Interactive decision planning method and device for autonomous vehicle

Abstract

The embodiment of the invention discloses an interactive decision planning method and device for an autonomous vehicle. The specific implementation mode of the method comprises the steps that in response to the detected lane changing intention of a vehicle, path planning is conducted according to at least one of driving state information of the vehicle, driving state information of an interactive vehicle and road environment information, and a target planning path of the vehicle is obtained; calculating a conflict area between the own vehicle and the interactive vehicle based on the target planned path and the driving state information of the interactive vehicle; and performing speed planning based on the conflict area to obtain a speed planning result of the own vehicle and a speed planning result of the interactive vehicle. According to the embodiment, path planning and speed planning are decoupled, and the automatic driving safety is improved.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of autonomous driving technology, and more particularly, to an interactive decision-making and planning method and apparatus for autonomous vehicles. Background Art

[0002] Currently, when autonomous vehicles face interactive scenarios, how to plan effective driving behaviors remains an unsolved research topic. Currently, the mainstream interactive processing technologies include vehicle-to-vehicle collaborative planning and forward deduction planning. Vehicle-to-vehicle collaborative planning depends on infrastructure construction and policies and regulations. It is expected that in the next period of time, autonomous driving technology will still mainly focus on the direction of single-vehicle intelligence. Forward deduction planning predicts the running trajectories of the host vehicle and the interactive vehicle through a closed-loop simulation process and selects the planning trajectory with the highest benefit. This method has strong robustness and is relatively easy to implement, but its ability to estimate the traffic environment of autonomous vehicles is limited, which may lead to overly conservative decision-making results. Summary of the Invention

[0003] Embodiments of the present disclosure propose an interactive decision-making and planning method and apparatus for autonomous vehicles.

[0004] In a first aspect, an embodiment of the present disclosure provides an interactive decision-making and planning method for an autonomous vehicle, including: in response to detecting an intention of the host vehicle to change lanes, performing path planning based on at least one of the driving state information of the host vehicle, the driving state information of the interactive vehicle, and the road environment information to obtain a target planning path of the host vehicle; calculating a conflict area between the host vehicle and the interactive vehicle based on the target planning path and the driving state information of the interactive vehicle; and performing speed planning based on the conflict area to obtain a speed planning result of the host vehicle and a speed planning result of the interactive vehicle.

[0005] In some embodiments, the road environment information includes the road width of the target lane; and performing path planning based on at least one of the driving state information of the host vehicle, the driving state information of the interactive vehicle, and the environment information to obtain a target planning path of the host vehicle includes: determining whether there is a lateral movement space for the host vehicle according to the road width; and in response to determining that there is no lateral movement space for the host vehicle, using the center line of the lane during the lane change process as the target planning path of the host vehicle.

[0006] In some embodiments, the road environment information includes the road width of the target lane; and path planning is performed based on at least one of the driving state information of the host vehicle, the driving state information of the interacting vehicle, and the environment information to obtain an initial planned path of the host vehicle, including: determining whether there is a lateral movement space for the host vehicle according to the road width; in response to determining that there is a lateral movement space for the host vehicle, detecting the congestion condition of the road ahead and the distance between the interacting vehicle and the host vehicle; in response to detecting that some lanes of the road ahead are congested and the distance is greater than a predetermined distance threshold, using the non-congested lane as the target planned path of the host vehicle.

[0007] In some embodiments, the method further includes: in response to detecting that some lanes of the road ahead are congested and the distance is less than or equal to the predetermined distance threshold, determining a connection node of the planned path of the host vehicle according to the position of the vehicle ahead in the congested lane; determining the path from the current position of the host vehicle to the connection node as the first path, and determining the path from the connection node to the non-congested lane as the second path, where the first path and the second path are combined as the target planned path of the host vehicle.

[0008] In some embodiments, path planning is performed based on at least one of the driving state information of the host vehicle, the driving state information of the interacting vehicle, and the road environment information to obtain a target planned path of the host vehicle, including: performing path planning based on at least one of the driving state information of the host vehicle, the driving state information of the interacting vehicle, and the road environment information to obtain an initial planned path of the host vehicle; optimizing the initial planned path based on the constraint conditions of path planning to obtain the target planned path of the host vehicle.

[0009] In some embodiments, the constraint conditions include: continuity constraint conditions, smoothness constraint conditions, and boundary constraint conditions; and optimizing the initial planned path based on the constraint conditions of path planning to obtain the target planned path of the host vehicle, including: performing curve fitting on the initial planned path to obtain a connection function; constructing continuity constraint conditions based on the second derivative of the connection function; constructing boundary constraint conditions based on the corner positions of the host vehicle; constructing smoothness constraint conditions based on the first derivative of the connection function; solving the connection function based on the constraint conditions to obtain a target planned path that satisfies at least one of the following optimization objectives: the distance between the target planned path and the lane reference line is the closest, the number of turning points of the target planned path is the least, and the gap between the target planned path and the initial planned path is the smallest.

[0010] In some embodiments, speed planning is performed based on a conflict area to obtain the speed planning result of the host vehicle and the speed planning result of the interacting vehicle, including: respectively constructing an optimal control problem for the host vehicle and an optimal control problem for the interacting vehicle based on the conflict area, and fusing the optimal control problem for the interacting vehicle and the optimal control problem for the host vehicle into a single optimal control problem; solving the single optimal control problem to obtain the speed planning result of the host vehicle and the speed planning result of the interacting vehicle.

[0011] In some embodiments, an ego-vehicle optimal control problem and an interactive vehicle optimal control problem are respectively constructed based on a conflict area, including: constructing an objective function of the ego-vehicle based on the distance, speed, and acceleration of the ego-vehicle along a target planned path; constructing an objective function of the interactive vehicle based on the distance, speed, and acceleration of the interactive vehicle along the driving direction; constructing an interactive vehicle optimal control problem based on inequality constraint conditions, the objective function of the interactive vehicle, and the equality constraint conditions of the interactive vehicle; and constructing an ego-vehicle optimal control problem based on inequality constraint conditions, the objective function of the ego-vehicle, and the equality constraint conditions of the ego-vehicle.

[0012] In some embodiments, the equality constraint conditions include that the second derivative of the objective function is continuous, and the inequality constraint conditions include that the sum of the conflict geometric distances of the ego-vehicle and the interactive vehicle is greater than or equal to zero, where the conflict geometric distance is the product of a first geometric distance and a second geometric distance, the first geometric distance is the difference between the front boundary of the conflict area and the vehicle position, and the second geometric distance is the difference between the rear boundary of the conflict area and the vehicle position.

[0013] In some embodiments, the method further includes: correcting the conflict geometric distance by adjusting a parameter, where the adjusting parameter is proportional to the sum of the conflict area size and the vehicle length.

[0014] In some embodiments, the objective function includes an objective cost, a second-order control quantity cost, and a third-order control quantity cost. The objective cost is obtained by the following method: obtaining the following-following state functions of the ego-vehicle and the interactive vehicle based on an intelligent driving model; respectively solving the following-following state functions of the ego-vehicle and the interactive vehicle based on the anti-rollover constraint condition and the anti-skid constraint condition to obtain the target state of the ego-vehicle and the target state of the interactive vehicle; and respectively constructing the objective cost of the ego-vehicle and the objective cost of the interactive vehicle based on the target state of the ego-vehicle and the target state of the interactive vehicle.

[0015] In a second aspect, an embodiment of the present disclosure provides an interactive decision-making and planning device for an autonomous vehicle, including: a path planning unit configured to perform path planning based on at least one of the driving state information of the ego-vehicle, the driving state information of the interactive vehicle, and the road environment information in response to detecting an intention of the ego-vehicle to change lanes, so as to obtain a target planned path of the ego-vehicle; a conflict calculation unit configured to calculate a conflict area between the ego-vehicle and the interactive vehicle based on the target planned path and the driving state information of the interactive vehicle; and a speed planning unit configured to perform speed planning based on the conflict area to obtain a speed planning result of the ego-vehicle and a speed planning result of the interactive vehicle.

[0016] In some embodiments, the road environment information includes the road width of the target lane; and the path planning unit is further configured to: determine whether there is a lateral movement space for the host vehicle according to the road width; in response to determining that there is no lateral movement space for the host vehicle, use the lane center line during the lane change process as the target planned path of the host vehicle.

[0017] In some embodiments, the road environment information includes the road width of the target lane; and the path planning unit is further configured to: determine whether there is a lateral movement space for the host vehicle according to the road width; in response to determining that there is a lateral movement space for the host vehicle, detect the congestion condition of the road ahead and the distance between the interacting vehicle and the host vehicle; in response to detecting that some lanes of the road ahead are congested and the distance is greater than a predetermined distance threshold, use the non-congested lane as the target planned path of the host vehicle.

[0018] In some embodiments, the path planning unit is further configured to: in response to detecting that some lanes of the road ahead are congested and the distance is less than or equal to the predetermined distance threshold, determine the connection node of the planned path of the host vehicle according to the position of the vehicle ahead in the congested lane; determine the path from the current position of the host vehicle to the connection node as the first path, determine the path from the connection node to the non-congested lane as the second path, and combine the first path and the second path as the target planned path of the host vehicle.

[0019] In some embodiments, the device further includes a path optimization unit, configured to perform path planning according to at least one of the driving state information of the host vehicle, the driving state information of the interacting vehicle, and the road environment information to obtain the initial planned path of the host vehicle; optimize the initial planned path based on the constraint conditions of the path planning to obtain the target planned path of the host vehicle.

[0020] In some embodiments, the constraint conditions include: continuity constraint conditions, smoothness constraint conditions, and boundary constraint conditions; and the path optimization unit is further configured to: perform curve fitting on the initial planned path to obtain a connection function; construct continuity constraint conditions based on the second derivative of the connection function; construct boundary constraint conditions based on the corner positions of the host vehicle; construct smoothness constraint conditions based on the first derivative of the connection function; solve the connection function based on the constraint conditions to obtain a target planned path that satisfies at least one of the following optimization objectives: the distance between the target planned path and the lane reference line is the closest, the number of turning points of the target planned path is the least, and the gap between the target planned path and the initial planned path is the smallest.

[0021] In some embodiments, the speed planning unit is further configured to: respectively construct an optimal control problem for the host vehicle and an optimal control problem for the interacting vehicle based on the conflict area, and fuse the optimal control problem for the interacting vehicle and the optimal control problem for the host vehicle into a single optimal control problem based on the KKT conditions; solve the single optimal control problem to obtain the speed planning result of the host vehicle and the speed planning result of the interacting vehicle.

[0022] In some embodiments, the speed planning unit is further configured to: construct an objective function for the host vehicle based on the distance, speed, and acceleration of the host vehicle along the target planned path; construct an objective function for the interacting vehicle based on the distance, speed, and acceleration of the interacting vehicle along the driving direction; construct an optimal control problem for the interacting vehicle based on the inequality constraint conditions, the objective function of the interacting vehicle, and the equality constraint conditions of the interacting vehicle; construct an optimal control problem for the host vehicle based on the inequality constraint conditions, the objective function of the host vehicle, and the equality constraint conditions of the host vehicle.

[0023] In some embodiments, the equality constraint conditions include: the second derivative of the objective function is continuous, and the inequality constraint conditions include: the sum of the conflict geometric distances of the host vehicle and the conflict geometric distances of the interacting vehicle is greater than or equal to zero, where the conflict geometric distance is the product of a first geometric distance and a second geometric distance, where the first geometric distance is the difference between the front boundary of the conflict area and the vehicle position, and the second geometric distance is the difference between the rear boundary of the conflict area and the vehicle position.

[0024] In some embodiments, the speed planning unit is further configured to: correct the conflict geometric distance by adjusting a parameter, where the adjusting parameter is proportional to the sum of the conflict area size and the vehicle length.

[0025] In some embodiments, the objective function includes an objective cost, a second-order control quantity cost, and a third-order control quantity cost, and the objective cost is obtained by the following method: obtain the following-following state function of the host vehicle and the following-following state function of the interacting vehicle based on the intelligent driving model; respectively solve the following-following state function of the host vehicle and the following-following state function of the interacting vehicle based on the anti-rollover constraint condition and the anti-skid constraint condition to obtain the target state of the host vehicle and the target state of the interacting vehicle; respectively construct the objective cost of the host vehicle and the objective cost of the interacting vehicle based on the target state of the host vehicle and the target state of the interacting vehicle.

[0026] In a third aspect, embodiments of the present disclosure provide an electronic device, including: one or more processors; a storage device storing one or more computer programs thereon, and when the one or more computer programs are executed by the one or more processors, the one or more processors are caused to implement the method according to any one of the first aspect.

[0027] Fourthly, an embodiment of the present disclosure provides a computer-readable medium, on which a computer program is stored, wherein when the computer program is executed by a processor, the method according to any one of the first aspect is implemented.

[0028] Fifthly, an embodiment of the present disclosure provides an autonomous vehicle, including the electronic device according to the third aspect.

[0029] The interactive decision-making and path planning method and device for the autonomous vehicle provided by the embodiment of the present disclosure perform path planning by collecting information of the host vehicle, interactive vehicles and environmental information, and obtain an initial planned path of the host vehicle. Calculate the drivable area according to reference line information, road boundaries, static obstacles and other information. Optimize the initial planned path within the drivable area. Calculate the information of the interactive conflict area according to the optimized planned path. Perform speed planning according to the information of the interactive conflict area. A lane-changing decision-making and planning method applicable under a complete path planning and speed planning decoupling framework is proposed to handle the dynamic game interaction requirements during lane-changing in congested scenarios.

[0030] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Other features, objectives and advantages of the present disclosure will become more obvious by reading the detailed description of the non-limiting embodiments with reference to the following drawings:

[0032] Figure 1 is an exemplary system architecture diagram to which an embodiment of the present disclosure can be applied;

[0033] Figure 2 is a flowchart of an embodiment of the interactive decision-making and path planning method for the autonomous vehicle according to the present disclosure;

[0034] Figures 3a - 3f is a schematic diagram of an application scenario of the interactive decision-making and path planning method for the autonomous vehicle according to the present disclosure;

[0035] Figure 4 is a flowchart of another embodiment of the interactive decision-making and path planning method for the autonomous vehicle according to the present disclosure;

[0036] Figure 5 is a schematic structural diagram of an embodiment of the interactive decision-making and path planning device for the autonomous vehicle according to the present disclosure;

[0037] Figure 6 is a schematic structural diagram of a computer system of an electronic device suitable for implementing the embodiments of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0038] The present disclosure will be further described in detail below with reference to the accompanying drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the related invention and are not intended to limit the invention. Additionally, it should be noted that for ease of description, only the parts related to the relevant invention are shown in the drawings.

[0039] It should be noted that, without conflict, the embodiments in the present disclosure and the features in the embodiments may be combined with each other. The present disclosure will be described in detail below with reference to the drawings and embodiments.

[0040] Figure 1 An exemplary system architecture 100 is shown, which shows embodiments of an interactive decision-making and planning method for an autonomous vehicle or an interactive decision-making and planning device for an autonomous vehicle to which the present disclosure can be applied.

[0041] As Figure 1 shown, the system architecture 100 may include a first in-vehicle terminal 101, a second in-vehicle terminal 102, a network 103, and a server 104. The network 103 is used to provide a medium for communication links between the first in-vehicle terminal 101, the second in-vehicle terminal 102, and the server 104. The network 103 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0042] Users can use the first in-vehicle terminal 101 and the second in-vehicle terminal 102 to interact with the server 104 through the network 103 to receive or send messages, etc. Various communication client applications, such as video playback applications, navigation applications, search applications, instant messaging tools, email clients, etc., may be installed on the first in-vehicle terminal 101 and the second in-vehicle terminal 102. The first in-vehicle terminal 101 and the second in-vehicle terminal 102 are also referred to as agents, in-vehicle computers, in-vehicle intelligent devices, intelligent in-vehicle terminals, vehicle dispatching and monitoring terminals, in-vehicle wireless terminals, etc.

[0043] The first in-vehicle terminal 101 and the second in-vehicle terminal 102 may be, for example, the in-vehicle computer systems of vehicles such as autonomous vehicles and delivery robots. This system can be implemented in a hardware manner, or in a software manner, or in a manner combining hardware and software.

[0044] The server 104 can provide various services. For example, the server 104 can perform path planning based on the driving state information of the host vehicle, the driving state information of the interactive vehicle, and the road environment information, and then perform speed planning. The speed planning results of the host vehicle and the interactive vehicle are fed back to the host vehicle and the interactive vehicle, thereby ensuring the safety of autonomous driving.

[0045] It should be noted that the server 104 can be hardware or software. When the server 104 is hardware, it can be implemented as a distributed server cluster composed of multiple servers or as a single server. When the server 104 is software, it can be implemented as multiple software or software modules (for example, used to provide distributed services) or as a single software or software module. No specific limitation is made here.

[0046] It should be pointed out that the execution subject (hereinafter simply referred to as the "execution subject") of the interactive decision-making and planning method for the autonomous driving vehicle provided by the present disclosure can be executed by the server in the above system architecture or can be implemented through the above vehicle-mounted terminal. (Supplement, for example, when the server executes, the control command is sent to the vehicle-mounted terminal, such as the in-vehicle system of the vehicle, and then the vehicle is controlled according to the control command through the in-vehicle system.

[0047] In addition, another applicable scenario is to directly complete the speed planning by the in-vehicle system without passing through the server and control the vehicle).

[0048] It should be understood, Figure 1 the numbers of the first vehicle-mounted terminal, the second vehicle-mounted terminal, the network and the server in [[ ]] are only illustrative. According to the implementation requirements, there can be any number of the first vehicle-mounted terminal, the second vehicle-mounted terminal, the network and the server.

[0049] Continuing to refer to Figure 2 , a flow 200 of an embodiment of the interactive decision-making and planning method for the autonomous driving vehicle according to the present disclosure is shown. The interactive decision-making and planning method for the autonomous driving vehicle includes the following steps:

[0050] Step 201, in response to detecting the intention of the host vehicle to change lanes, perform path planning based on one of the driving state information of the host vehicle, the driving state information of the interactive vehicle, and the road environment information to obtain the target planned path of the host vehicle.

[0051] In this embodiment, the execution subject can determine whether the host vehicle has the intention to change lanes through the navigation information of the host vehicle. For example, if it is detected that there is a T-junction ahead of the vehicle, it is determined that the host vehicle has the intention to change lanes. The interactive vehicle refers to other vehicles that may collide with the host vehicle near the planned route. The driving state information may include the pose, speed, and acceleration of the vehicle. The road environment information may include the road width, static obstacle information, etc.

[0052] The planned path refers to the driving route of the host vehicle that avoids the interactive vehicle and static obstacles and reaches the destination. The planned path can be a rough plan only based on the road environment information and can be further optimized later.

[0053] Optionally, the result of step 201 is used as the initial planned path, and the initial planned path is optimized based on the constraint conditions of path planning to obtain the target planned path of the host vehicle.

[0054] In this embodiment, the initial planned path is a set of discrete trajectory points, which can be fitted into a curve through a connection function. During the curve fitting process, some constraint conditions can be set to make the fitted curve smoother, so that the vehicle runs more smoothly during driving and the riding experience of passengers is better.

[0055] The constraint conditions may include at least one of the following: continuity constraint condition, smoothness constraint condition, and boundary constraint condition.

[0056] The continuity constraint condition can be realized by limiting the second derivative of the connection function to be continuous.

[0057] The smoothness constraint condition can be realized by limiting the sum of squares of the first derivatives of the connection function to be the smallest, that is, the total amount of change in the position of the trajectory points is the smallest.

[0058] The boundary constraint condition can be realized by limiting the corner position of the vehicle within the drivable area (i.e., convex space).

[0059] The discrete trajectory points of the initial planned path are adjusted in position so that the curve fitted by the trajectory points after position adjustment conforms to the constraint conditions of path planning, and then the optimized target planned path of the host vehicle is obtained.

[0060] Step 202, based on the target planned path and the driving state information of the interacting vehicle, calculate the conflict area between the host vehicle and the interacting vehicle.

[0061] In this embodiment, based on the driving state information of the interacting vehicle, the planned path of the interacting vehicle is obtained. Then, based on the target planned path of the host vehicle and the planned path of the interacting vehicle, the path overlapping area between the host vehicle and the interacting vehicle is obtained. Based on the path overlapping area between the host vehicle and the interacting vehicle, the conflict area is calculated.

[0062] Both the interacting vehicle and the host vehicle can pass through the conflict area before the other, or choose to wait and let the other pass through the conflict area first. The interacting vehicle and the host vehicle cannot be in the conflict area at the same time.

[0063] Step 203, based on the conflict area, perform speed planning to obtain the speed planning result of the host vehicle and the speed planning result of the interacting vehicle.

[0064] In this embodiment, variables to be optimized are determined, such as acceleration, speed, etc., and an optimization objective function is defined. Generally, the purpose of the objective function is to make the acceleration small, the distance from obstacles far, and meet the vehicle's acceleration and deceleration requirements. In the lane-changing scenario, the constraint conditions for speed planning need to consider the centrifugal force and turning radius of the vehicle, calculate a suitable speed curve, and ensure that the vehicle can pass through the curve smoothly.

[0065] Optionally, speed planning can construct a Stackelberg game model, initiate a game with the ego vehicle as the leader, and the interacting vehicle as the follower to respond to the ego vehicle. The system state at time t is representing the system state of the ego vehicle at time t, representing the system state of the interacting vehicle at time t. x represents all system states. The objective functions of the ego vehicle and the interacting vehicle are respectively denoted as J L and J F . Under the decoupling framework, the planned trajectory is planned by simple kinematics. The vehicle's planned trajectory needs to satisfy the passing constraints.

[0066] In the leader-follower game, the follower formulates a response trajectory based on the possible planned trajectories within the future planning horizon of the leader.

[0067] Similarly, the leader will consider the possible response strategies of the follower to formulate its own future trajectory.

[0068] The method provided in the above embodiment of the present disclosure proposes a complete lane-changing decision-making and planning method applicable under the decoupling framework of path planning and speed planning, which can avoid collisions between the ego vehicle and the interacting vehicle and improve the driving safety of the ego vehicle in different scenarios. It can handle the dynamic game interaction requirements during lane-changing in congested scenarios.

[0069] In some optional implementation manners of this embodiment, the road environment information includes the road width of the target lane; and path planning is performed according to the driving state information of the ego vehicle, the driving state information of the interacting vehicle, and the environment information to obtain the initial planned path of the ego vehicle, including: determining whether there is a lateral movement space for the ego vehicle according to the road width; in response to determining that there is no lateral movement space for the ego vehicle, using the lane center line during the lane-changing process as the initial planned path of the ego vehicle.

[0070] When the road is narrow, there is no lateral offset space for the ego vehicle to merge into the target lane. Therefore, the initial planned path can be along the lane center as much as possible, as shown in Figure 3a .

[0071] In some alternative implementation manners of this embodiment, the road environment information includes the road width of the target lane; and path planning is performed according to the driving state information of the host vehicle, the driving state information of the interacting vehicle, and the environment information to obtain the initial planned path of the host vehicle, including: determining whether there is a lateral movement space for the host vehicle according to the road width; in response to determining that there is a lateral movement space for the host vehicle, detecting the congestion condition of the road ahead and the distance between the interacting vehicle and the host vehicle; in response to detecting that some lanes of the road ahead are congested and the distance is greater than a predetermined distance threshold, using the non-congested lane as the initial planned path of the host vehicle.

[0072] When the front part of the road is congested and the vehicle behind is far away, the initial planned path of the host vehicle should seek to quickly merge into a non-congested lane. For example, Figure 3b As shown, there is an interacting vehicle 1 blocking the front in the right lane, so the left lane is selected for merging.

[0073] In some alternative implementation manners of this embodiment, the method further includes: in response to detecting that some lanes of the road ahead are congested and the distance is less than or equal to a predetermined distance threshold, determining the connection node of the planned path of the host vehicle according to the position of the vehicle ahead in the congested lane; determining the path from the current position of the host vehicle to the connection node as the first path, determining the path from the connection node to the non-congested lane as the second path, and combining the first path and the second path as the initial planned path of the host vehicle.

[0074] When the front part of the road is congested and the vehicle behind is close and the host vehicle does not have the condition to quickly merge in, the host vehicle should adopt a two-stage merging method to ensure a sufficiently long deceleration safety path, as Figure 3c shown. It should be noted that the connection node of the two-stage merging needs to be earlier than the latest merging point, and the position of the latest merging point is the position of the rear end of interacting vehicle 2 when interacting vehicle 2 decelerates and queues up behind interacting vehicle 1.

[0075] In some alternative implementation manners of this embodiment, the constraint conditions include: continuity constraint conditions, smoothness constraint conditions, and boundary constraint conditions; and optimizing the initial planned path based on the constraint conditions of path planning to obtain the target planned path of the host vehicle, including: performing curve fitting on the initial planned path to obtain a connection function; constructing continuity constraint conditions based on the second derivative of the connection function; constructing boundary constraint conditions based on the corner positions of the host vehicle; constructing smoothness constraint conditions based on the first derivative of the connection function; solving the connection function based on the constraint conditions to obtain a target planned path that satisfies at least one of the following optimization objectives: the target planned path is closest to the lane reference line, the target planned path has the fewest turning points, and the target planned path has the smallest gap from the initial planned path.

[0076] The planned path can be expressed as l = f(s), where f(s) represents a function related to s, s is the longitudinal position in the SL coordinate system, and l is the lateral position in the SL coordinate system. By optimizing and solving l = f(s), the target planned path is obtained. Among them, the SL coordinate system is composed of the longitudinal distance (the distance along the center line of the road) and the lateral distance (the distance perpendicular to the center line of the road), which helps to accurately express the vehicle position and attitude.

[0077] The optimization result needs to meet at least one of the following constraints:

[0078] (1) The second derivative of f(s) is continuous;

[0079] (2) Boundary constraints, that is i = 1, 2,..., n; where n is the planned frame, a total of n frames are planned. Generally, the planning period of one frame is 100 ms, and the empirical value plans 50 frames, that is, the trajectory in the next 5 s. s i represents the longitudinal position of the i-th frame, represents the minimum value of the lateral position of the i-th frame, represents the maximum value of the lateral position of the i-th frame.

[0080] (3) Ensure the smoothness of f(s), that is represents the sum of the squares of the first derivatives of f(s i ) is the smallest.

[0081] The constraints can be divided into two types: equality constraints and inequality constraints:

[0082] A Equality constraints

[0083] Denote l i = f(s i ), l i ′ = f ′ (s i ), l i ″ = f″(s i ).

[0084] l i ′ represents the first derivative of l i , l i ″ represents the second derivative of l i . Assume that the connecting function f(s) between l i and l i+1 has a third derivative l i ″ ′ as a fixed constant Δs represents the change in the longitudinal position between two adjacent frames, which can ensure the continuity of the second derivative. After a finite number of Taylor expansions, we have

[0085]

[0086] Substitute and transpose to get

[0087]

[0088] That is, let

[0089]

[0090] There is

[0091] A eq [l i l i ′ l i ″l i+1 l i+1 ′ l i+1 ″] T = 0

[0092] Among them, A eq represents the equation matrix, and T represents the matrix transpose.

[0093] In the optimized time domain, there is an equation constraint

[0094]

[0095] Among them, the state X = [l1 l1 ′ l1″l2 l2 ′ l2″...l n l n ′ l n ″] T .

[0096] B inequality constraint

[0097] The inequality constraint includes the boundary constraint conditions formed by the convex space of the vehicle corner points. Let

[0098] a. The distance from the vehicle planning center to the front of the vehicle is d1;

[0099] b. The distance from the vehicle planning center to the rear of the vehicle is d2;

[0100] c. The width of the vehicle is W;

[0101] d. The head orientation angle of the vehicle in the SL coordinate system is θ;

[0102] e. The vehicle geometric corner points are P1, P2, P3, and P4, as Figure 3d shown.

[0103] To handle the non - linear trigonometric function, make the following approximation

[0104] sinθ ≤ tanθ = l ′

[0105] cosθ ≤ 1

[0106] The abscissas of the corner points P1, P2, P3, and P4 are approximately respectively

[0107]

[0108] In the convex space, denote the minimum l value of the Upper bound (upper boundary) in the interval (s i - d2, s i + d1) as The maximum l value of the Lowerbound (lower boundary) is Ensure that the following corner point convex space constraints hold

[0109]

[0110] Let

[0111]

[0112]

[0113] There is

[0114]

[0115] where A neq represents the inequality matrix, represents the boundary matrix.

[0116] In the optimization time domain, the inequality constraints are as follows

[0117]

[0118] The C objective function

[0119] The objective function (i.e., the cost function) may include at least one of the following: a. reference line cost; b. smoothing cost; c. rough solution cost.

[0120] (1) The reference line cost C ref

[0121]

[0122] In the SL coordinate system, the reference line is the coordinate axis. w ref represents the weight of the reference line cost, represents the distance between the trajectory points on the planned path and the reference line. C ref represents the reference line cost, C refThe smaller it is, the closer the planned path is to the center line of the lane, which is one of the goals of path optimization.

[0123] (2) Smoothness cost C smooth

[0124]

[0125] w dl 、w ddl 、w dddl respectively represent ∑ i l i ′ 2 、∑ i l i ″ 2 and ∑ i (l i+1 ″ - l i ″) 2 weights. The smoothness of the planned path is measured from multiple perspectives, making the optimized path smooth, turning as little as possible, and even if turning, with the smallest possible curvature.

[0126] (3) Coarse solution cost C rough

[0127]

[0128] represents the trajectory points of the initial planned path, C rough represents that the optimized planned path is as close as possible to the initial planned path, that is, the moving distance of the trajectory points is as small as possible. w rough represents the weight of the coarse solution cost.

[0129] There is a total cost function C

[0130] C = C ref + C smooth + C rough

[0131] For the application scenario of this application, the coarse solution cost should be much greater than the reference line cost.

[0132] By solving for the state X when C is minimized, the target planned path of the vehicle itself is obtained.

[0133] For further reference Figure 4 , which shows the process 400 of another embodiment of the interactive decision-making planning method for an autonomous vehicle. The process 400 of the interactive decision-making planning method for the autonomous vehicle includes the following steps:

[0134] Step 401, in response to detecting the intention of the host vehicle to change lanes, perform path planning based on the driving state information of the host vehicle, the driving state information of the interacting vehicle, and the road environment information to obtain the initial planned path of the host vehicle.

[0135] Step 402, optimize the initial planned path based on the constraint conditions of path planning to obtain the target planned path of the host vehicle.

[0136] Step 403, calculate the conflict area between the host vehicle and the interacting vehicle based on the target planned path and the driving state information of the interacting vehicle.

[0137] Steps 401 - 403 are basically the same as steps 201 - 203, so they will not be elaborated here.

[0138] Step 404, construct the optimal control problem of the host vehicle and the optimal control problem of the interacting vehicle respectively based on the conflict area.

[0139] In this embodiment, the objective functions of the optimal control problem of the host vehicle and the optimal control problem of the interacting vehicle may both include at least one of the following: minimum acceleration, minimum jerk. The constraint conditions may both include: the two vehicles do not appear in the conflict area at the same time, the jerk is continuous, the speed cannot be too fast to prevent rollover and skidding, etc.

[0140] In some alternative implementation manners of this embodiment, construct the objective function of the host vehicle based on the distance, speed, and acceleration of the host vehicle along the target planned path; construct the objective function of the interacting vehicle based on the distance, speed, and acceleration of the interacting vehicle along the driving direction; construct the optimal control problem of the interacting vehicle based on the inequality constraint conditions, the objective function of the interacting vehicle, and the equality constraint conditions of the interacting vehicle; construct the optimal control problem of the host vehicle based on the inequality constraint conditions, the objective function of the host vehicle, and the equality constraint conditions of the host vehicle.

[0141] A vehicle model

[0142] Vehicle state The control quantity is described by the distance and speed of the vehicle along the planned path S direction Characterized by a single control variable acceleration. Where s represents the longitudinal position in the SL coordinate system, represents the first derivative of s with respect to time t (i.e., speed), represents the second derivative of s with respect to time t (i.e., acceleration), represents the third derivative of s with respect to time t (i.e., jerk). The vehicle state can be established for the host vehicle and the interacting vehicle respectively.

[0143] In addition, define the vehicle conflict flag z as follows:

[0144] The distance of the vehicle from the boundary of the conflict area is S c

[0145]

[0146] Among them, S and are the current position of the vehicle and the position of the conflict boundary point respectively.

[0147] The vehicle conflict flag z

[0148]

[0149] Among them, sign represents the sign function, respectively represent the distances from the current position of the vehicle to the front and rear boundaries of the conflict area, that is, when the vehicle has not entered the conflict area, z = 1; when the vehicle enters the conflict area, z = -1; when the vehicle exits the conflict area, z = 1.

[0150] B equality constraint

[0151] Assume that the minimum planning period of the vehicle is Δt, and there is

[0152]

[0153] Similarly, the second derivative of the objective function is continuous. Assume is a fixed constant, and there is

[0154]

[0155] Let

[0156]

[0157] There is

[0158]

[0159] Among them, A eqsub represents the equality matrix, and T represents the matrix transpose.

[0160] In the optimization time domain, there is an equality constraint

[0161]

[0162] Among them, the state

[0163] C inequality constraint

[0164] In speed optimization, the vehicle collision constraint is mainly considered and converted into a conflict form for processing. Here, it is assumed that the position of the front boundary point of the conflict area of the host vehicle is S n , and the position of the rear boundary point of the conflict area of the host vehicle is S r . The position of the front boundary point of the conflict area of the interacting vehicle is The position of the rear boundary point of the conflict area of the interacting vehicle is Considering the independence of the sign function transmission, therefore, the conflict geometric distance in the state expression is retained.

[0165] The conflict geometric distance z of the host vehicle i :

[0166] z i = (S n - s i ) * (S r - s i ), where S n - s i represents the first geometric distance of the host vehicle, and S r - s i represents the second geometric distance of the host vehicle.

[0167] Similarly, the conflict geometric distance z of the interacting vehicle j

[0168] where represents the first geometric distance of the interacting vehicle, represents the second geometric distance of the interacting vehicle.

[0169] where i and j are corresponding frames.

[0170] In the above calculation process, the host vehicle is imagined as a mass point. However, considering the geometric characteristics of the vehicle, assuming the vehicle length is L, there is

[0171] z = (S n - s f ) 2 + (S r - S n + L) (S n - s f )

[0172] where s f represents the position of the front boundary of the host vehicle body

[0173] In some alternative implementation manners of this embodiment, the conflict geometric distance is corrected by adjusting parameters, where the adjustment parameter is proportional to the sum of the conflict area size and the vehicle length.

[0174] Considering the above general form of the conflict geometric distance, an adjustment parameter σ is defined for different vehicles and different driving environments, and the above formula is written as

[0175]

[0176] To represent that the host vehicle and the interacting vehicle cannot appear in the conflict area simultaneously, it is expressed by the conflict inequality constraint as follows:

[0177] z i +z j ≥0

[0178] Replace s in the above formula f with the current geometric center position s of the host vehicle i , and the difference between the two is exactly half of the vehicle length. So s f = s i + L / 2. After arrangement, we get For the adjustment parameter σ, take different conflict areas (S r - S n ) and vehicle lengths (L), and plot the z image as shown in Figures 3e - 3f shown below.

[0179] (1) Different conflict areas

[0180] Fix the vehicle length at 2m, the distance of one vehicle to the conflict area is fixed at 2m, and the distances of the other vehicle to the conflict area are taken as 5m and 10m respectively. As shown in Figure 3e shown, it can be seen from the figure that the greater the distance between a vehicle and the conflict area, the greater the interaction buffer (interaction buffer, the distance to the conflict area) that the other interacting vehicle needs to reserve.

[0181] (2) Vehicle length

[0182] The distances of the two vehicles to the conflict area are fixed at 2m, the length of one vehicle is fixed at 2m, and the lengths of the other vehicle are taken as 2m and 10m respectively. As shown in Figure 3f shown, it can be seen from the figure that the longer the vehicle, the greater the interaction buffer that the other interacting vehicle needs to reserve.

[0183] Obviously, the interaction buffer of one vehicle is proportional to the size of the interaction buffer of the other vehicle and the vehicle length. The popular explanation is as follows: (1) For the conflict area, generally larger conflict areas occur in the lane-changing scenario. At this time, retaining a larger interaction buffer can ensure lateral safety; (2) For the vehicle length, the driving uncertainty of large vehicles is relatively large. Its characteristics such as large memory difference when turning determine that such vehicles have greater safety risks. Therefore, a larger interaction buffer should be reserved for safe avoidance.

[0184] Finally, to ensure that the interaction buffer is controllable, by adjusting the parameter σ, the interaction buffer is controlled within the empirical value range of [s min , s max , as shown below

[0185]

[0186] s min represents the minimum longitudinal distance, s max represents the maximum longitudinal distance.

[0187] D objective function

[0188] In some alternative implementation manners of this embodiment, the objective function includes at least one of the following: objective cost, second-order control quantity cost, and third-order control quantity cost. The objective cost represents the cost of achieving the desired driving speed without considering the interacting vehicle and road conditions. The second-order control quantity cost represents the minimum acceleration, and the third-order control quantity cost represents the minimum jerk.

[0189] The objective cost is obtained in the following manner: obtaining the following-following state function of the host vehicle and the following-following state function of the interacting vehicle based on the intelligent driving model; solving the following-following state function of the host vehicle and the following-following state function of the interacting vehicle respectively based on the anti-rollover constraint condition and the anti-skid constraint condition to obtain the target state of the host vehicle and the target state of the interacting vehicle; constructing the objective cost of the host vehicle and the objective cost of the interacting vehicle respectively based on the target state of the host vehicle and the target state of the interacting vehicle.

[0190] (1) Objective cost

[0191] Target state X ref is the expected state at the end of the vehicle trajectory planning, without considering the interacting vehicle and road conditions. The acceleration u of the following-following state of the host vehicle can be obtained from the IDM model (Intelligent Driver Model). constant :

[0192]

[0193] where represents the safe following distance, s0 is the minimum safe distance, T is the safe following time, v0 is the speed of the leading vehicle, v is the speed of the host vehicle, Δv is the speed difference, s is the longitudinal position in the SL coordinate system, a is the expected acceleration, b is the expected deceleration, and δ is a parameter, generally taken as 4.

[0194] When there is no following vehicle in the target lane,

[0195]

[0196] where v0 is the expected driving speed in the target lane.

[0197]

[0198] where i = 1, 2,..., n, denote

[0199]

[0200] Among them, X free represents the unconstrained state.

[0201] Considering the curvature constraint of the planned path, to prevent rollover, anti-rollover constraint conditions are set according to the gravitational acceleration, center of mass height, path point curvature, and wheelbase, as follows:

[0202]

[0203] Among them, g is the gravitational acceleration, H is the center of mass height, k i is the path point curvature, and W is the vehicle width.

[0204] To prevent sideslip, anti-sideslip constraint conditions are set according to the gravitational acceleration, path point curvature, and road surface adhesion coefficient, as follows:

[0205]

[0206] Among them, g is the gravitational acceleration, k i is the path point curvature, and u is the road surface adhesion coefficient.

[0207] J = (X - X free ) T R x (X - X free ) + X T R u X

[0208] s.t.

[0209] f r X ≤ X rollover

[0210] f s X ≤ X sideslip

[0211] Among them, J represents the cost function, R x , R u , f r , f s are parameter matrices, and the matrix form is [0, 1, 0, 0, 1, 0, 0, 1......]. X rollover represents the upper limit speed to prevent rollover X sideslip represents the upper limit speed to prevent sideslip

[0212] Solve the above problem to obtain the target state X ref . Based on this, there is a target cost J ref

[0213] Jref =(X - X ref ) T R f (X - X ref )

[0214] where R f is a parameter matrix in the matrix form of [0, 1, 0, 0, 1, 0, 0, 1......].

[0215] (2) Second - order control variable cost

[0216] The second - order control variable cost, i.e., the acceleration cost, is denoted as J u

[0217] J u = X T R u X

[0218] where R u is a parameter matrix in the matrix form of [0, 1, 0, 0, 1, 0, 0, 1......].

[0219] (3) Third - order control variable cost

[0220] The third - order control variable cost, i.e., the jerk cost, is denoted as

[0221]

[0222] where is a parameter matrix in the matrix form of [0, 1, 0, 0, 1, 0, 0, 1......].

[0223] Therefore, the total cost function J is

[0224]

[0225] Step 405: Based on the KKT conditions, fuse the optimal control problem of the interacting vehicle and the optimal control problem of the ego - vehicle into a single optimal control problem.

[0226] In this embodiment, for each party in the interaction process, considering the cost function comprehensively, that is, the cost function of one party should partially consider the cost of the other party, it is defined as follows

[0227] J union = αJ F +(1 - α)J L

[0228] where J union is the comprehensive cost of fusing the leader and the follower, J F is the cost of the follower, JL is the leader cost, and α ∈ [0, 1] is the trade-off coefficient. The larger its value, the more inclined to consider the follower (interactive vehicle).

[0229] For the follower, its optimal control problem (OCP) is characterized as follows

[0230]

[0231] The above equation represents solving for X when the follower cost is minimized F and u F , where X F represents the state of the follower, X L represents the state of the leader, and u F represents the acceleration of the follower.

[0232] s.t.

[0233] Equality constraint of the follower: h F (X L , X F , u F ) = 0, where h F represents the equality constraint function of the follower.

[0234] Inequality constraint of the follower: g F (X L , X F , u F ) ≤ 0, where g F represents the inequality constraint function of the follower.

[0235] Among them, the follower is assumed to know the leader state X L .

[0236] For the leader, its optimal control problem (OCP) is characterized as follows

[0237]

[0238] The above equation represents solving for X L , X F , u L and u F , where X F represents the state of the follower, X L represents the state of the leader, and u FDenote the acceleration of the follower as \(u\). L Denote the acceleration of the leader.

[0239] s.t.

[0240] Equality constraint of the leader: \(h\) L (\(X\) L , \(X\) F , \(u\) L ) = 0, where \(h\) F represents the equality constraint function of the leader.

[0241] Inequality constraint of the leader: \(g\) L (\(X\) L , \(X\) F , \(u\) L ) ≤ 0, where \(g\) L represents the inequality constraint function of the leader.

[0242] Among them, the state \(X\) of the follower F and the acceleration \(u\) F are obtained by solving the OCP problem of the follower.

[0243] When the OCP problem of the follower is convex, it is transformed into the KKT conditions. The KKT conditions (Karush-Kuhn-Tucker Conditions) are a necessary and sufficient condition for a non-linear programming problem to have an optimal solution under some regular conditions.

[0244] Construct the Lagrangian function \(L(X\) L , \(X\) F , \(u\) F , \(\lambda,\mu)\) as follows

[0245] \(L(X\) L , \(X\) F , \(u\) F , \(\lambda,\mu)\) = \(J\) F (\(X\) L , \(X\) F , \(u\) F ) + \(\lambda\) T \(h\) F (\(X\) L , \(X\) F , \(u\) F ) + \(\mu\) T \(g\) F (\(X\) L , \(X\) F , \(u\) F )

[0246] Among them, \(\lambda\) and \(\mu\) represent the Lagrange multipliers, and \(T\) represents the matrix transpose.

[0247] There is

[0248]

[0249] Among them, the above formula represents solving for X when the comprehensive cost is minimized L , X F , u L , u F , λ, and μ.

[0250] s.t.

[0251] Equality constraint of the leader: h L (X L , X F , u L ) = 0

[0252] Inequality constraint of the leader: g L (X L , X F , u L ) ≤ 0

[0253] Gradient condition:

[0254] Equality constraint of the follower: h F (X L , X F , u F ) = 0

[0255] Inequality constraint of the follower: g F (X L , X F , u F ) ≤ 0

[0256] μ ≥ 0

[0257]

[0258] Step 406, solve the single optimal control problem to obtain the speed planning result of the host vehicle and the speed planning result of the interacting vehicle.

[0259] In this embodiment, solve for (X union , X L , u F , λ, μ) when J F is minimized to obtain the speed planning result X L of the host vehicle and the speed planning result X F of the interacting vehicle.

[0260] The method provided by the above embodiments of the present disclosure proposes an efficient one-dimensional conflict constraint expression method under a decoupled framework of path planning and speed planning, reducing the computing power requirements for collision calculation.

[0261] Further referring to Figure 5 , as an implementation of the methods shown in the above figures, the present disclosure provides an embodiment of an interactive decision-making and planning device for an autonomous vehicle. This device embodiment corresponds to Figure 2 the method embodiment shown, and this device can be specifically applied to various electronic devices.

[0262] As shown in Figure 5 , the interactive decision-making and planning device 500 for an autonomous vehicle in this embodiment includes: a path planning unit 501, a conflict calculation unit 502, and a speed planning unit 503. Among them, the path planning unit 501 is configured to, in response to detecting the intention of the host vehicle to change lanes, perform path planning based on at least one of the driving state information of the host vehicle, the driving state information of the interactive vehicle, and the road environment information, and obtain the target planned path of the host vehicle; the conflict calculation unit 502 is configured to calculate the conflict area between the host vehicle and the interactive vehicle based on the target planned path and the driving state information of the interactive vehicle; the speed planning unit 503 is configured to perform speed planning based on the conflict area to obtain the speed planning result of the host vehicle and the speed planning result of the interactive vehicle.

[0263] In this embodiment, the specific processing of the path planning unit 501, the conflict calculation unit 502, and the speed planning unit 503 of the interactive decision-making and planning device 500 for an autonomous vehicle can refer to Figure 2 Steps 201, 202, and 203 in the corresponding embodiment.

[0264] In some alternative implementation manners of this embodiment, the road environment information includes the road width of the target lane; and the path planning unit 501 is further configured to: determine whether there is a lateral movement space for the host vehicle according to the road width; in response to determining that there is no lateral movement space for the host vehicle, use the lane center line during the lane change process as the target planned path of the host vehicle.

[0265] In some alternative implementation manners of this embodiment, the road environment information includes the road width of the target lane; and the path planning unit 501 is further configured to: determine whether there is a lateral movement space for the host vehicle according to the road width; in response to determining that there is a lateral movement space for the host vehicle, detect the congestion condition of the road ahead and the distance between the interactive vehicle and the host vehicle; in response to detecting that some lanes of the road ahead are congested and the distance is greater than a predetermined distance threshold, use the non-congested lane as the target planned path of the host vehicle.

[0266] In some alternative implementation manners of this embodiment, the path planning unit 501 is further configured to: in response to detecting that some lanes of the road ahead are congested and the distance is less than or equal to a predetermined distance threshold, determine the connection node of the planned path of the vehicle itself according to the position of the vehicle ahead in the congested lane; determine the path from the current position of the vehicle itself to the connection node as the first path, determine the path from the connection node to the non-congested lane as the second path, and merge the first path and the second path as the target planned path of the vehicle itself.

[0267] In some alternative implementation manners of this embodiment, the device 500 further includes a path optimization unit, which is configured to perform path planning according to at least one of the driving state information of the vehicle itself, the driving state information of the interacting vehicle, and the road environment information, to obtain the initial planned path of the vehicle itself; optimize the initial planned path based on the constraint conditions of the path planning, to obtain the target planned path of the vehicle itself.

[0268] In some alternative implementation manners of this embodiment, the constraint conditions include: continuity constraint conditions, smoothness constraint conditions, and boundary constraint conditions; and the path optimization unit is further configured to: perform curve fitting based on the initial planned path to obtain a connection function; construct continuity constraint conditions based on the second derivative of the connection function; construct boundary constraint conditions based on the corner positions of the vehicle itself; construct smoothness constraint conditions based on the first derivative of the connection function; solve the connection function based on the constraint conditions to obtain a target planned path that satisfies at least one of the following optimization objectives: the distance between the target planned path and the lane reference line is the closest, the number of turning points of the target planned path is the least, and the gap between the target planned path and the initial planned path is the smallest.

[0269] In some alternative implementation manners of this embodiment, the speed planning unit 503 is further configured to: respectively construct an optimal control problem for the vehicle itself and an optimal control problem for the interacting vehicle based on the conflict area, and fuse the optimal control problem for the interacting vehicle and the optimal control problem for the vehicle itself into a single optimal control problem based on the KKT conditions; solve the single optimal control problem to obtain the speed planning result of the vehicle itself and the speed planning result of the interacting vehicle.

[0270] In some alternative implementation manners of this embodiment, the speed planning unit 503 is further configured to: construct an objective function for the vehicle itself based on the distance, speed, and acceleration of the vehicle itself along the target planned path; construct an objective function for the interacting vehicle based on the distance, speed, and acceleration of the interacting vehicle along the driving direction; construct an optimal control problem for the interacting vehicle based on the inequality constraint conditions, the objective function of the interacting vehicle, and the equality constraint conditions of the interacting vehicle; construct an optimal control problem for the vehicle itself based on the inequality constraint conditions, the objective function of the vehicle itself, and the equality constraint conditions of the vehicle itself.

[0271] In some alternative implementation manners of this embodiment, the equality constraint condition includes that the second derivative of the objective function is continuous, and the inequality constraint condition includes that the sum of the conflict geometric distances of the host vehicle and the interacting vehicle is greater than or equal to zero, where the conflict geometric distance is the product of a first geometric distance and a second geometric distance, where the first geometric distance is the difference between the front boundary of the conflict area and the vehicle position, and the second geometric distance is the difference between the rear boundary of the conflict area and the vehicle position.

[0272] In some alternative implementation manners of this embodiment, the speed planning unit 504 is further configured to: correct the conflict geometric distance by adjusting a parameter, where the adjustment parameter is proportional to the sum of the conflict area size and the vehicle length.

[0273] In some alternative implementation manners of this embodiment, the objective function includes an objective cost, a second-order control quantity cost, and a third-order control quantity cost. The objective cost is obtained by the following method: obtaining the following-following state function of the host vehicle and the following-following state function of the interacting vehicle based on an intelligent driving model; respectively solving the following-following state function of the host vehicle and the following-following state function of the interacting vehicle based on the anti-rollover constraint condition and the anti-skid constraint condition to obtain the target state of the host vehicle and the target state of the interacting vehicle; respectively constructing the objective cost of the host vehicle and the objective cost of the interacting vehicle based on the target state of the host vehicle and the target state of the interacting vehicle.

[0274] It should be noted that in the technical solution of the present disclosure, in terms of the collection, collection, update, analysis, processing, use, transmission, storage, etc. of the user's personal information, it complies with the provisions of relevant laws and regulations, is used for legal purposes, and does not violate public order and good customs. Necessary measures are taken for the user's personal information to prevent illegal access to the user's personal information data, and to safeguard the user's personal information security, network security, and national security.

[0275] According to the embodiments of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and an autonomous vehicle.

[0276] An electronic device includes: one or more processors; a storage device on which one or more computer programs are stored. When the one or more computer programs are executed by the one or more processors, the one or more processors implement the method described in process 200 or 400.

[0277] A computer-readable medium has a computer program stored thereon, where the computer program, when executed by a processor, implements the method described in process 200 or 400.

[0278] An autonomous vehicle includes the above-mentioned electronic device.

[0279] Figure 6FIG. 0 is a schematic block diagram of an exemplary electronic device 600 that may be used to implement embodiments of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as, for example, a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as, for example, a personal digital processor, a cellular phone, a smart phone, a wearable device, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely exemplary and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0280] As Figure 6 shown, the device 600 includes a computing unit 601 that can perform various appropriate actions and processes in accordance with a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the device 600 can also be stored. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0281] A plurality of components in the device 600 are connected to the I / O interface 605, including: an input unit 606, such as a keyboard, a mouse, etc.; an output unit 607, such as various types of displays, speakers, etc.; a storage unit 608, such as a magnetic disk, an optical disk, etc.; and a communication unit 609, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 609 allows the device 600 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0282] The computing unit 601 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 601 executes the various methods and processes described above, such as the road area planning method. For example, in some embodiments, the road area planning method can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 608. In some embodiments, part or all of the computer program can be loaded and / or installed onto the device 600 via the ROM 602 and / or the communication unit 609. When the computer program is loaded into the RAM 603 and executed by the computing unit 601, one or more steps of the road area planning method described above can be executed. Alternatively, in other embodiments, the computing unit 601 can be configured to execute the road area planning method by any other suitable means (e.g., by means of firmware).

[0283] The various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuitry, integrated circuit systems, field-programmable gate arrays (FPGA), application-specific integrated circuits (ASIC), application-specific standard products (ASSP), systems-on-a-chip (SOC), complex programmable logic devices (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a dedicated or general-purpose programmable processor, and can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0284] The program code for implementing the methods of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to the processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the program codes are executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The program codes can be executed entirely on the machine, partially on the machine, as an independent software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0285] In the context of this disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0286] In order to provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, speech input, or tactile input).

[0287] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0288] A computer system may include a client and a server. The client and the server are generally far from each other and usually interact via a communication network. The relationship between the client and the server is generated by computer programs running on the respective computers and having a client-server relationship with each other. The server may be a server of a distributed system or a server incorporating a blockchain. The server may also be a cloud server, or an intelligent cloud computing server or an intelligent cloud host with artificial intelligence technology. The server may be a server of a distributed system or a server incorporating a blockchain. The server may also be a cloud server, or an intelligent cloud computing server or an intelligent cloud host with artificial intelligence technology.

[0289] It should be understood that various forms of the processes shown above can be used, with steps reordered, added, or deleted. For example, the steps recited in this disclosure can be executed 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, and no limitations are imposed herein.

[0290] The above specific embodiments do not constitute a limitation on the protection scope of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the protection scope of this disclosure.

Claims

1. An interactive decision-making planning method for an autonomous driving vehicle, comprising: In response to detecting the intention of the ego vehicle to change lanes, performing path planning according to at least one of the driving state information of the ego vehicle, the driving state information of the interactive vehicle, and the road environment information to obtain a target planned path of the ego vehicle; Calculating a conflict area between the self-vehicle and the interactive vehicle based on the target planning path and the driving state information of the interactive vehicle; Speed ​​planning is performed based on the conflict area to obtain speed planning results of the vehicle and the interacting vehicle.

2. The method according to claim 1, wherein: The road environment information includes the road width of the target lane; and The process of performing path planning according to at least one of the driving state information of the own vehicle, the driving state information of the interactive vehicle, and the road environment information to obtain a target planned path of the own vehicle includes: Determining whether there is space for the vehicle to move laterally based on the road width; In response to determining that there is no lateral movement space for the ego vehicle, a center line of the lane during the lane change process is used as a target planning path for the ego vehicle.

3. The method according to claim 1, wherein: The road environment information includes the road width of the target lane; and The process of performing path planning according to at least one of the driving state information of the own vehicle, the driving state information of the interactive vehicle, and the road environment information to obtain a target planned path of the own vehicle includes: Determining whether there is space for the vehicle to move laterally based on the road width; In response to determining that there is lateral movement space for the own vehicle, detecting a congestion condition of a road ahead and a distance between the interactive vehicle and the own vehicle; In response to detecting that a portion of the lanes of the road ahead are congested and the distance is greater than a predetermined distance threshold, a non-congested lane is used as a target planning path for the vehicle.

4. The method according to claim 3, wherein: The method further comprises: In response to detecting that a portion of the lane of the road ahead is congested and the distance is less than or equal to a predetermined distance threshold, determining a connection node of the planned path of the vehicle according to a position of a front vehicle in the congested lane; The path from the current position of the vehicle to the connecting node is determined as the first path segment, and the path from the connecting node to the non-congested lane is determined as the second path segment, wherein the first path segment and the second path segment are combined as the target planning path of the vehicle.

5. The method according to claim 1, wherein: The process of performing path planning according to at least one of the driving state information of the own vehicle, the driving state information of the interactive vehicle, and the road environment information to obtain a target planned path of the own vehicle includes: Performing path planning based on at least one of the driving state information of the own vehicle, the driving state information of the interactive vehicle, and the road environment information to obtain an initial planned path of the own vehicle; The initial planned path is optimized based on the constraint conditions of the path planning to obtain the target planned path of the vehicle.

6. The method according to claim 5, wherein: The constraints include: continuity constraints, smoothness constraints and boundary constraints; and The initial planned path is optimized based on the constraint conditions of the path planning to obtain the target planned path of the vehicle, including: Performing curve fitting based on the initial planned path to obtain a connection function; constructing a continuity constraint based on the second-order derivative of the link function; Construct boundary constraints based on the corner positions of the ego vehicle; constructing a smoothness constraint based on the first-order derivative of the link function; The connection function is solved based on the constraint conditions to obtain a target planning path that meets at least one of the following optimization objectives: the target planning path is closest to the lane reference line, the target planning path has the fewest turning points, and the target planning path has the smallest difference from the initial planning path.

7. The method according to claim 1, wherein: The speed planning is performed based on the conflict area to obtain the speed planning result of the vehicle and the speed planning result of the interactive vehicle, including: Based on the conflict area, an optimal control problem of the self-vehicle and an optimal control problem of the interactive vehicle are respectively constructed, and the optimal control problem of the interactive vehicle and the optimal control problem of the self-vehicle are merged into a single optimal control problem; Solve the single optimal control problem to obtain the speed planning results of the self-vehicle and the speed planning results of the interactive vehicle.

8. The method according to claim 7, wherein: The constructing of the optimal control problem of the self-vehicle and the optimal control problem of the interactive vehicle based on the conflict area respectively includes: constructing an objective function of the ego vehicle based on the distance, speed, and acceleration of the ego vehicle along the target planning path; Constructing the objective function of the interactive vehicle based on the distance, speed, and acceleration of the interactive vehicle along the travel direction; Constructing an optimal control problem for the interactive vehicle based on the inequality constraint, the objective function of the interactive vehicle and the equality constraint of the interactive vehicle; An optimal control problem of the ego vehicle is constructed based on the inequality constraints, the objective function of the ego vehicle and the equality constraints of the ego vehicle.

9. The method according to claim 8, wherein: The equality constraint condition includes: the second-order derivative of the objective function is continuous; the inequality constraint condition includes: the sum of the conflict geometric distance of the self-vehicle and the conflict geometric distance of the interactive vehicle is greater than or equal to zero, wherein the conflict geometric distance is the product of a first geometric distance and a second geometric distance, wherein the first geometric distance is the difference between the front boundary of the conflict area and the vehicle position, and the second geometric distance is the difference between the rear boundary of the conflict area and the vehicle position.

10. The method according to claim 9, wherein: The method further comprises: The conflict geometric distance is corrected by adjusting a parameter, wherein the adjustment parameter is proportional to the sum of the conflict area size and the vehicle length.

11. The method according to claim 8, wherein: The objective function includes the target cost, the second-order control amount cost and the third-order control amount cost, and the target cost is obtained by: Obtaining the following state function of the self-vehicle and the following state function of the interactive vehicle based on the intelligent driving model; Based on the rollover prevention constraint condition and the side slip prevention constraint condition, the following state function of the ego vehicle and the following state function of the interactive vehicle are solved respectively to obtain the target state of the ego vehicle and the target state of the interactive vehicle; A target cost of the ego vehicle and a target cost of the interactive vehicle are constructed based on the target state of the ego vehicle and the target state of the interactive vehicle, respectively.

12. An interactive decision-making planning device for an autonomous driving vehicle, comprising: a path planning unit configured to, in response to detecting the intention of the ego vehicle to change lanes, perform path planning based on at least one of the driving state information of the ego vehicle, the driving state information of the interactive vehicle, and the road environment information, and obtain a target planned path for the ego vehicle; a conflict calculation unit, configured to calculate a conflict area between the self-vehicle and the interactive vehicle based on the target planned path and the driving state information of the interactive vehicle; The speed planning unit is configured to perform speed planning based on the conflict area to obtain a speed planning result of the own vehicle and a speed planning result of the interactive vehicle.

13. An electronic device comprising: one or more processors; a storage device having one or more computer programs stored thereon, When the one or more computer programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 11.

14. A computer readable medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the method according to any one of claims 1 to 11 is implemented.

15. An autonomous driving vehicle comprising the electronic device as claimed in claim 13.

Images