Speed planning method and device, electronic device, storage medium and program product

Abstract

The present disclosure provides a speed planning method and apparatus, electronic device, storage medium, and program product. The method includes: determining a vehicle conflict area based on the planned trajectory information of an autonomous vehicle and the predicted trajectory information of an obstacle vehicle; determining the obstacle vehicle's driving intention with respect to the vehicle conflict area based on the obstacle vehicle's historical trajectory information; and adjusting the autonomous vehicle's driving speed based on the planned trajectory information, the predicted trajectory information, the vehicle conflict area, and the driving intention. The present disclosure can solve the speed planning problem for collisions between autonomous vehicles and obstacle vehicles in uncertain environments.

Description

Technical Field

[0001] The embodiments of the present disclosure relate to the field of computer technology, and in particular to a speed planning method, a speed planning device, an electronic device, a computer-readable storage medium, and a computer program product. Background Art

[0002] When the predicted trajectory of an autonomous vehicle conflicts with that of other vehicles, the speed of the autonomous vehicle needs to be planned to avoid collisions. Speed planning for autonomous vehicles in uncertain environments is a challenging task. Summary of the Invention

[0003] The embodiments of the present disclosure provide a speed planning method and device, electronic device, storage medium, and program product, which can solve or partially solve the above-mentioned deficiencies in the prior art or other deficiencies in the prior art.

[0004] According to the first aspect of the present disclosure, a speed planning method is provided, comprising: determining a vehicle conflict area based on planned trajectory information of an autonomous driving vehicle and predicted trajectory information of an obstacle vehicle; determining the obstacle vehicle's driving intention with respect to the vehicle conflict area based on historical trajectory information of the obstacle vehicle; and adjusting the driving speed of the autonomous driving vehicle based on the planned trajectory information, the predicted trajectory information, the vehicle conflict area, and the driving intention.

[0005] According to the second aspect of the present disclosure, a speed planning device is provided, including: a conflict detection module, configured to determine a vehicle conflict area based on planned trajectory information of an autonomous driving vehicle and predicted trajectory information of an obstacle vehicle; an intention prediction module, configured to determine the obstacle vehicle's driving intention with respect to the vehicle conflict area based on historical trajectory information of the obstacle vehicle; and a speed planning module, configured to adjust the driving speed of the autonomous driving vehicle based on the planned trajectory information, the predicted trajectory information, the vehicle conflict area, and the driving intention.

[0006] The electronic device provided according to the third aspect of the present disclosure may include: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed 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 execute the speed planning method described in the first aspect of the present disclosure.

[0007] The computer-readable storage medium provided according to the fourth aspect of the present disclosure includes a computer program, and when the computer program is executed by a processor, the speed planning method described in the first aspect of the present disclosure is implemented.

[0008] The computer program product provided according to the fifth aspect of the present disclosure stores a computer program, and when the computer program is executed by a processor, the speed planning method described in the first aspect of the present disclosure is implemented.

[0009] According to the speed planning method and device, electronic device, storage medium and program product provided in the embodiments of the present disclosure, the vehicle conflict area is determined through the planned trajectory information of the autonomous driving vehicle and the predicted trajectory information of the obstacle vehicle; and the driving intention of the obstacle vehicle with respect to the vehicle conflict area is determined through the historical trajectory information of the obstacle vehicle; the driving speed of the autonomous driving vehicle is adjusted according to the planned trajectory information, the predicted trajectory information, the vehicle conflict area and the driving intention; the historical trajectory information of the obstacle vehicle is used as prior information to implement contingency planning of the speed of the autonomous driving vehicle, solve the speed planning problem for collision conflicts between the autonomous driving vehicle and the obstacle vehicle in an uncertain environment, effectively avoid collision conflicts between the autonomous driving vehicle and the obstacle vehicle, and improve the driving safety of the autonomous driving vehicle.

[0010] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Other features, objects, and advantages of the present disclosure will become more apparent upon reading the detailed description of the non-limiting embodiments made with reference to the following drawings. The drawings are provided for a better understanding of the present invention and do not constitute a limitation of the present disclosure. Among them:

[0012] Figure 1 is a flow chart of a speed planning method according to an embodiment of the present disclosure;

[0013] Figure 2 is a flowchart of adjusting the driving speed of an autonomous vehicle according to some embodiments of the present disclosure;

[0014] Figure 3 is a flow chart for determining an objective function of a two-level optimization game model according to some embodiments of the present disclosure;

[0015] Figure 4 is a schematic diagram of a vehicle conflict area between an autonomous driving vehicle and an obstacle vehicle according to some embodiments of the present disclosure;

[0016] Figure 5 is a flow chart for determining constraints for a two-level optimization game model according to some embodiments of the present disclosure;

[0017] Figure 6is a flow chart of solving an objective function based on constraints according to some embodiments of the present disclosure;

[0018] Figure 7 is a flow chart of an application scenario of the speed planning method according to an embodiment of the present disclosure;

[0019] Figure 8A This is a schematic diagram of the trajectory of a collision between an autonomous vehicle and an obstacle vehicle in a meeting scenario;

[0020] Figures 8B to 8U Yes Figure 8A A schematic diagram of speed and trajectory obtained by performing speed planning according to a speed planning method according to an embodiment of the present disclosure;

[0021] Figure 9 is a block diagram of a speed planning device according to an embodiment of the present disclosure;

[0022] Figure 10 is a block diagram of an example electronic device that can be used to implement embodiments of the present disclosure. DETAILED DESCRIPTION

[0023] The following description of exemplary embodiments of the present disclosure is made in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0024] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in the present disclosure may be combined with each other. The present disclosure will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0025] An exemplary system architecture for implementing the speed planning method provided by the present disclosure may include a terminal device, a network, and a server. The network is used to provide a communication link between the terminal device and the server, and may include various connection types, such as a wired communication link, a wireless communication link, or an optical fiber cable.

[0026] Users can use terminal devices to interact with servers through the network to receive or send information, etc. Various client applications can be installed on the terminal devices, such as map, navigation, entertainment and other client applications.

[0027] The terminal device may be, for example, a vehicle-mounted system of a self-driving car, a delivery robot, or the like. The system may be implemented through hardware, software, or a combination of hardware and software.

[0028] The server can be hardware or software. When the server is hardware, it can be implemented as a distributed server cluster consisting of multiple servers, or it can be implemented as a single server. When the server is software, it can be implemented as multiple software or software modules (for example, to provide distributed services), or it can be implemented as a single software or software module. No specific limitations are given here.

[0029] It should be pointed out that the execution entity of the speed planning method provided by the present disclosure (hereinafter referred to as the "execution entity") can be the server in the above system architecture, the terminal device in the above system architecture, or the server and terminal device in the above system architecture.

[0030] When the speed planning method is executed by the server, the server can determine the vehicle conflict area based on the planned trajectory information of the autonomous driving vehicle and the predicted trajectory information of the obstacle vehicle, and determine the driving intention of the obstacle vehicle with respect to the vehicle conflict area based on the historical trajectory information of the obstacle vehicle. The server can then adjust the driving speed of the autonomous driving vehicle based on the planned trajectory information, the predicted trajectory information, the vehicle conflict area and the driving intention, and send the control command to the terminal device, such as the vehicle-mounted system of the autonomous driving vehicle, and control the speed of the autonomous driving vehicle through the vehicle-mounted system according to the control command.

[0031] Another applicable scenario is that speed planning is completed directly by terminal devices, such as the vehicle-mounted system of an autonomous vehicle, instead of through a server. At this time, the vehicle-mounted system can determine the vehicle conflict area based on the planned trajectory information of the autonomous vehicle and the predicted trajectory information of the obstacle vehicle, and determine the driving intention of the obstacle vehicle with respect to the vehicle conflict area based on the historical trajectory information of the obstacle vehicle. Then, the vehicle-mounted system can adjust the driving speed of the autonomous vehicle based on the planned trajectory information, predicted trajectory information, vehicle conflict area and driving intention, and control the speed of the autonomous vehicle.

[0032] In addition, speed planning can also be completed jointly by a server and a terminal device, such as the vehicle-mounted system of an autonomous vehicle. The present disclosure does not limit the operations performed by the server and the terminal device at this time. For example, the server can determine the vehicle conflict area based on the planned trajectory information of the autonomous vehicle and the predicted trajectory information of the obstacle vehicle, and determine the driving intention of the obstacle vehicle with respect to the vehicle conflict area based on the historical trajectory information of the obstacle vehicle, and then send the vehicle conflict area and driving intention to the vehicle-mounted system. The vehicle-mounted system can adjust the driving speed of the autonomous vehicle based on the planned trajectory information, predicted trajectory information, vehicle conflict area and driving intention, and control the speed of the autonomous vehicle.

[0033] In addition, in the technical solutions involved in this disclosure, the acquisition, storage, use, processing, transportation, provision and disclosure of vehicle speed, trajectory information, etc. are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0034] Figure 1 FIG. 1 shows a flow chart of a speed planning method 100 according to an embodiment of the present disclosure. Figure 1 As shown, the speed planning method 100 may include the following steps:

[0035] S101. Determine a vehicle conflict area based on the planned trajectory information of the autonomous driving vehicle and the predicted trajectory information of the obstacle vehicle.

[0036] In an embodiment of the present disclosure, the execution entity may determine the vehicle conflict area based on the planned trajectory information of the autonomous driving vehicle and the predicted trajectory information of the obstacle vehicle.

[0037] In an embodiment of the present disclosure, the execution entity may first obtain the trajectory information of the autonomous driving vehicle, where the trajectory information may refer to the planned trajectory information obtained by performing trajectory planning on the autonomous driving vehicle. The planned trajectory information may include information such as the vehicle's position and speed. For example, the execution entity may obtain the planned trajectory information of the autonomous driving vehicle from the planning module of the autonomous driving vehicle. The execution entity may also obtain the trajectory information of other vehicles, where other vehicles may refer to vehicles traveling on the road section through which the planned trajectory information of the autonomous driving vehicle passes, where the trajectory information may refer to predicted trajectory information obtained by performing trajectory prediction on other vehicles. The predicted trajectory information may include information such as the vehicle's position and speed. For example, the execution entity may obtain the predicted trajectory information of other vehicles from the prediction module of the autonomous driving vehicle.

[0038] Then, the execution entity can determine whether the planned trajectory information and the predicted trajectory information overlap at a certain moment based on the planned trajectory information and the predicted trajectory information. If the planned trajectory information and the predicted trajectory information overlap at a certain moment, it indicates that the autonomous vehicle will collide with other vehicles at that moment. The other vehicles that collide with the autonomous vehicle can be regarded as obstacle vehicles, and the area where the collision occurs can be regarded as the vehicle conflict area. It should be noted that the planned trajectory information of the autonomous vehicle can overlap with the predicted trajectory information of multiple other vehicles, thereby obtaining multiple obstacle vehicles and multiple vehicle conflict areas. The embodiments of the present disclosure do not limit the number of obstacle vehicles and the number of vehicle conflict areas.

[0039] S102: Determine the driving intention of the obstructing vehicle with respect to the vehicle conflict area based on the historical trajectory information of the obstructing vehicle.

[0040] In an embodiment of the present disclosure, the execution entity may determine the driving intention of the obstructing vehicle with respect to the vehicle conflict area based on the historical trajectory information of the obstructing vehicle.

[0041] In an embodiment of the present disclosure, the execution entity may also obtain the historical driving trajectory information of the obstructing vehicle, where the historical driving trajectory information may refer to the historical trajectory information that the obstructing vehicle has already traveled. The historical trajectory information may include information such as the location of the vehicle. For example, the execution entity may obtain the historical trajectory information of the obstructing vehicle from the electronic map server through the map application of the autonomous driving vehicle. Then, the execution entity may judge the driving intention of the obstructing vehicle with respect to the vehicle conflict area based on the historical trajectory information of the obstructing vehicle. For example, the driving intention of the obstructing vehicle with respect to the vehicle conflict area may include overtaking or giving way, where overtaking may mean that the obstructing vehicle passes through the vehicle conflict area before the autonomous driving vehicle, and giving way may mean that the obstructing vehicle allows the autonomous driving vehicle to pass through the vehicle conflict area first.

[0042] S103: Adjust the driving speed of the autonomous driving vehicle based on the planned trajectory information, the predicted trajectory information, the vehicle conflict area, and the driving intention.

[0043] In an embodiment of the present disclosure, the execution entity can adjust the driving speed of the autonomous driving vehicle based on the planned trajectory information, the predicted trajectory information, the vehicle conflict area and the driving intention.

[0044] In an embodiment of the present disclosure, the execution entity may adjust the driving speed of the autonomous vehicle based on the planned trajectory information of the autonomous vehicle, the predicted trajectory information of the obstructing vehicle, the vehicle conflict area between the autonomous vehicle and the obstructing vehicle, and the obstructing vehicle's driving intention with respect to the vehicle conflict area, so that the driving speed of the autonomous vehicle matches the driving intention of the obstructing vehicle with respect to the vehicle conflict area, thereby preventing the autonomous vehicle and the obstructing vehicle from appearing in the vehicle conflict area at the same time and avoiding a collision between the autonomous vehicle and the obstructing vehicle. For example, when the obstructing vehicle's driving intention with respect to the vehicle conflict area is to cut in, the execution entity may adjust the driving speed of the autonomous vehicle to yield, allowing the obstructing vehicle to pass through the vehicle conflict area first; when the obstructing vehicle's driving intention with respect to the vehicle conflict area is to cut in, the execution entity may adjust the driving speed of the autonomous vehicle to cut in, allowing the obstructing vehicle to pass through the vehicle conflict area first.

[0045] The embodiments of the present disclosure do not limit the implementation method of adjusting the driving speed of the autonomous driving vehicle based on the planned trajectory information, predicted trajectory information, vehicle conflict area and driving intention. In an optional example, the execution entity may adjust the driving speed of the autonomous driving vehicle through a large language model based on the planned trajectory information, predicted trajectory information, vehicle conflict area and driving intention. In another optional example, the execution entity may adjust the driving speed of the autonomous driving vehicle through a game model based on the planned trajectory information, predicted trajectory information, vehicle conflict area and driving intention. In yet another optional example, the execution entity may adjust the driving speed of the autonomous driving vehicle through a decision tree and speed planning method based on the planned trajectory information, predicted trajectory information, vehicle conflict area and driving intention.

[0046] The speed planning method 100 of the embodiment of the present disclosure determines the vehicle conflict area through the planned trajectory information of the autonomous driving vehicle and the predicted trajectory information of the obstacle vehicle; and determines the driving intention of the obstacle vehicle with respect to the vehicle conflict area through the historical trajectory information of the obstacle vehicle; adjusts the driving speed of the autonomous driving vehicle according to the planned trajectory information, the predicted trajectory information, the vehicle conflict area and the driving intention; uses the historical trajectory information of the obstacle vehicle as prior information to implement contingency planning of the speed of the autonomous driving vehicle, solve the speed planning problem for collision conflicts between the autonomous driving vehicle and the obstacle vehicle in an uncertain environment, effectively avoid collision conflicts between the autonomous driving vehicle and the obstacle vehicle, and improve the driving safety of the autonomous driving vehicle.

[0047] It should be understood that the steps shown in method 100 are not exclusive, and other steps may be performed before, after, or between any of the steps shown. In addition, some of the steps shown may be performed simultaneously, or may be performed in a different order than the steps shown in the method 100. Figure 1 Executed in the order shown.

[0048] Figure 2 FIG. 1 shows a flow chart of adjusting the driving speed of an autonomous vehicle according to some embodiments of the present disclosure. Figure 2 As shown, step S103 adjusts the driving speed of the autonomous driving vehicle based on the planned trajectory information, the predicted trajectory information, the vehicle conflict area, and the driving intention, and may include the following steps:

[0049] S201. Based on the planned trajectory information, the predicted trajectory information, the vehicle conflict area and the driving intention, a two-level optimization game model between the autonomous driving vehicle and the obstacle vehicle is established.

[0050] In an embodiment of the present disclosure, the execution entity can establish a two-level optimization game model between the autonomous driving vehicle and the obstacle vehicle based on the planned trajectory information, the predicted trajectory information, the vehicle conflict area and the driving intention.

[0051] In an embodiment of the present disclosure, a game model can be used to characterize the interaction between an autonomous driving vehicle and an obstacle vehicle. The autonomous driving vehicle can be used as a leader and the obstacle vehicle can be used as a follower. The two-level optimization game model can include two processes: in the first process, the follower predicts the leader's future movement and optimizes its own movement; in the second process, the leader optimizes its own movement based on the follower's movement. The two processes of the two-level optimization game model can constitute a nested optimization problem. In some optional embodiments of the present disclosure, step S201 establishes a two-level optimization game model between the autonomous driving vehicle and the obstacle vehicle based on the planned trajectory information, the predicted trajectory information, the vehicle conflict area, and the driving intention, and can include: determining the objective function of the two-level optimization game model based on the planned trajectory information, the predicted trajectory information, and the driving intention; and determining the constraints of the two-level optimization game model based on the planned trajectory information, the predicted trajectory information, and the vehicle conflict area.

[0052] In an optional example, the autonomous vehicle and the obstacle vehicle can be placed in t The states at each moment are represented as and , the trajectories of the autonomous vehicle and the obstacle vehicle can be represented as and ,The effective execution of the two processes of the two-level optimization game model relies on the following assumptions:

[0053] Strong assumption: In a short period of time, the obstacle vehicle can well predict the trajectory of the autonomous vehicle, making .in, is the trajectory of the autonomous vehicle predicted by the obstacle vehicle, It is the actual driving trajectory of the autonomous vehicle.

[0054] Based on the above assumptions, the first process of the model can be expressed as:

[0055] (Formula 1)

[0056] The second process of the model can be expressed as:

[0057] (Formula 2)

[0058] in, and are the objective functions of the autonomous driving vehicle and the obstacle vehicle, and are the driving trajectories of the autonomous vehicle and the obstacle vehicle, and They are the optimized driving trajectories of the autonomous driving vehicle and the obstacle vehicle respectively.

[0059] For a follower, given the leader's state sequence , the optimal control problem of the follower can be expressed as follows:

[0060] (Formula 3)

[0061] st

[0062] ,

[0063] .

[0064] Known follower reaction status and reaction actions , the leader’s optimal control problem can be formulated as follows:

[0065] (Formula 4)

[0066] st

[0067] ,

[0068] .

[0069] in, and are the state sequences corresponding to the driving trajectories of the autonomous vehicle and the obstacle vehicle, and are the control quantities corresponding to the driving trajectories of the autonomous driving vehicle and the obstacle vehicle, and are the inequality constraints for the optimal control problems of the autonomous driving vehicle and the obstacle vehicle, and are the equality constraints for the optimal control problems of the autonomous vehicle and the obstacle vehicle, respectively.

[0070] S202. Obtain an optimization solution by solving a two-level optimization game model, and adjust the driving speed of the autonomous driving vehicle based on the optimization solution.

[0071] In an embodiment of the present disclosure, the execution entity can obtain an optimization solution by solving a two-level optimization game model and adjust the driving speed of the autonomous driving vehicle according to the optimization solution.

[0072] In the embodiments of the present disclosure, there is no limitation on the implementation method for solving the two-level optimization game model to obtain the optimized solution. In an optional example, the two-level optimization game model can be solved by the diagonal method to obtain the optimized solution. In another optional example, the two-level optimization game model can be solved by the Karush-Kuhn-Tucker condition (KKT) to obtain the optimized solution. In another optional example, the two-level optimization game model can be solved by an artificial intelligence method to obtain the optimized solution. In some optional embodiments of the present disclosure, after constructing the two-level optimization game model by determining the objective function and the constraints, step S202 of obtaining the optimized solution by solving the two-level optimization game model can include: solving the objective function based on the constraints to obtain the optimized solution.

[0073] In some optional embodiments of the present disclosure, when an obstructing vehicle intends to overtake the autonomous vehicle in a conflict zone, it will accelerate, overtaking the autonomous vehicle and passing through the conflict zone first. When an obstructing vehicle intends to yield to the autonomous vehicle, it will decelerate, avoiding the autonomous vehicle and allowing it to pass through the conflict zone first. Therefore, the probability of an obstructing vehicle accelerating or decelerating can be used to represent its intention to move through the conflict zone.

[0074] Optionally, the executing entity may assess the acceleration or deceleration probability of the obstructing vehicle based on historical data trends using the historical trajectory information of the obstructing vehicle. Step S102, determining the obstructing vehicle's driving intention with respect to the vehicle conflict area based on the historical trajectory information of the obstructing vehicle, may include: establishing an acceleration prediction model for the obstructing vehicle based on the historical trajectory information; determining the acceleration of the obstructing vehicle's current trajectory point using the acceleration prediction model; and determining the acceleration or deceleration probability of the obstructing vehicle's current trajectory point based on the acceleration of the obstructing vehicle's current trajectory point and the acceleration or deceleration probability of the obstructing vehicle's previous trajectory point, wherein the determined acceleration probability indicates the obstructing vehicle's driving intention to cut in with respect to the vehicle conflict area, and the determined deceleration probability indicates the obstructing vehicle's driving intention to yield with respect to the vehicle conflict area.

[0075] In an optional example, the most recent 6 frames of driving trajectory information can be taken as the historical trajectory information of the obstacle vehicle. For example, the coordinate set of the trajectory points of the 6 frames of driving trajectory information is: , the speed of the obstacle vehicle's trajectory point in each frame of the driving trajectory information can be calculated by the position difference, and its calculation formula is as follows:

[0076] (Formula 5)

[0077] The acceleration of the vehicle at each frame of trajectory information can be calculated by the velocity difference, and the formula is as follows:

[0078] (Formula 6)

[0079] The acceleration prediction model is fitted by a cubic polynomial, and its formula is as follows:

[0080] (Formula 7)

[0081] in, To fit the parameters, the historical acceleration information obtained by differential calculation is substituted into the fitting parameters to obtain the fitting parameters.

[0082] If the acceleration probability of the obstacle vehicle at the trajectory point of the previous frame's driving trajectory information is , the maximum intention fluctuation between frames is no more than , then the prior probability of the trajectory point of the current frame driving trajectory information is p The calculation formula is as follows:

[0083] (Formula 8)

[0084] Among them, sign() is the sign function, is a fixed parameter, is the acceleration of the trajectory point of the current frame driving trajectory information.

[0085] Figure 3 FIG. 1 shows a flow chart for determining the objective function of a two-level optimization game model according to some embodiments of the present disclosure. Figure 3 As shown, based on the planned trajectory information, the predicted trajectory information and the driving intention, determining the objective function of the two-level optimization game model may include the following steps:

[0086] S301. Obtain an objective function of the autonomous driving vehicle based on the planned trajectory information and the acceleration or deceleration probability of the obstacle vehicle's current trajectory point.

[0087] In an embodiment of the present disclosure, an execution entity may derive an objective function for the autonomous vehicle based on the planned trajectory information and the probability of acceleration or deceleration of the obstacle vehicle's current trajectory point. In some optional embodiments, the planned trajectory information for the autonomous vehicle may be a sequence of trajectory points, each of which may include position information, velocity information, acceleration information, and the like. A cost function for the autonomous vehicle may be constructed based on the state changes of the trajectory points in the sequence of trajectory points in the planned trajectory information, and the objective function of the autonomous vehicle may be derived based on the cost function of the autonomous vehicle.

[0088] Optionally, step S301 obtains the objective function of the autonomous driving vehicle based on the planned trajectory information and the acceleration or deceleration probability of the current trajectory point of the obstacle vehicle, and may include: determining the cost function of the autonomous driving vehicle based on the planned trajectory information; and obtaining the objective function of the autonomous driving vehicle based on the cost function of the autonomous driving vehicle and the acceleration or deceleration probability of the current trajectory point of the obstacle vehicle.

[0089] In some optional embodiments, determining the cost function of the autonomous driving vehicle based on the planned trajectory information may include: determining a state sequence and an acceleration sequence of the trajectory points of the autonomous driving vehicle based on the planned trajectory information, wherein the state of each trajectory point in the state sequence includes posture information, velocity information, and acceleration information; determining a first state change cost of the state sequence of the trajectory points of the autonomous driving vehicle relative to the target state of the autonomous driving vehicle, and a first smooth transition cost of the acceleration sequence of the trajectory points of the autonomous driving vehicle, wherein the cost function of the autonomous driving vehicle includes a first state change cost and a first smooth transition cost.

[0090] It should be noted that the target state of the autonomous vehicle can be set as needed, and the embodiments of this disclosure do not limit this. For example, the speed limit of the road section through which the planned trajectory information passes can be used as the speed information of the target state. The position information of the target state can be determined based on the speed limit and time of the road section through which the planned trajectory information passes. The maximum acceleration of the autonomous vehicle can be used as the acceleration information of the target state.

[0091] In some optional implementations, an objective function for the autonomous vehicle can be constructed based on the prior probability of the obstacle vehicle's current trajectory point and the autonomous vehicle's cost function. The objective function for the autonomous vehicle can include an acceleration cost function and a deceleration cost function. The prior probability of the obstacle vehicle's current trajectory point can be either an acceleration probability or a deceleration probability.

[0092] Optionally, obtaining the objective function of the autonomous driving vehicle based on the cost function of the autonomous driving vehicle and the acceleration or deceleration probability of the current trajectory point of the obstacle vehicle may include: obtaining the acceleration cost function and deceleration cost function of the autonomous driving vehicle based on the cost function of the autonomous driving vehicle; determining the weights of the acceleration cost function and the deceleration cost function based on the acceleration or deceleration probability of the current trajectory point of the obstacle vehicle; and obtaining the objective function of the autonomous driving vehicle by taking the weighted sum of the acceleration cost function and the deceleration cost function.

[0093] S302: Determine the objective function of the obstacle vehicle based on the predicted trajectory information.

[0094] In embodiments of the present disclosure, an execution entity may determine an objective function for an obstructing vehicle based on the predicted trajectory information. In some optional embodiments, the predicted trajectory information for the obstructing vehicle may be a sequence of trajectory points, each of which may include position, velocity, and acceleration information. A cost function for the obstructing vehicle may be constructed based on the state changes of the trajectory points in the sequence of predicted trajectory information, and the objective function for the obstructing vehicle may be derived based on the cost function.

[0095] Optionally, step S302, determining the objective function of the obstructing vehicle based on the predicted trajectory information, may include: determining a cost function of the obstructing vehicle based on the predicted trajectory information as the objective function of the obstructing vehicle. In some optional embodiments, determining the cost function of the obstructing vehicle based on the predicted trajectory information may include: determining a state sequence and a jerk sequence of trajectory points of the obstructing vehicle based on the predicted trajectory information, wherein the state of each trajectory point in the state sequence includes position information, velocity information, and acceleration information; determining a second state change cost of the state sequence of the obstructing vehicle's trajectory points relative to the target state of the obstructing vehicle, and a second smooth transition cost of the jerk sequence of the obstructing vehicle's trajectory points, wherein the cost function of the obstructing vehicle includes the second state change cost and the second smooth transition cost.

[0096] It should be noted that the target state of the obstructing vehicle can be set as needed and is not limited in the embodiments of this disclosure. For example, the speed limit of the road section where the predicted trajectory information passes can be used as the target state speed information, the target state posture information can be determined based on the speed limit and time of the road section where the predicted trajectory information passes, and the maximum acceleration of the obstructing vehicle can be used as the target state acceleration information.

[0097] S303. Based on the objective function of the autonomous driving vehicle and the objective function of the obstacle vehicle, the objective function of the two-level optimization game model is obtained.

[0098] In embodiments of the present disclosure, the executing entity may derive the objective function of the two-level optimization game model based on the objective function of the autonomous vehicle and the objective function of the obstacle vehicle. In some alternative embodiments, the objective function of the two-level optimization game model may be obtained by performing a weighted summation of the objective function of the autonomous vehicle and the objective function of the obstacle vehicle. The weights of the objective function of the autonomous vehicle and the objective function of the obstacle vehicle may be set as desired, and the embodiments of the present disclosure are not limited thereto.

[0099] In an optional example, the state sequence of the trajectory point is , the acceleration sequence of the trajectory points is ,in It's posture. It's speed, is the acceleration, is the jerk.

[0100] The objective function includes the state change cost and the smooth transition cost. The target state of the autonomous driving vehicle is , the target state of the obstacle vehicle is , then the cost function of the autonomous vehicle is as follows:

[0101] (Formula 9)

[0102] In Public Notice 9, is the state sequence of the autonomous vehicle The state at time k, is the jerk at time k in the jerk sequence of the autonomous vehicle.

[0103] The cost function of the obstacle vehicle is as follows:

[0104] (Formula 10)

[0105] In formula 10, is the state sequence of the obstacle vehicle The state at time k, is the jerk of the obstacle vehicle at time k in the jerk sequence.

[0106] In Formula 9 and Formula 10, Q and R are the weight parameter matrices of the obstructing vehicle and the autonomous driving vehicle. The weight parameter matrices of the obstructing vehicle and the autonomous driving vehicle can be the same or different. For ease of explanation, in this embodiment, the obstructing vehicle and the autonomous driving vehicle use the same weight parameter matrix. Q and R can be defined as:

[0107] ,

[0108] .

[0109] The objective function of an autonomous vehicle can be expressed as:

[0110] (Formula 11)

[0111] in, p is the prior probability that the obstacle vehicle will slow down in response to the autonomous vehicle, is the acceleration cost function, is the deceleration cost function.

[0112] The weight of the autonomous vehicle is , the weight of the obstacle vehicle is , then the objective function of the two-level optimization game model is as follows:

[0113] (Formula 12)

[0114] in, , The closer the value is to 1, the more emphasis is placed on the driving goal and feeling of the autonomous vehicle. On the contrary, it means that the emphasis is placed on the feeling of the obstacle vehicle.

[0115] This embodiment uses the acceleration or deceleration probability of the obstructing vehicle as prior information, and can implement contingency planning of the speed of the autonomous driving vehicle. It can take into account both the situations of accelerating to overtake and decelerating to give way. It can not only meet the requirements of overtaking efficiency, but also have the robustness of giving way and deceleration to accidental events, and can meet the requirements of situations such as sudden changes in environmental conditions.

[0116] In some optional embodiments of the present disclosure, after the execution entity determines the vehicle conflict area based on the planned trajectory information of the autonomous driving vehicle and the predicted trajectory information of the obstacle vehicle, it can further determine the boundary of the vehicle conflict area. Optionally, the projection position of the overlapping area of the planned trajectory of the autonomous driving vehicle and the predicted trajectory of the obstacle vehicle on their respective trajectories can be used as the boundary of the vehicle conflict area. Step S101 determines the vehicle conflict area based on the planned trajectory information of the autonomous driving vehicle and the predicted trajectory information of the obstacle vehicle, which may include: determining the overlapping area of the planned trajectory and the predicted trajectory at the same time as the vehicle conflict area based on the planned trajectory information and the predicted trajectory information; and determining the boundary of the vehicle conflict area based on the projection of the overlapping area on the planned trajectory and the predicted trajectory.

[0117] In an alternative example, Figure 4 As shown in Figure 1, CDEF is the vehicle conflict area where the trajectories of the autonomous vehicle and the obstacle vehicle overlap, where C is the front boundary of the autonomous vehicle (self-vehicle) reaching the vehicle conflict area, D is the rear boundary of the autonomous vehicle reaching the vehicle conflict area, E is the front boundary of the obstacle vehicle (other vehicle) reaching the vehicle conflict area, and F is the rear boundary of the obstacle vehicle reaching the vehicle conflict area.

[0118] Figure 5 FIG. 1 shows a flow chart of determining the constraints of a two-level optimization game model according to some embodiments of the present disclosure. Figure 5 As shown, based on the planned trajectory information, the predicted trajectory information and the vehicle conflict area, determining the constraints of the two-level optimization game model may include the following steps:

[0119] S501: Determine the equality constraints of the two-level optimization game model based on the planned trajectory information and the predicted trajectory information.

[0120] In embodiments of the present disclosure, the execution entity may determine the equality constraints of the two-level optimization game model based on the planned trajectory information and the predicted trajectory information. The equality constraints of the two-level optimization game model may be set as needed and are not limited in the embodiments of the present disclosure. In some optional embodiments, the equality constraints of the two-level optimization game model may include: initial state constraints, state transition constraints, and contingency planning constraints.

[0121] Optionally, step S501 determines the equality constraints of the two-level optimization game model based on the planned trajectory information and the predicted trajectory information, which may include: obtaining the initial state constraint of the autonomous driving vehicle based on the state of the first trajectory point in the planned trajectory information being the initial state of the autonomous driving vehicle; obtaining the initial state constraint of the obstacle vehicle based on the state of the first trajectory point in the predicted trajectory information being the initial state of the obstacle vehicle; obtaining the state transfer constraint of the autonomous driving vehicle based on the states and jerk of adjacent trajectory points in the planned trajectory information satisfying the transfer matrix; obtaining the state transfer constraint of the obstacle vehicle based on the states and jerk of adjacent trajectory points in the predicted trajectory information satisfying the transfer matrix; and obtaining the contingency planning constraint of the autonomous driving vehicle based on the jerk of the first trajectory point in the planned trajectory information being equal to the acceleration and deceleration of the obstacle vehicle.

[0122] In an optional example, the equality constraint may include: initial state constraint, state transition constraint and contingency planning constraint.

[0123] The initial state constraint can be expressed as:

[0124] (Formula 13)

[0125] in, is the current state of the vehicle, is the state of the trajectory point in the first frame of the state sequence.

[0126] The state transition constraint can be expressed as:

[0127] (Formula 14)

[0128] Among them, A and B are transfer matrices, is the state of the trajectory point in the k+1th frame of the state sequence, is the state of the trajectory point of the kth frame of the state sequence, The jerk of the trajectory point at the kth frame of the jerk sequence. A and B can be defined as:

[0129] ,

[0130] .

[0131] The contingency planning constraints can be expressed as:

[0132] (Formula 15)

[0133] in, and are the jerk of the trajectory point in the first frame of the jerk sequence of the autonomous vehicle’s response to the acceleration and deceleration of the obstacle vehicle.

[0134] This embodiment uses contingency planning constraints and integrates the speed planning results of the autonomous driving vehicle's acceleration or deceleration. This can improve the smoothness of the autonomous driving vehicle's state changes while ensuring the autonomous driving vehicle's safe driving, and can greatly enhance the autonomous driving vehicle's control capabilities.

[0135] S502 : Determine inequality constraints of a two-level optimization game model based on the planned trajectory information, the predicted trajectory information, and the boundary of the vehicle conflict area.

[0136] In embodiments of the present disclosure, the execution entity may determine the inequality constraints of the two-level optimization game model based on the planned trajectory information, the predicted trajectory information, and the boundaries of the vehicle conflict zone. The inequality constraints of the two-level optimization game model may be set as needed and are not limited in the embodiments of the present disclosure. In some optional embodiments, the equality constraints of the two-level optimization game model may include: state constraints, reversing constraints, and collision constraints.

[0137] Optionally, step S502 determines the inequality constraints of the two-level optimization game model based on the planned trajectory information, the predicted trajectory information and the boundary of the vehicle conflict area, which may include: obtaining the state constraint of the autonomous driving vehicle based on the state of each trajectory point in the planned trajectory information being between the upper boundary state and the lower boundary state of the autonomous driving vehicle; obtaining the state constraint of the obstacle vehicle based on the state of each trajectory point in the predicted trajectory information being between the upper boundary state and the lower boundary state of the obstacle vehicle; obtaining the reversing constraint of the autonomous driving vehicle based on the position of the subsequent trajectory point of adjacent trajectory points in the planned trajectory information being not less than the position of the previous trajectory point; obtaining the reversing constraint of the obstacle vehicle based on the position of the subsequent trajectory point of adjacent trajectory points in the predicted trajectory information being not less than the position of the previous trajectory point; and obtaining the collision constraint based on the sum of the distance between the trajectory point of the planned trajectory information and the boundary of the vehicle conflict area and the distance between the trajectory point of the predicted trajectory information and the boundary of the vehicle conflict area at the same time being not less than a preset safety distance.

[0138] It should be noted that the upper and lower boundary states can be set as needed and are not limited in the embodiments of the present disclosure. For example, the speed limit of the road section can be used as the upper limit of the speed, and the speed of zero can be used as the lower limit of the speed. The upper limit of the posture can be determined based on the speed limit of the road section and time, and the posture of zero can be used as the lower limit of the posture. The maximum acceleration can be used as the upper limit of the acceleration, and the maximum deceleration can be used as the lower limit of the acceleration.

[0139] Optionally, obtaining a collision constraint based on the sum of the distances between a trajectory point in the planned trajectory information and the boundary of the vehicle conflict area and the distances between a trajectory point in the predicted trajectory information and the boundary of the vehicle conflict area at the same time being greater than a preset safety distance may include: determining a first geometric distance based on the front and rear boundaries of the vehicle conflict area on the planned trajectory and the position of the current trajectory point in the planned trajectory information; determining a second geometric distance based on the front and rear boundaries of the vehicle conflict area on the predicted trajectory and the position of the current trajectory point in the predicted trajectory information; and obtaining the collision constraint based on the sum of the first and second geometric distances being no less than a preset safety distance. The collision constraint method of this embodiment can simplify the collision detection process.

[0140] In an optional example, the inequality constraints include: state constraints, reversing constraints, and collision constraints.

[0141] State constraints mainly include state boundary constraints, k The upper boundary state at time , the lower boundary state at time k is , then the state constraint can be expressed as:

[0142] (Formula 16)

[0143] (Formula 17)

[0144] The reversing constraint can be expressed as:

[0145] (Formula 18)

[0146] The state sequence is formed into a table with 3 rows and N columns, with the posture information as the first row, the velocity information as the second row, and the acceleration information as the third row. is the position of the vehicle at time k, is the position of the vehicle at time k+1.

[0147] The collision constraint can be expressed as:

[0148] (Formula 19)

[0149] in, It's a safe distance. is the geometric distance of the autonomous driving vehicle at time k, is the geometric distance of the obstacle vehicle at time k.

[0150] The geometric distance of the autonomous vehicle can be expressed as:

[0151] (Formula 20)

[0152] The geometric distance of the obstacle vehicle can be expressed as:

[0153] (Formula 21)

[0154] in, is the position of the autonomous vehicle at time k, is the boundary point position of the front boundary C of the vehicle conflict area, is the boundary point position of the rear boundary D of the vehicle conflict area, is the position of the obstacle vehicle at time k, is the boundary point position of the front boundary E of the vehicle conflict area, is the boundary point position of the rear boundary F of the vehicle conflict area.

[0155] Among them, if the obstacle vehicle is driven by a human, the safety distance reserved when taking the acceleration response strategy is It can be longer, to reserve a safe distance when adopting a deceleration response strategy It can be shorter. Because the self-driving vehicle adopts contingency speed planning, it takes into account the situations of accelerating to overtake and slowing down to give way, so the reserved safety distance A compromise can be expressed as:

[0156] (Formula 22)

[0157] Figure 6 FIG. 1 shows a flow chart of solving an objective function based on constraint conditions according to some embodiments of the present disclosure. Figure 6 As shown, step S202 solves the objective function based on the constraint conditions to obtain the optimized solution, which may include the following steps:

[0158] S601: Convert the objective function of the obstacle vehicle into equality constraints and inequality constraints to obtain new equality constraints and new inequality constraints.

[0159] S602. Based on the new equality constraint, the new inequality constraint, the equality constraint of the two-level optimization game model, and the inequality constraint of the two-level optimization game model, the objective function of the autonomous driving vehicle is solved to obtain an optimized solution.

[0160] In an optional example, the objective function of the obstacle vehicle can be converted into equality constraints and inequality constraints through KKT conditions. The principle of conversion through KKT conditions is as follows.

[0161] The Lagrangian function of the follower optimal control problem is expressed as follows:

[0162] (Formula 23)

[0163] Where λ is the Lagrange multiplier corresponding to the equality constraint, μ is the KKT multiplier corresponding to the inequality constraint, and the KKT conditions include:

[0164] ,

[0165] ,

[0166] ,

[0167] ,

[0168] .

[0169] After KKT condition transformation, the nested optimization problem is expressed as follows:

[0170] (Formula 24)

[0171] st

[0172] , (1)

[0173] , (2)

[0174] , (3)

[0175] , (4)

[0176] , (5)

[0177] , (6)

[0178] , (7)

[0179] , (8)

[0180] , (9)

[0181] , (10)

[0182] , (11)

[0183] , (12)

[0184] (13)

[0185] , (14)

[0186] (15)

[0187] Among them, p is the prior probability that the obstacle vehicle will take a deceleration response, and 1-p is the prior probability that the obstacle vehicle will take an acceleration response.

[0188] Figure 7 FIG. 1 is a flow chart of an application scenario of the speed planning method according to an embodiment of the present disclosure. Figure 7 As shown, the speed planning method includes the following steps:

[0189] Step 1: Obtain the planned trajectory information of the self-driving vehicle (including position, speed, and other information), and obtain the predicted trajectory information and historical trajectory information of the other vehicle (obstacle vehicle). The predicted trajectory information includes position, speed, and other information;

[0190] Step 2: Evaluate the probability of the other vehicle cutting in or giving way based on historical trajectory information;

[0191] Step 3: Determine the vehicle conflict area based on the planned trajectory information and the predicted trajectory information, and calculate the boundary of the vehicle conflict area;

[0192] Step 4: Construct a two-level optimization game model;

[0193] Step 5: Adjust the vehicle's speed by solving the two-level optimization game model to obtain the speed planning result.

[0194] Figure 8A This is a schematic diagram of the trajectory of the collision between the autonomous vehicle and the obstacle vehicle in the oncoming vehicle scene. Figure 8A As shown in the figure, the red line aa represents the predicted trajectory of the obstacle vehicle, and the blue line bb represents the planned trajectory of the autonomous vehicle. In the oncoming vehicle scenario, the planned trajectory of the autonomous vehicle and the predicted trajectory of the obstacle vehicle overlap in an area cc, where the autonomous vehicle and the obstacle vehicle will collide.

[0195] Figures 8B to 8U Yes Figure 8A Schematic diagram of the speed and trajectory obtained by speed planning according to the speed planning method of the embodiment of the present disclosure. Figures 8B to 8UAs shown in the figure, the purple line dd is the planned speed of the autonomous vehicle, the red line ee is the trajectory of the autonomous vehicle corresponding to the purple line dd, the blue line ff is the speed of the obstacle vehicle, and the blue line gg is the trajectory of the obstacle vehicle corresponding to the blue line ff. As can be seen from the figure, after the autonomous vehicle detects the obstacle vehicle's deceleration intention, the contingency planning process gradually favors acceleration over overtaking and abandons deceleration avoidance. Here, Pa is the probability of the autonomous vehicle accelerating, Pd is the probability of decelerating, Traj_F is the trajectory of the obstacle vehicle, Traj_L is the trajectory of the autonomous vehicle, dec is deceleration, acc is acceleration, vxF is the speed of the obstacle vehicle, and vxL is the speed of the autonomous vehicle.

[0196] The embodiment of the present disclosure further provides a speed planning device 900, Figure 9 FIG. 9 is a block diagram of a speed planning device 900 according to an embodiment of the present disclosure. The speed planning device 900 according to an embodiment of the present disclosure can execute the above-mentioned speed planning method 100. Figure 9 As shown, the speed planning device 900 may include:

[0197] The conflict detection module 901 is configured to determine a vehicle conflict area based on the planned trajectory information of the autonomous driving vehicle and the predicted trajectory information of the obstacle vehicle;

[0198] The intention prediction module 902 is configured to determine the driving intention of the obstacle vehicle with respect to the vehicle conflict area based on the historical trajectory information of the obstacle vehicle;

[0199] The speed planning module 903 is configured to adjust the driving speed of the autonomous driving vehicle based on the planned trajectory information, the predicted trajectory information, the vehicle conflict area and the driving intention.

[0200] In some optional implementations, the speed planning module 903 includes:

[0201] a model building submodule configured to build a two-level optimization game model between the autonomous driving vehicle and the obstacle vehicle based on the planned trajectory information, the predicted trajectory information, the vehicle conflict area, and the driving intention;

[0202] The model solving submodule is configured to obtain an optimization solution by solving the two-level optimization game model, and adjust the driving speed of the autonomous driving vehicle based on the optimization solution.

[0203] In some optional implementations, the model building submodule is further configured to:

[0204] determining an objective function of the two-level optimization game model based on the planned trajectory information, the predicted trajectory information, and the driving intention;

[0205] Determining constraints of the two-level optimization game model based on the planned trajectory information, the predicted trajectory information, and the vehicle conflict area;

[0206] The model solving submodule is further configured as follows:

[0207] The objective function is solved based on the constraint conditions to obtain the optimized solution.

[0208] In some optional implementations, the intention prediction module 902 is further configured to:

[0209] Based on the historical trajectory information, establishing an acceleration prediction model for the obstacle vehicle;

[0210] Determining the acceleration of the current trajectory point of the obstacle vehicle through the acceleration prediction model;

[0211] Based on the acceleration of the current trajectory point of the obstructing vehicle and the acceleration or deceleration probability of the previous trajectory point of the obstructing vehicle, the acceleration or deceleration probability of the current trajectory point of the obstructing vehicle is determined as the driving intention of the obstructing vehicle to overtake or yield in the vehicle conflict area.

[0212] In some optional implementations, the model building submodule is further configured to:

[0213] Obtaining an objective function of the autonomous driving vehicle based on the planned trajectory information and the acceleration or deceleration probability of the current trajectory point of the obstacle vehicle;

[0214] Determining an objective function of the obstacle vehicle based on the predicted trajectory information;

[0215] Based on the objective function of the autonomous driving vehicle and the objective function of the obstacle vehicle, the objective function of the two-level optimization game model is obtained. In some optional embodiments, the model building submodule is further configured to:

[0216] Determining a cost function for the autonomous driving vehicle based on the planned trajectory information;

[0217] Obtaining an objective function of the autonomous driving vehicle based on the cost function of the autonomous driving vehicle and the probability of acceleration or deceleration of the obstacle vehicle's current trajectory point; and / or

[0218] Based on the predicted trajectory information, a cost function of the obstacle vehicle is determined as an objective function of the obstacle vehicle.

[0219] In some optional implementations, the model building submodule is further configured to:

[0220] Based on the planned trajectory information, determining a state sequence and an acceleration sequence of trajectory points of the autonomous driving vehicle, wherein the state of each trajectory point in the state sequence includes posture information, velocity information, and acceleration information;

[0221] Determining a first state change cost of a sequence of states of the autonomous vehicle's trajectory points relative to a target state of the autonomous vehicle, and a first smooth transition cost of a sequence of accelerations of the autonomous vehicle's trajectory points, wherein the cost function of the autonomous vehicle includes the first state change cost and the first smooth transition cost; and / or

[0222] Based on the predicted trajectory information, determining a state sequence and a jerk sequence of the trajectory points of the obstacle vehicle, wherein the state of each trajectory point in the state sequence includes posture information, velocity information, and acceleration information;

[0223] Determine a second state change cost of a state sequence of the obstacle vehicle's trajectory points relative to a target state of the obstacle vehicle, and a second smooth transition cost of an acceleration sequence of the obstacle vehicle's trajectory points, wherein the cost function of the obstacle vehicle includes the second state change cost and the second smooth transition cost.

[0224] In some optional implementations, the model building submodule is further configured to:

[0225] Based on the cost function of the autonomous driving vehicle, obtaining an acceleration cost function and a deceleration cost function of the autonomous driving vehicle;

[0226] Determining the weights of the acceleration cost function and the deceleration cost function based on the acceleration or deceleration probability of the current trajectory point of the obstacle vehicle;

[0227] The acceleration cost function and the deceleration cost function are weighted and summed to obtain the objective function of the autonomous driving vehicle.

[0228] In some optional implementations, the conflict detection module 901 is further configured to:

[0229] Based on the planned trajectory information and the predicted trajectory information, determining an overlapping area between the planned trajectory and the predicted trajectory at the same time as a vehicle conflict area;

[0230] A boundary of the vehicle conflict area is determined based on the projection of the overlapping area on the planned trajectory and the predicted trajectory.

[0231] In some optional implementations, the model building submodule is further configured to:

[0232] Determining equality constraints of the two-level optimization game model based on the planned trajectory information and the predicted trajectory information;

[0233] Based on the planned trajectory information, the predicted trajectory information and the boundary of the vehicle conflict area, an inequality constraint condition of the two-level optimization game model is determined.

[0234] In some optional implementations, the model building submodule is further configured to:

[0235] Obtaining an initial state constraint of the autonomous driving vehicle based on a state of a first trajectory point in the planned trajectory information being an initial state of the autonomous driving vehicle;

[0236] Obtaining an initial state constraint of the obstacle vehicle based on the state of the first trajectory point in the predicted trajectory information being the initial state of the obstacle vehicle;

[0237] Obtaining a state transition constraint of the autonomous driving vehicle based on states and jerk of adjacent trajectory points in the planned trajectory information satisfying a transfer matrix;

[0238] Obtaining a state transition constraint of the obstacle vehicle based on states and jerk of adjacent trajectory points in the predicted trajectory information satisfying the transfer matrix;

[0239] Based on the fact that the jerk of the first trajectory point in the planned trajectory information is equal to the acceleration and deceleration of the obstacle vehicle, a contingency planning constraint of the autonomous driving vehicle is obtained.

[0240] In some optional implementations, the model building submodule is further configured to:

[0241] Obtaining a state constraint of the autonomous driving vehicle based on a state of each trajectory point in the planned trajectory information being between an upper boundary state and a lower boundary state of the autonomous driving vehicle;

[0242] Obtaining a state constraint of the obstacle vehicle based on a state of each trajectory point in the predicted trajectory information being between an upper boundary state and a lower boundary state of the obstacle vehicle;

[0243] Obtaining a reversing constraint for the autonomous driving vehicle based on a position of a subsequent trajectory point of adjacent trajectory points in the planned trajectory information being no less than a position of a previous trajectory point;

[0244] Obtaining a reversing constraint for the obstacle vehicle based on a position of a subsequent trajectory point of adjacent trajectory points in the predicted trajectory information being no less than a position of a previous trajectory point;

[0245] A collision constraint is obtained based on that the sum of the distance between the trajectory point of the planned trajectory information and the boundary of the vehicle conflict area and the distance between the trajectory point of the predicted trajectory information and the boundary of the vehicle conflict area at the same time is not less than a preset safety distance.

[0246] In some optional implementations, the model building submodule is further configured to:

[0247] determining a first geometric distance based on a front boundary and a rear boundary of the vehicle conflict area on the planned trajectory and a position of a current trajectory point in the planned trajectory information;

[0248] determining a second geometric distance based on a front boundary and a rear boundary of the vehicle conflict area on the predicted trajectory and a position of a current trajectory point in the predicted trajectory information;

[0249] The collision constraint is obtained based on that the sum of the first geometric distance and the second geometric distance is not less than the preset safety distance.

[0250] In some optional implementations, the model solving submodule is further configured to:

[0251] Converting the objective function of the obstacle vehicle into equality constraints and inequality constraints to obtain new equality constraints and new inequality constraints;

[0252] Based on the new equality constraint, the new inequality constraint, the equality constraint of the two-level optimization game model, and the inequality constraint of the two-level optimization game model, the objective function of the autonomous driving vehicle is solved to obtain the optimized solution.

[0253] In addition, an embodiment of the present disclosure further provides an electronic device, which includes at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed 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 above-mentioned speed planning method 100.

[0254] The embodiment of the present disclosure further provides a computer-readable storage medium storing a computer program, which implements the above-mentioned speed planning method 100 when executed by a processor.

[0255] The embodiment of the present disclosure further provides a computer program product, including a computer program, which implements the above-mentioned speed planning method 100 when executed by a processor.

[0256] Figure 10A block diagram of an example electronic device 1000 that can be used to implement embodiments of the present disclosure is shown. Electronic device 1000 is intended to represent various forms of digital computers. Electronic device 1000 can also represent various forms of mobile devices capable of running computing programs. The components shown herein, their connections and relationships, and their functions are provided for example purposes only and are not intended to limit the implementation of the present disclosure as described and / or claimed herein.

[0257] like Figure 10 As shown, electronic device 1000 includes a computing unit 1010, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 1020 or a computer program loaded from a storage unit 1080 into a random access memory (RAM) 1030. Various programs and data required for the operation of electronic device 1000 may also be stored in RAM 1030. Computing unit 1010, ROM 1020, and RAM 1030 are connected to each other via a bus 1040. An input / output (I / O) interface 1050 is also connected to bus 1040.

[0258] Multiple components in the electronic device 1000 are connected to the I / O interface 1050, including an input unit 1060, such as a touch screen, an output unit 1070, such as various types of displays, speakers, etc., a storage unit 1080, such as a disk, etc., and a communication unit 1090, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1090 allows the electronic device 1000 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0259] The computing unit 1010 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the computing unit 1010 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized 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 1010 performs the various methods and processes described above, such as the speed planning method 100. For example, in some embodiments, the speed planning method 100 can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 1080. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 1000 via the ROM 1020 and / or the communication unit 1090. When the computer program is loaded into the RAM 1030 and executed by the computing unit 1010, one or more steps of the speed planning method 100 described above can be performed. Alternatively, in other embodiments, the computing unit 1010 may be configured to execute the speed planning method 100 in any other appropriate manner (eg, by means of firmware).

[0260] Various embodiments of the systems and techniques described above 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), 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 are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0261] The program code for implementing the method of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0262] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of machine-readable storage media may include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), optical fibers, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0263] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved. This is not a limitation herein.

[0264] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

Claims

1. A speed planning method, characterized in that: include: Determine the vehicle conflict area based on the planned trajectory information of the autonomous vehicle and the predicted trajectory information of the obstacle vehicle; Determining, based on the historical trajectory information of the obstructing vehicle, a probability of the obstructing vehicle accelerating or decelerating with respect to the vehicle conflict area, wherein the acceleration probability indicates the obstructing vehicle's driving intention to overtake the vehicle in the vehicle conflict area, and the deceleration probability indicates the obstructing vehicle's driving intention to yield the vehicle in the vehicle conflict area; Based on the planned trajectory information, the predicted trajectory information, and the acceleration or deceleration probability, establishing an objective function of a two-level optimization game model between the autonomous driving vehicle and the obstacle vehicle, wherein the objective function includes an objective function of the autonomous driving vehicle and an objective function of the obstacle vehicle, the objective function of the autonomous driving vehicle includes an acceleration cost function and a deceleration cost function, and the acceleration or deceleration probability determines the weights of the acceleration cost function and the deceleration cost function in the objective function; Determining constraints of the two-level optimization game model based on the planned trajectory information, the predicted trajectory information, and the vehicle conflict area, wherein the constraints include a contingency planning constraint of the autonomous driving vehicle, the contingency planning constraint being obtained based on the jerk of the first trajectory point in the planned trajectory information being equal to the acceleration and deceleration of the obstacle vehicle; The objective function is solved based on the constraint conditions to obtain an optimized solution of the two-level optimization game model, and the driving speed of the autonomous driving vehicle is adjusted based on the optimized solution.

2. The method according to claim 1, characterized in that The determining, based on the historical trajectory information of the obstacle vehicle, the driving intention of the obstacle vehicle with respect to the vehicle conflict area includes: Based on the historical trajectory information, establishing an acceleration prediction model for the obstacle vehicle; Determining the acceleration of the current trajectory point of the obstacle vehicle through the acceleration prediction model; The acceleration or deceleration probability of the current trajectory point of the obstructing vehicle is determined based on the acceleration of the current trajectory point of the obstructing vehicle and the acceleration or deceleration probability of the previous trajectory point of the obstructing vehicle.

3. The method according to claim 2, characterized in that The objective function of the two-level optimization game model between the autonomous driving vehicle and the obstacle vehicle is established based on the planned trajectory information, the predicted trajectory information, and the acceleration or deceleration probability, including: Determining a cost function for the autonomous driving vehicle based on the planned trajectory information; Based on the cost function of the autonomous driving vehicle, obtaining an acceleration cost function and a deceleration cost function of the autonomous driving vehicle; Determining the weights of the acceleration cost function and the deceleration cost function based on the acceleration or deceleration probability of the current trajectory point of the obstacle vehicle; The acceleration cost function and the deceleration cost function are weighted and summed to obtain the objective function of the autonomous driving vehicle.

4. The method according to claim 3, characterized in that Determining a cost function of the autonomous driving vehicle based on the planned trajectory information includes: Based on the planned trajectory information, determining a state sequence and an acceleration sequence of trajectory points of the autonomous driving vehicle, wherein the state of each trajectory point in the state sequence includes posture information, velocity information, and acceleration information; Determine a first state change cost of a state sequence of trajectory points of the autonomous driving vehicle relative to a target state of the autonomous driving vehicle, and a first smooth transition cost of an acceleration sequence of the trajectory points of the autonomous driving vehicle, wherein the cost function of the autonomous driving vehicle includes the first state change cost and the first smooth transition cost.

5. The method according to claim 3, characterized in that The objective function of establishing a two-level optimization game model between the autonomous driving vehicle and the obstacle vehicle based on the planned trajectory information, the predicted trajectory information, and the acceleration or deceleration probability further includes: Based on the predicted trajectory information, a cost function of the obstacle vehicle is determined as an objective function of the obstacle vehicle.

6. The method according to claim 5, characterized in that The step of determining a cost function of the obstacle vehicle based on the predicted trajectory information includes: Based on the predicted trajectory information, determining a state sequence and a jerk sequence of the trajectory points of the obstacle vehicle, wherein the state of each trajectory point in the state sequence includes posture information, velocity information, and acceleration information; Determine a second state change cost of a state sequence of the obstacle vehicle's trajectory points relative to a target state of the obstacle vehicle, and a second smooth transition cost of an acceleration sequence of the obstacle vehicle's trajectory points, wherein the cost function of the obstacle vehicle includes the second state change cost and the second smooth transition cost.

7. The method according to any one of claims 3 to 6, characterized in that: The determining of the vehicle conflict area based on the planned trajectory information of the autonomous driving vehicle and the predicted trajectory information of the obstacle vehicle includes: Based on the planned trajectory information and the predicted trajectory information, determining an overlapping area between the planned trajectory and the predicted trajectory at the same time as a vehicle conflict area; A boundary of the vehicle conflict area is determined based on the projection of the overlapping area on the planned trajectory and the predicted trajectory.

8. The method according to claim 7, characterized in that The determining of the constraint conditions of the two-level optimization game model based on the planned trajectory information, the predicted trajectory information, and the vehicle conflict area includes: Determining, based on the planned trajectory information and the predicted trajectory information, an equality constraint condition of the two-level optimization game model, wherein the equality constraint condition includes the contingency planning constraint; Based on the planned trajectory information, the predicted trajectory information and the boundary of the vehicle conflict area, an inequality constraint condition of the two-level optimization game model is determined.

9. The method according to claim 8, characterized in that The determining of the equality constraint conditions of the two-level optimization game model based on the planned trajectory information and the predicted trajectory information further includes: Obtaining an initial state constraint of the autonomous driving vehicle based on a state of a first trajectory point in the planned trajectory information being an initial state of the autonomous driving vehicle; Obtaining an initial state constraint of the obstacle vehicle based on the state of the first trajectory point in the predicted trajectory information being the initial state of the obstacle vehicle; Obtaining a state transition constraint of the autonomous driving vehicle based on states and jerk of adjacent trajectory points in the planned trajectory information satisfying a transfer matrix; Based on the states and jerk of adjacent trajectory points in the predicted trajectory information satisfying the transfer matrix, a state transfer constraint of the obstacle vehicle is obtained.

10. The method according to claim 8, characterized in that The determining of the inequality constraints of the two-level optimization game model based on the planned trajectory information, the predicted trajectory information, and the boundary of the vehicle conflict area includes: Obtaining a state constraint of the autonomous driving vehicle based on a state of each trajectory point in the planned trajectory information being between an upper boundary state and a lower boundary state of the autonomous driving vehicle; Obtaining a state constraint of the obstacle vehicle based on a state of each trajectory point in the predicted trajectory information being between an upper boundary state and a lower boundary state of the obstacle vehicle; Obtaining a reversing constraint for the autonomous driving vehicle based on a position of a subsequent trajectory point of adjacent trajectory points in the planned trajectory information being no less than a position of a previous trajectory point; Obtaining a reversing constraint for the obstacle vehicle based on a position of a subsequent trajectory point of adjacent trajectory points in the predicted trajectory information being no less than a position of a previous trajectory point; A collision constraint is obtained based on that the sum of the distance between the trajectory point of the planned trajectory information and the boundary of the vehicle conflict area and the distance between the trajectory point of the predicted trajectory information and the boundary of the vehicle conflict area at the same time is not less than a preset safety distance.

11. The method according to claim 10, characterized in that The obtaining of the collision constraint based on the sum of the distance between the trajectory point of the planned trajectory information and the boundary of the vehicle conflict area and the distance between the trajectory point of the predicted trajectory information and the boundary of the vehicle conflict area at the same time being greater than a preset safety distance includes: determining a first geometric distance based on a front boundary and a rear boundary of the vehicle conflict area on the planned trajectory and a position of a current trajectory point in the planned trajectory information; determining a second geometric distance based on a front boundary and a rear boundary of the vehicle conflict area on the predicted trajectory and a position of a current trajectory point in the predicted trajectory information; The collision constraint is obtained based on that the sum of the first geometric distance and the second geometric distance is not less than the preset safety distance.

12. The method according to any one of claims 8 to 11, characterized in that Solving the objective function based on the constraint conditions to obtain an optimized solution to the two-level optimization game model includes: Converting the objective function of the obstacle vehicle into equality constraints and inequality constraints to obtain new equality constraints and new inequality constraints; Based on the new equality constraint, the new inequality constraint, the equality constraint of the two-level optimization game model, and the inequality constraint of the two-level optimization game model, the objective function of the autonomous driving vehicle is solved to obtain the optimized solution.

13. A speed planning device, characterized in that: include: a conflict detection module configured to determine a vehicle conflict area based on the planned trajectory information of the autonomous driving vehicle and the predicted trajectory information of the obstacle vehicle; an intention prediction module configured to determine, based on historical trajectory information of the obstructing vehicle, a probability of acceleration or deceleration of the obstructing vehicle with respect to the vehicle conflict area, wherein the acceleration probability indicates the obstructing vehicle's intention to overtake the vehicle in the vehicle conflict area, and the deceleration probability indicates the obstructing vehicle's intention to yield the vehicle in the vehicle conflict area; Speed planning module, including: a model building submodule configured to establish an objective function of a two-level optimization game model between the autonomous driving vehicle and the obstacle vehicle based on the planned trajectory information, the predicted trajectory information, and the acceleration or deceleration probability, wherein the objective function includes an objective function of the autonomous driving vehicle and an objective function of the obstacle vehicle, the objective function of the autonomous driving vehicle includes an acceleration cost function and a deceleration cost function, and the acceleration or deceleration probability determines the weights of the acceleration cost function and the deceleration cost function in the objective function; and determine constraints of the two-level optimization game model based on the planned trajectory information, the predicted trajectory information, and the vehicle conflict area, wherein the constraints include a contingency planning constraint of the autonomous driving vehicle, and the contingency planning constraint is obtained based on the jerk of the first trajectory point in the planned trajectory information being equal to the acceleration and deceleration of the obstacle vehicle; The model solving submodule is configured to solve the objective function based on the constraint conditions to obtain an optimized solution of the two-level optimization game model, and adjust the driving speed of the autonomous driving vehicle based on the optimized solution.

14. An electronic device, characterized in that: include: 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, the instructions being executed by the at least one processor so as to enable the at least one processor to execute the speed planning method described in any one of claims 1 to 12.

15. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the speed planning method as claimed in any one of claims 1 to 12 is implemented.

16. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the speed planning method as claimed in any one of claims 1 to 12 is implemented.

Images