshold. The number of historical failure frames is used to represent the number of frames that have failed to plan continuously before the corresponding frame.

[0059] Specifically, obstacles may include surrounding stationary obstacles, roadside edges, lane lines, and rear moving obstacles, etc. Then, by determining whether the lateral trajectory of the previous frame has an intersection with these obstacles, it can be determined whether there is no collision risk between the lateral trajectory of the previous frame and the obstacle.

[0060] In addition, the types of trajectories planned for each frame (including horizontal and vertical trajectories) can generally include three types: successful trajectory planning (Norm), trajectory planning failed but the trajectory of the previous frame is available (LastCycle), and trajectory planning failed but the trajectory of the previous frame is not available (FallBack). The number of historical failed frames can be used to characterize the number of frames of continuous LastCycle type trajectories. Exemplarily, for the LastCycle type, assuming that it is the nth frame trajectory, since this frame uses the n-1th frame trajectory, if the n-1th frame trajectory is also the LastCycle type, and the n-2th frame trajectory is used, then the number of historical failed frames corresponding to the nth frame trajectory is 2 frames.

[0061] Based on the first preset condition of the above example, the horizontal trajectory of the previous frame can be determined by the following steps:

[0062] The first step is to obtain the current position of the obstacle perceived by the main vehicle.

[0063] Specifically, the positions of surrounding stationary obstacles, road edges, lane lines, and rear moving obstacles perceived by the main vehicle at this time can be obtained.

[0064] The second step is to obtain the number of historical failure frames corresponding to the horizontal trajectory of the previous frame.

[0065] The third step is to generate a usable result of the lateral trajectory of the previous frame based on the current obstacle position, when it is confirmed that there is no intersection between the lateral trajectory of the previous frame and the current obstacle position, and the number of historical failure frames is less than the preset failure threshold.

[0066] In the fourth step, based on the position of the current obstacle, when it is confirmed that the lateral trajectory of the previous frame intersects with the position of the current obstacle, and / or the number of historical failure frames is greater than or equal to a preset failure threshold, a result that the lateral trajectory of the previous frame is unavailable is generated.

[0067] In this way, whether the horizontal track of the previous frame is available can be detected in the above manner, and when the horizontal track of the previous frame is not available, the horizontal track state is determined to be the first state, and step 302 is started.

[0068] See also Figure 4 , Figure 4 The shape of the first lateral trajectory in the embodiment of the present application is illustrated. In some embodiments, when the lane line A of the current lane is not detected 1 and A 2 When the vehicle is in a state of being ... 1 This can be achieved in the following ways:

[0069] Assume that the current coordinates of the autonomous driving device (Host Vehicle) are (x, y, θ, kappa, dkappa), keep the current heading angle θ for fixed heading angle planning, and output the first lateral backup trajectory G 1 is a set of multiple lateral trajectory points, and the i-th lateral trajectory point is expressed as formula (1):

[0070] (x next+i ,y next+i ,θ,kappa,dkappa) (1)

[0071] In formula (1), x next+i ,y next+i is the coordinate of the i-th lateral trajectory point, θ is the current heading angle, kappa is the current curvature, dkappa is the rate of change of the current curvature, i is greater than or equal to zero and less than or equal to N p integer, N p is the total number of planned steps.

[0072] By the above method, when the lane line of the current lane in which the automatic driving device is traveling is not detected, a first lateral backup trajectory G is generated that maintains the current heading angle. 1 ,At this time, the automatic driving equipment can maintain the current heading angle planning, and the collision risk is guaranteed through subsequent longitudinal planning.

[0073] In some other embodiments, the first state is configured to satisfy the second horizontal trajectory information constraint. The second horizontal trajectory information constraint is configured such that the horizontal trajectory of the previous frame is the first horizontal trajectory.

[0074] That is to say, another application scenario of the first state is that the current frame horizontal trajectory planning fails, and the previous frame horizontal trajectory is already the first horizontal trajectory. Since the previous frame horizontal trajectory is already a spare trajectory, when the current frame horizontal trajectory planning fails, there is no need to determine whether the previous frame horizontal trajectory is available, but directly use the method of step 302 to determine the first horizontal trajectory of the current frame. The relevant specific steps can be referred to the above embodiment, which will not be repeated here.

[0075] Step 303: In response to the longitudinal trajectory state being the second state, determining a first longitudinal trajectory based on the speed information, where the first longitudinal trajectory is used to characterize a speed trajectory of smooth deceleration.

[0076] In some embodiments, the second state is configured to satisfy the first longitudinal trajectory information constraint, and the collision risk between the autonomous driving device and the target driving device satisfies a preset safety threshold.

[0077] In some examples, the longitudinal trajectory in the first longitudinal trajectory information constraint includes a current frame longitudinal trajectory and a previous frame longitudinal trajectory. The first longitudinal trajectory information constraint is configured such that the current frame longitudinal trajectory planning fails and the previous frame longitudinal trajectory is unavailable.

[0078] Exemplarily, one application scenario of the second state is that when the longitudinal trajectory planning of the current frame fails and the longitudinal trajectory of the previous frame is unavailable, when the collision risk between the autonomous driving device and the target driving device meets a preset safety threshold, the method of step 303 is used to determine the first longitudinal trajectory for smooth deceleration.

[0079] Specifically, when the longitudinal planning device 260 fails to plan the longitudinal trajectory of the current frame, the storage module for temporarily storing the longitudinal trajectory of the current frame will be empty. At this time, a second label indicating the failure of longitudinal trajectory planning, such as Lon_Plan_Failed, can be added. After the backup trajectory planning device 270 detects the second label, it can be known that the longitudinal trajectory planning of the current frame has failed.

[0080] In some embodiments, before executing step 303, the trajectory planning method of the embodiment of the present application may further include the following steps:

[0081] In response to failure in planning the longitudinal trajectory of the current frame, the longitudinal trajectory of the previous frame is determined based on the second preset condition.

[0082] In some examples, the second preset condition can be configured as the previous frame longitudinal trajectory includes the current moment, and the collision risk between the target trajectory point and the target driving device meets a preset safety threshold. The target trajectory point is the trajectory point in the previous frame longitudinal trajectory corresponding to the current moment.

[0083] Based on the second preset condition in the above example, the vertical trajectory of the previous frame can be determined by the following steps:

[0084] The first step is to obtain the first absolute time corresponding to the end moment from the vertical trajectory of the previous frame.

[0085] The second step is to obtain the second absolute time corresponding to the start time of the current frame.

[0086] It can be understood that if the second absolute time is less than the first absolute time, it indicates that the previous frame of the longitudinal trajectory has not yet ended, and the previous frame of the longitudinal trajectory includes the current time.

[0087] The third step is to obtain the third speed information of the target trajectory point close to the current moment from the longitudinal trajectory of the previous frame.

[0088] It is understandable that the current moment is the planned start moment of the current frame. The target trajectory point can be the trajectory point in the longitudinal trajectory of the previous frame that is closest to the current moment. The embodiment of the present application uses the target trajectory point to replace the automatic driving device for calculation.

[0089] Specifically, the third speed information may include the third speed v of the target trajectory point. ego , the third acceleration, and the third displacement.

[0090] The fourth step is to obtain the second speed information of the target driving device.

[0091] Specifically, the second speed information may include the second speed v of the target driving device. front , the second acceleration a front and the second displacement.

[0092] The fifth step is to generate a result in which the longitudinal trajectory of the previous frame is available when the second absolute time is less than the first absolute time and it is confirmed based on the third speed information and the second speed information that there is no collision risk between the target trajectory point and the target driving device.

[0093] Step 6. When the second absolute time is greater than or equal to the first absolute time, and / or when it is confirmed based on the third speed information and the second speed information that the target trajectory point has a risk of collision with the target driving device, a result that the previous frame longitudinal trajectory is unavailable is generated.

[0094] In some examples, the following steps may be used to determine whether there is a collision risk between the target trajectory point and the target driving device based on the third speed information and the second speed information:

[0095] First, the relative speed v between the target trajectory point and the target driving device is calculated by the following formula (2): relative :

[0096] vrelative =fmax(v ego -v front ,0.0) (2)

[0097] In formula (2), fmax means taking the larger of the two values ​​in brackets, v ego is the third velocity of the target trajectory point at the current moment, v front The second speed of the target driving device at the current moment.

[0098] Then, the relative distance d between the target trajectory point and the target driving device is determined based on the third displacement and the second displacement. relative Then, the relative distance d between the target trajectory point and the target driving device for subsequent calculation is determined by the following formula (3): relative :

[0099] d relative =fmax(d relative -d stop ,0.0) (3)

[0100] In formula (3), fmax means the larger of the two values ​​in brackets, and d stop The preset parking safety distance.

[0101] Next, the minimum acceleration of the vehicle without collision is determined by formula (4): avoidcollision :

[0102]

[0103] In formula (4), v relative is the relative speed between the target trajectory point and the target driving device, d relative is the relative distance between the target trajectory point and the target driving device, a front is the second acceleration of the target driving device.

[0104] Finally, if the third acceleration of the target trajectory point is less than the minimum acceleration of the vehicle without collision, accel avoidcollision , then it indicates that the target trajectory point has a very small collision risk at the current moment, that is, there is no collision risk between the target trajectory point and the target driving device. Otherwise, there is a collision risk between the target trajectory point and the target driving device.

[0105] In this way, whether the longitudinal track of the previous frame is available can be detected in the above manner, and when the longitudinal track of the previous frame is not available, it is determined that the longitudinal track state is the second state, and step 303 is started.

[0106] In some embodiments, the following steps may be used to confirm that the collision risk between the autonomous driving device and the target driving device meets a preset safety threshold:

[0107] Step 1: Obtain second speed information of the target driving device.

[0108] Specifically, the second speed information may include a second speed, a second acceleration, and a second displacement of the target driving device.

[0109] Step 2: Obtain the first speed information of the automatic driving device at the current moment.

[0110] Specifically, the first speed information may include the current first speed, first acceleration, and first displacement of the automatic driving device.

[0111] Step 3: Based on the first speed information and the second speed information, obtain the encounter time TTC between the automatic driving device and the target driving device. The encounter time TTC is used to characterize the collision risk.

[0112] Specifically, based on the first speed and the second speed, the relative speed between the autonomous driving device and the target driving device can be obtained, and then the encounter time TTC when the autonomous driving device collides with the target driving device can be obtained by dividing the relative distance between the autonomous driving device and the target driving device by the relative speed.

[0113] Step 4: When the encounter time TTC is greater than the preset time threshold T, it is confirmed that the collision risk between the autonomous driving device and the target driving device meets the preset safety threshold, that is, there is no collision risk. When the encounter time TTC is less than or equal to the preset time threshold T, it is confirmed that the collision risk between the autonomous driving device and the target driving device exceeds the preset safety threshold, that is, there is a collision risk.

[0114] In addition, it can also be understood that if there is no target driving device in the driving direction of the autonomous driving device, it can also be determined that there is no risk of collision between the autonomous driving device and the target driving device.

[0115] In some embodiments, when it is confirmed that there is no risk of collision between the autonomous driving device and the target driving device, step 303 may specifically generate a first longitudinal trajectory for representing a smooth deceleration based on the speed information. This may be achieved in the following manner:

[0116] First, construct the objective function as shown in formula (5):

[0117]

[0118] In formula (5), p1, p2, p3, p4, and p5 are coefficients, s is displacement, v is velocity, a is acceleration, jerk is jerk, and v set To set the speed.

[0119] Then, perform quadratic programming. Its quadratic form is shown in formula (6):

[0120]

[0121]

[0122] In formula (6), P is x 2 The coefficient matrix, q T is the coefficient matrix of x, s i 、v i 、a i , jerk i Respectively represent the displacement, velocity, acceleration, and jerk at the current moment, s i-1 、v i-1 、a i-1 Respectively represent the displacement, velocity, and acceleration at the previous moment, t cycle is the planning time for each frame, s min 、v min 、a min Respectively represent the set minimum displacement, minimum velocity, and minimum acceleration, s max 、v max 、a max Respectively represent the maximum displacement, maximum velocity, and maximum acceleration, s start 、v start 、a start They represent initial displacement, initial velocity, and initial acceleration respectively.

[0123] Finally, after solving, the first longitudinal trajectory can be obtained.

[0124] By generating the first longitudinal trajectory in the above manner, braking can be performed faster on the basis of comfort. While taking into account comfort, it can also ensure that the autonomous driving equipment can stop in the shortest possible time to better take into account safety. At the same time, there will not be too much abruptness when the successfully planned trajectory is subsequently adopted, which can greatly improve the driving experience.

[0125] In some other embodiments, the second state is configured to satisfy a second longitudinal track information constraint. The second longitudinal track information constraint is configured such that the longitudinal track of the previous frame is the first longitudinal track.

[0126] That is to say, another application scenario of the second state is that the longitudinal trajectory planning of the current frame fails, and the longitudinal trajectory of the previous frame is already the first longitudinal trajectory. Since the longitudinal trajectory of the previous frame is already the backup trajectory, when the longitudinal trajectory planning of the current frame fails, there is no need to determine whether the longitudinal trajectory of the previous frame is available, but directly use the method of step 303 to determine the first longitudinal trajectory of the current frame. The relevant specific steps can be referred to the above embodiment, which will not be repeated here.

[0127] Step 304: Based on the first lateral trajectory and the first longitudinal trajectory, the current frame trajectory planning is performed for the autonomous driving device.

[0128] Specifically, after the first lateral trajectory and the first longitudinal trajectory are combined into a trajectory of the current frame, the trajectory is sent to the control system 300 to control the vehicle to execute the trajectory.

[0129] It can be understood that the trajectory planning method of the embodiment of the present application can generate a lateral trajectory based on the current heading angle when the current frame trajectory planning fails, and select a comfortable braking trajectory based on the fact that the target driving device does not pose a danger to the automatic driving device. Therefore, it can greatly improve the discontinuity caused by the failure of the current frame trajectory planning, improve the discomfort of passengers caused by strong braking, and enhance the passengers' driving experience.

[0130] In addition, in some embodiments, the trajectory planning method of the embodiment of the present application may further include the following steps in addition to the above steps 301 to 304:

[0131] Step 1: in response to the longitudinal trajectory state being the third state, determining a second longitudinal trajectory based on the speed information, where the second longitudinal trajectory is used to characterize a speed trajectory of rapid deceleration.

[0132] In some examples, the third state is configured to satisfy the first longitudinal trajectory information constraint, and the collision risk between the autonomous driving device and the target driving device exceeds a preset safety threshold.

[0133] That is to say, the application scenario of the third state is that when the longitudinal trajectory planning of the current frame fails and the longitudinal trajectory of the previous frame is unavailable, when the collision risk between the autonomous driving device and the target driving device exceeds the preset safety threshold, a second longitudinal trajectory of rapid deceleration is determined based on the speed information.

[0134] In some embodiments, when it is confirmed that the collision risk between the autonomous driving device and the target driving device exceeds a preset safety threshold, that is, when it is confirmed that there is a collision risk between the autonomous driving device and the target driving device, step one may specifically generate a second longitudinal trajectory for characterizing rapid deceleration in the following manner:

[0135] The benchmark formula for constructing motion planning is shown in formula (7):

[0136] x t =a 0 +a 1 ×t+a 2 ×t 2 +a 3 ×t 3 +a 4 ×t 4+a 5 ×t 5

[0137] v t =a 1 +2×a 2 *t+3×a 3 ×t 2 +4*a 4 ×t 3 +5*a 5 ×t 4

[0138] accel t =2×a 2 +6×a 3 ×t+12×a 4 ×t 2 +20×a 5 ×t 3

[0139] jerk t =6×a 3 +24×a 4 ×t+60×a 5 ×t 2 (7)

[0140] In formula (7), x t 、v t 、accel t , jerk t are displacement, velocity, acceleration, and jerk, respectively, and a 0 ~a 5 are the coefficients of the quintic polynomial, and t is the time.

[0141] Since the second longitudinal track is a strong braking method, it is necessary to use the minimum jerk t To decelerate, i.e. jerk t =jerk min , while the jerk t It should be independent of time t and the maximum negative jerk should be maintained at all times. 5 =0,a 4 =0,a 3 =jerk min / 6, a 2 =accel init / 2, a 1 =velo init , a 0 =dist init , that is, at this time, the maximum negative jerk will be used to decelerate until the maximum deceleration is reached. init、velo init 、dist init are the acceleration, velocity and displacement of the initial state.

[0142] By generating the second longitudinal trajectory in the above manner, safety can be improved.

[0143] Step 2: Based on the first lateral trajectory and the second longitudinal trajectory, the current frame trajectory is planned for the autonomous driving device.

[0144] In addition, in some embodiments, the trajectory planning method of the embodiment of the present application may also include the following steps:

[0145] Step three, in response to the lateral trajectory state being the fourth state, determining a second lateral trajectory based on the posture information and the lane line information, where the second lateral trajectory is used to extend along the current lane.

[0146] In some examples, the fourth state is configured to satisfy the first lateral trajectory information constraint and detect lane line information from the environmental information.

[0147] Exemplarily, the application scenario of the fourth state is a scenario where lane line information is detected from environmental information. Then, when the lateral trajectory planning of the current frame fails and the lateral trajectory of the previous frame is unavailable, when lane line information is detected from the environmental information, the method of step three is used to determine a second lateral trajectory extending along the current lane.

[0148] See also Figure 5 , Figure 5 The shape of the second lateral trajectory in the embodiment of the present application is illustrated. In some embodiments, when the lane line A of the current lane in which the autonomous driving device (Host Vehicle) is traveling is detected, 1 and A 2 When step 3 is performed, the second lateral trajectory G extending along the current lane can be generated by the following method: 2 :

[0149] Assume that the center line of the current lane is represented by the equation: y = a 0 +a 1 x+a 2 x 2 +a 3 x 3 . Assume that the current planning starting position is l 0 , the position accuracy is Δl, the second lateral trajectory G 2 is a set of multiple lateral trajectory points, and the i-th lateral trajectory point is expressed as formula (8):

[0150]

[0151] In formula (8), i is greater than or equal to zero and less than or equal to N p integer, N p is the total number of planning steps, l 0 is the current planning starting position, Δl is the position accuracy, is the current lane centerline position corresponding to the i-th lateral trajectory point, is the current heading angle, is the current curvature, is the rate of change of the current curvature.

[0152] Specifically, each parameter in formula (8) is determined by the following formula (9):

[0153]

[0154] In formula (9), θ is the initial current heading angle, a 1 、a 2 、a 3 is the coefficient in the lane centerline equation of the current lane, l is the planned position of the i-th lateral trajectory point, theta l is the current heading angle, kappa l is the current curvature, dy represents the distance between the i-th lateral trajectory point and the centerline of the lane, ddy represents the derivative of dy, and dddy represents the second-order derivative of dy.

[0155] In the above manner, when the lane line of the current lane in which the autonomous driving device is traveling is detected, a second lateral trajectory G extending along the current lane is generated. 2 At this time, the autonomous driving equipment can move in the direction of the lane line and maintain the lateral distance from the lane line unchanged, which will not cause danger to surrounding vehicles.

[0156] Step 4: Based on the second lateral trajectory and the first longitudinal trajectory, the current frame trajectory planning is performed for the autonomous driving device.

[0157] In addition, in some embodiments, the trajectory planning method of the embodiment of the present application may also include the following steps:

[0158] Step five: in response to the longitudinal trajectory state being the third state, determining a second longitudinal trajectory based on the speed information, where the second longitudinal trajectory is used to characterize a speed trajectory of rapid deceleration.

[0159] Step six: Based on the second lateral trajectory and the second longitudinal trajectory, the current frame trajectory planning is performed for the autonomous driving device.

[0160] In addition, in some embodiments, the trajectory planning method of the embodiment of the present application may also include the following steps:

[0161] Step seven: in response to the lateral trajectory state being the fifth state, the fifth state is configured to perform trajectory planning based on the lateral trajectory of the previous frame.

[0162] Exemplarily, the application scenario of the fifth state is that the horizontal trajectory planning of the current frame fails, and the horizontal trajectory of the previous frame is available.

[0163] Step eight, determining a third lateral trajectory based on the lateral trajectory of the previous frame.

[0164] Step nine: performing trajectory planning for the autonomous driving device based on the third lateral trajectory and the first longitudinal trajectory.

[0165] In addition, in some embodiments, the trajectory planning method of the embodiment of the present application may also include the following steps:

[0166] Step ten: in response to the longitudinal trajectory state being the sixth state, the sixth state is configured to perform trajectory planning based on the longitudinal trajectory of the previous frame.

[0167] Exemplarily, the application scenario of the sixth state is that the longitudinal trajectory planning of the current frame fails, and the longitudinal trajectory of the previous frame is available.

[0168] Step eleven: determining the third longitudinal trajectory based on the longitudinal trajectory of the previous frame.

[0169] Step 12: Based on the first lateral trajectory and the third longitudinal trajectory, the current frame trajectory planning is performed for the automatic driving device.

[0170] It can be understood that, through the above method, if the lateral planning is successful but the longitudinal planning fails, the lateral planning is used to obtain the trajectory, and the longitudinal direction first determines whether the trajectory of the previous moment is available. If available, the longitudinal trajectory of the previous moment is output. If not available, the comfortable braking trajectory and the emergency braking trajectory are selected according to whether the target driving device exists or whether the target driving device is dangerous to the automatic driving device. If the lateral and longitudinal planning fail at the same time, the lateral and longitudinal directions respectively determine whether the trajectory of the previous moment is available. If available, the trajectory of the previous moment is used. If not available, the lateral trajectory is generated according to the lane line direction / the vehicle heading angle direction, and the longitudinal direction is selected according to whether the target driving device exists or whether the target driving device is dangerous to the automatic driving device. Therefore, a trajectory line along the lane line / the current vehicle heading angle is finally generated, and its speed direction is slow deceleration. When a frame is successfully planned, the successfully planned trajectory can be used without too much abruptness, and the driving experience of passengers can be improved.

[0171] See also Figure 6 , Figure 6 The example process of the trajectory planning method of the embodiment of the present application is illustrated. In some examples, the trajectory planning method of the embodiment of the present application may specifically include the following steps:

[0172] Step 601: In response to the failure of the current frame horizontal trajectory planning, check whether the previous frame horizontal trajectory is available. If the previous frame horizontal trajectory is not available, execute step 602; if the previous frame horizontal trajectory is available, execute step 615.

[0173] Step 602: Check whether the lane line of the current lane can be detected. When the lane line of the current lane is not detected, step 603 is executed, and when the lane line of the current lane in which the autonomous driving device is traveling is detected, step 608 is executed.

[0174] Step 603: Based on the current heading angle of the automatic driving device, generate a first lateral trajectory extending while maintaining the current heading angle.

[0175] Step 604: In response to the failure of planning the longitudinal trajectory of the current frame, checking whether the longitudinal trajectory of the previous frame is available. If the longitudinal trajectory of the previous frame is not available, executing step 605.

[0176] Step 605: Confirm whether there is a collision risk between the autonomous driving device and the target driving device. When it is confirmed that there is no collision risk between the autonomous driving device and the target driving device, step 606 is executed. When it is confirmed that there is a collision risk between the autonomous driving device and the target driving device, step 613 is executed.

[0177] Step 606: Generate a first longitudinal trajectory, where the first longitudinal trajectory is used to characterize a velocity trajectory of smooth deceleration.

[0178] Step 607: Generate a trajectory of the current frame based on the first horizontal trajectory and the first vertical trajectory.

[0179] Step 608: Based on the lane centerline of the current lane and the current heading angle of the autonomous driving device, generate a second lateral trajectory extending along the current lane.

[0180] Step 609: In response to the failure of planning the longitudinal trajectory of the current frame, check whether the longitudinal trajectory of the previous frame is available. If the longitudinal trajectory of the previous frame is not available, execute step 610; if the longitudinal trajectory of the previous frame is available, execute step 620.

[0181] Step 610: Determine whether there is a risk of collision between the autonomous driving device and the target driving device. When it is determined that there is a risk of collision between the autonomous driving device and the target driving device, step 611 is executed.

[0182] Step 611: Generate a second longitudinal trajectory, where the second longitudinal trajectory is used to characterize a velocity trajectory of rapid deceleration.

[0183] Step 612: Generate a trajectory of the current frame based on the second horizontal trajectory and the second vertical trajectory.

[0184] Step 613: Generate a second longitudinal trajectory, where the second longitudinal trajectory is used to characterize a velocity trajectory of rapid deceleration.

[0185] Step 614: Generate a trajectory of the current frame based on the first horizontal trajectory and the second vertical trajectory.

[0186] Step 615: Determine the horizontal trajectory of the previous frame as the horizontal trajectory of the current frame.

[0187] Step 616: In response to the failure of planning the longitudinal trajectory of the current frame, check whether the longitudinal trajectory of the previous frame is available. If the longitudinal trajectory of the previous frame is not available, execute step 617.

[0188] Step 617: Confirm whether there is a risk of collision between the autonomous driving device and the target driving device. When it is confirmed that there is no risk of collision between the autonomous driving device and the target driving device, step 618 is executed.

[0189] Step 618: Generate a first longitudinal trajectory, where the first longitudinal trajectory is used to represent a velocity trajectory of smooth deceleration.

[0190] Step 619: Generate the trajectory of the current frame based on the horizontal trajectory of the current frame and the first vertical spare trajectory.

[0191] Step 620: Determine the longitudinal trajectory of the previous frame as the longitudinal trajectory of the current frame.

[0192] Step 621: Generate a trajectory of the current frame based on the second horizontal trajectory and the vertical trajectory of the current frame.

[0193] It can be understood that in other embodiments, if the horizontal trajectory planning of the current frame is successful, or the vertical trajectory planning of the current frame is successful, the successfully planned trajectory is determined as the horizontal trajectory or vertical trajectory of the current frame. For the horizontal trajectory or vertical trajectory that is not successfully planned, the method of the embodiment of the present application can be used to perform corresponding backup trajectory planning, which will not be described one by one here.

[0194] In addition, in order to more clearly reflect the method of the embodiment of the present application, please refer to Figure 7 , Figure 7 The simplified process of the trajectory planning method of the embodiment of the present application is illustrated. First, determine whether the trajectory planning of the current frame fails. If it fails, determine whether the trajectory of the previous frame can be used. If it cannot be used, the lateral trajectory can be calculated according to the shape of the lane line, or according to the heading angle of the vehicle. The longitudinal trajectory can use the comfortable braking trajectory or the emergency braking trajectory. The lateral trajectory and the longitudinal trajectory are combined to output the final trajectory.

[0195] It can be understood that, through the above method, if the lateral planning is successful but the longitudinal planning fails, the lateral planning is used to obtain the trajectory, and the longitudinal direction first determines whether the trajectory of the previous moment is available. If available, the longitudinal trajectory of the previous moment is output. If not available, the comfortable braking trajectory and the emergency braking trajectory are selected according to whether the target driving device exists or whether the target driving device is dangerous to the automatic driving device. If the lateral and longitudinal planning fail at the same time, the lateral and longitudinal directions respectively determine whether the trajectory of the previous moment is available. If available, the trajectory of the previous moment is used. If not available, the lateral trajectory is generated according to the lane line direction / the vehicle heading angle direction, and the longitudinal direction is selected according to whether the target driving device exists or whether the target driving device is dangerous to the automatic driving device. Therefore, a trajectory line along the lane line / the current vehicle heading angle is finally generated, and its speed direction is slow deceleration. When a frame is successfully planned, the successfully planned trajectory can be used without too much abruptness, and the driving experience of passengers can be improved.

[0196] Accordingly, see Figure 8 , Figure 8 The backup trajectory planning device 270 provided in the embodiment of the present application is configured in the automatic driving device, and the backup trajectory planning device 270 includes an information acquisition unit 271 , a first planning unit 272 , a second planning unit 273 and a third planning unit 274 .

[0197] The information acquisition unit 271 is used to acquire the perception information, lateral trajectory state and longitudinal trajectory state of the automatic driving device. The perception information includes posture information, speed information and environmental information.

[0198] The first planning unit 272 is configured to determine a first lateral trajectory based on the posture information in response to the lateral trajectory state being the first state, where the first lateral trajectory is configured to maintain the current heading angle extension.

[0199] The second planning unit 273 is configured to determine a first longitudinal trajectory based on the speed information in response to the longitudinal trajectory state being the second state, where the first longitudinal trajectory is used to characterize a speed trajectory of smooth deceleration.

[0200] The third planning unit 274 is used to plan the current frame trajectory of the automatic driving device based on the first lateral trajectory and the first longitudinal trajectory.

[0201] In some embodiments, the first state is configured to satisfy a first lateral trajectory information constraint.

[0202] The second state is configured to satisfy the first longitudinal trajectory information constraint, and the collision risk between the autonomous driving device and the target driving device satisfies a preset safety threshold.

[0203] In some embodiments, the horizontal trajectory in the first horizontal trajectory information constraint includes a current frame horizontal trajectory and a previous frame horizontal trajectory.

[0204] The longitudinal trajectory in the first longitudinal trajectory information constraint includes the longitudinal trajectory of the current frame and the longitudinal trajectory of the previous frame.

[0205] In some embodiments, the first planning unit 272 is further configured to:

[0206] The lateral trajectory of the previous frame is determined based on the first preset condition. The environmental information includes the current obstacle. The first preset condition is configured as:

[0207] The current obstacle intersects with the lateral trajectory of the previous frame.

[0208] And / or, the number of historical failed frames corresponding to the previous frame of the lateral trajectory is less than a preset failure threshold, and the number of historical failed frames is used to represent the number of frames that have failed to be planned continuously before the corresponding frame.

[0209] In some embodiments, the second planning unit 273 is further configured to:

[0210] The longitudinal track condition of the previous frame is determined based on the second preset condition.

[0211] The second preset condition is configured as that the previous frame longitudinal trajectory includes the current moment, and the collision risk between the target trajectory point and the target driving device meets the preset safety threshold, and the target trajectory point is the trajectory point in the previous frame longitudinal trajectory corresponding to the current moment.

[0212] In some embodiments, the first state is configured to satisfy a second horizontal trajectory information constraint, and the second horizontal trajectory information constraint is configured such that the horizontal trajectory of the previous frame is the first horizontal trajectory.

[0213] The second state is configured to satisfy a second longitudinal trajectory information constraint, and the second longitudinal trajectory information constraint is configured that the longitudinal trajectory of the previous frame is the first longitudinal trajectory.

[0214] In some embodiments, the speed information includes first speed information of the autonomous driving device and second speed information of the target driving device.

[0215] The second planning unit 273 is further specifically used for:

[0216] Based on the first speed information and the second speed information, an encounter time between the automatic driving device and the target driving device is obtained, and the encounter time is used to characterize the collision risk.

[0217] When the encounter time is greater than the preset time threshold, it is confirmed that the collision risk between the autonomous driving device and the target driving device meets the preset safety threshold.

[0218] In some embodiments, the second planning unit 273 is further configured to determine a second longitudinal trajectory based on the speed information in response to the longitudinal trajectory state being the third state, where the second longitudinal trajectory is used to characterize a speed trajectory of rapid deceleration.

[0219] The third planning unit 274 is further configured to plan a current frame trajectory for the automatic driving device based on the first lateral trajectory and the second longitudinal trajectory.

[0220] In some embodiments, the third state is configured to satisfy the first longitudinal trajectory information constraint, and the collision risk between the autonomous driving device and the target driving device exceeds a preset safety threshold.

[0221] In some embodiments, the first planning unit 272 is further configured to determine, in response to the lateral trajectory state being the fourth state, a second lateral trajectory based on the posture information and the lane line information, where the second lateral trajectory is configured to extend along the current lane.

[0222] The third planning unit 274 is further configured to plan a current frame trajectory for the automatic driving device based on the second lateral trajectory and the first longitudinal trajectory.

[0223] In some embodiments, the fourth state is configured to satisfy the first lateral trajectory information constraint, and lane line information is detected from the environment information.

[0224] In some embodiments, the second planning unit 273 is further configured to determine a second longitudinal trajectory based on the speed information in response to the longitudinal trajectory state being the third state, where the second longitudinal trajectory is used to characterize a speed trajectory of rapid deceleration.

[0225] The third planning unit 274 is further configured to plan the current frame trajectory of the automatic driving device based on the second lateral trajectory and the second longitudinal trajectory.

[0226] In some embodiments, the first planning unit 272 is further configured to respond to the lateral trajectory state being the fifth state, the fifth state being configured to perform trajectory planning based on the lateral trajectory of the previous frame, and determine the third lateral trajectory based on the lateral trajectory of the previous frame.

[0227] The third planning unit 274 is further configured to perform trajectory planning for the automatic driving device based on the third lateral trajectory and the first longitudinal trajectory.

[0228] In some embodiments, the second planning unit 273 is further configured to respond to the longitudinal trajectory state being the sixth state, the sixth state being configured to perform trajectory planning based on the longitudinal trajectory of the previous frame, and determine the third longitudinal trajectory based on the longitudinal trajectory of the previous frame.

[0229] The third planning unit 274 is further specifically configured to perform current frame trajectory planning for the automatic driving device based on the first lateral trajectory and the third longitudinal trajectory.

[0230] It can be understood that the trajectory planning device of the embodiment of the present application can generate a lateral trajectory based on the current heading angle direction when the current frame trajectory planning fails, and select a comfortable braking trajectory based on the fact that the target driving device does not pose a danger to the automatic driving device. Therefore, it can greatly improve the discontinuity caused by the failure of the current frame trajectory planning, improve the discomfort caused by strong braking to the passengers, and enhance the passengers' driving experience.

[0231] Accordingly, see Fig. 9 , Fig. 9 The structure of the trajectory planning device of the automatic driving device of the embodiment of the present application is illustrated. The embodiment of the present application also provides a trajectory planning device of the automatic driving device, including a memory 901 and a processor 902. The memory 901 is used to store programs. The processor 902 is used to execute the program stored in the memory 901. When the program stored in the memory 901 is executed, the processor 902 executes the trajectory planning method of the aforementioned embodiment of the present application.

[0232] Correspondingly, the autonomous driving device provided in the embodiment of the present application can generate a lateral trajectory based on the current heading angle direction when the current frame trajectory planning fails, and select a comfortable braking trajectory based on the fact that the target driving device does not pose a danger to the autonomous driving device. Therefore, it can greatly improve the discontinuity caused by the failure of the current frame trajectory planning, improve the discomfort of passengers caused by strong braking, and enhance the passengers' driving experience.

[0233] Correspondingly, an embodiment of the present application also provides a computer-readable storage medium, which stores instructions for execution by a computing device. When the computing device executes the instructions, the trajectory planning method as described in the aforementioned embodiment of the present application is implemented.

[0234] The above is a detailed introduction to a trajectory planning method, device, autonomous driving equipment and readable storage medium provided in the embodiments of the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the technical solution and its core idea of ​​the present application. Ordinary technicians in this field should understand that they can still modify the technical solutions recorded in the aforementioned embodiments, or replace some of the technical features therein with equivalents; and these modifications or replacements do not make the essence of the corresponding technical solution deviate from the scope of the technical solution of the embodiments of the present application.

Claims

1. A trajectory planning method, It is characterized in that Applied to an autonomous driving device, the method comprises: Acquire perception information, lateral trajectory state, and longitudinal trajectory state of the autonomous driving device, wherein the perception information includes posture information, speed information, and environmental information; In response to the lateral trajectory state being a first state, determining a first lateral trajectory based on the position information, the first lateral trajectory being used to maintain the current heading angle extension; In response to the longitudinal trajectory state being the second state, determining a first longitudinal trajectory based on the speed information, the first longitudinal trajectory being used to characterize a speed trajectory of smooth deceleration; Based on the first lateral trajectory and the first longitudinal trajectory, current frame trajectory planning is performed for the automatic driving device.

2. A trajectory planning method according to claim 1, It is characterized in that The first state is configured to satisfy a first lateral trajectory information constraint; The second state is configured to satisfy the first longitudinal trajectory information constraint, and the collision risk between the autonomous driving device and the target driving device satisfies a preset safety threshold.

3. A trajectory planning method according to claim 2, It is characterized in that The lateral trajectory in the first lateral trajectory information constraint includes the lateral trajectory of the current frame and the lateral trajectory of the previous frame; The longitudinal trajectory in the first longitudinal trajectory information constraint includes a longitudinal trajectory of a current frame and a longitudinal trajectory of a previous frame.

4. A trajectory planning method according to claim 3, It is characterized in that The method further comprises: The lateral trajectory of the previous frame is determined based on a first preset condition; the environmental information includes the current obstacle; the first preset condition is configured as: The current obstacle and the lateral trajectory of the previous frame have an intersection; And / or, the number of historical failed frames corresponding to the previous frame of the lateral trajectory is less than a preset failure threshold, and the number of historical failed frames is used to represent the number of frames that have failed to be planned continuously before the corresponding frame.

5. A trajectory planning method according to claim 3, It is characterized in that The method further comprises: Determining the longitudinal trajectory of the previous frame based on a second preset condition; The second preset condition is configured as that the previous frame longitudinal trajectory includes the current moment, and the collision risk between the target trajectory point and the target driving device meets the preset safety threshold, and the target trajectory point is the trajectory point in the previous frame longitudinal trajectory corresponding to the current moment.

6. A trajectory planning method according to claim 1, It is characterized in that The first state is configured to satisfy a second lateral trajectory information constraint, and the second lateral trajectory information constraint is configured that the lateral trajectory of the previous frame is the first lateral trajectory; The second state is configured to satisfy a second longitudinal trajectory information constraint, and the second longitudinal trajectory information constraint is configured such that the longitudinal trajectory of a previous frame is the first longitudinal trajectory.

7. A trajectory planning method according to claim 2, It is characterized in that The speed information includes first speed information of the automatic driving device and second speed information of the target driving device; The step of confirming that the collision risk between the autonomous driving device and the target driving device meets a preset safety threshold comprises: Based on the first speed information and the second speed information, acquiring an encounter time between the automatic driving device and the target driving device, wherein the encounter time is used to characterize the collision risk; When the encounter time is greater than a preset time threshold, it is confirmed that the collision risk between the automatic driving device and the target driving device meets a preset safety threshold.

8. A trajectory planning method according to claim 1, It is characterized in that The method further comprises: In response to the longitudinal trajectory state being a third state, determining a second longitudinal trajectory based on the speed information, the second longitudinal trajectory being used to characterize a speed trajectory of rapid deceleration; Based on the first lateral trajectory and the second longitudinal trajectory, current frame trajectory planning is performed on the automatic driving device.

9. A trajectory planning method according to claim 8, It is characterized in that The third state is configured to satisfy the first longitudinal trajectory information constraint, and the collision risk between the autonomous driving device and the target driving device exceeds a preset safety threshold.

10. A trajectory planning method according to claim 1, It is characterized in that The method further comprises: In response to the lateral trajectory state being a fourth state, determining a second lateral trajectory based on the posture information and the lane line information, the second lateral trajectory being used to extend along the current lane; Based on the second lateral trajectory and the first longitudinal trajectory, current frame trajectory planning is performed on the automatic driving device.

11. A trajectory planning method according to claim 10, It is characterized in that The fourth state is configured to satisfy the first lateral trajectory information constraint, and lane line information is detected from the environment information.

12. A trajectory planning method according to claim 11, It is characterized in that The method further comprises: In response to the longitudinal trajectory state being a third state, determining a second longitudinal trajectory based on the speed information, the second longitudinal trajectory being used to characterize a speed trajectory of rapid deceleration; Based on the second lateral trajectory and the second longitudinal trajectory, current frame trajectory planning is performed on the automatic driving device.

13. A trajectory planning method according to claim 1, It is characterized in that The method further comprises: In response to the lateral trajectory state being a fifth state, the fifth state is configured to perform the trajectory planning based on a previous frame of lateral trajectory; Determining a third lateral trajectory based on the lateral trajectory of the previous frame; Based on the third lateral trajectory and the first longitudinal trajectory, trajectory planning is performed for the automatic driving device.

14. A trajectory planning method according to claim 1, It is characterized in that The method further comprises: In response to the longitudinal trajectory state being a sixth state, the sixth state is configured to perform the trajectory planning based on a previous frame of longitudinal trajectory; Determining the third longitudinal trajectory based on the longitudinal trajectory of the previous frame; Based on the first lateral trajectory and the third longitudinal trajectory, current frame trajectory planning is performed for the automatic driving device.

15. A trajectory planning device, It is characterized in that Configured in an automatic driving device, the device includes: an information acquisition unit, configured to acquire perception information, a lateral trajectory state, and a longitudinal trajectory state of the autonomous driving device, wherein the perception information includes position information, speed information, and environmental information; A first planning unit, configured to determine, in response to the lateral trajectory state being a first state, a first lateral trajectory based on the posture information, wherein the first lateral trajectory is configured to maintain the extension of a current heading angle; a second planning unit, configured to determine, in response to the longitudinal trajectory state being a second state, a first longitudinal trajectory based on the speed information, the first longitudinal trajectory being used to characterize a speed trajectory of smooth deceleration; The third planning unit is used to plan the current frame trajectory of the automatic driving device based on the first lateral trajectory and the first longitudinal trajectory.

16. A trajectory planning device for an autonomous driving device, It is characterized in that include: Memory, used to store programs; A processor, configured to execute a program stored in the memory; When the program stored in the memory is executed, the processor executes the trajectory planning method according to any one of claims 1 to 14.

17. An autonomous driving device, It is characterized in that Includes the trajectory planning device as described in claim 15.

18. A computer-readable storage medium, It is characterized in that The computer-readable medium stores instructions for execution by a computing device, and when the computing device executes the instructions, the method according to any one of claims 1 to 14 is implemented.

Images