An autonomous vehicle cooperative platoon control method and device and a storage medium

Abstract

The application discloses an automatic driving car cooperative type formation control method and device, and a storage medium. The method comprises the following steps: determining vehicle driving information of a target vehicle in a preset geographical range; determining reference driving information of the target vehicle based on the vehicle driving information and acquired static environment information in the preset geographical range, wherein the static environment information comprises road information of a plurality of lanes, and the reference driving information comprises at least one candidate path of the target vehicle; determining corresponding car-following preceding vehicles of the target vehicle on each candidate path by using the reference driving information based on acquired surrounding vehicle driving information of the target vehicle; and constructing a finite time domain optimal control problem in real time for the candidate path and the car-following preceding vehicles, so as to determine a target driving track of the target vehicle and a corresponding target car-following preceding vehicle by using the finite time domain optimal control problem. The method realizes the automatic driving function of multi-vehicle cooperative type formation decision control.

Description

Technical Field

[0001] This application relates to the field of autonomous decision-making and control technology for autonomous vehicles, and in particular to a cooperative formation control method, device and storage medium for autonomous vehicles. Background Technology

[0002] In the field of autonomous driving, intelligent connected vehicles, as the intersection of intelligent vehicles and vehicle networks, rely on the communication advantages of vehicle networks and advanced perception and monitoring systems to obtain rich information about surrounding vehicles and the environment, thereby achieving high-performance intelligent connected vehicle decision-making and control.

[0003] Existing queuing or grouping methods for intelligent connected vehicles mainly consider single-lane queuing control problems. In multi-lane scenarios, these methods suffer from low lane utilization and long queues can hinder lane changes for surrounding vehicles. Therefore, a distributed, multi-vehicle cooperative queuing decision-making and control method is needed for large-scale intelligent connected vehicles to better leverage their advantages. Summary of the Invention

[0004] The purpose of this application is to provide a method, device and storage medium for cooperative formation control of autonomous vehicles. This method realizes multi-vehicle cooperative formation decision control, thereby realizing various autonomous driving functions such as autonomous formation, formation maintenance and switching of intelligent connected vehicle groups.

[0005] To achieve the above objectives, this application provides a cooperative platooning control method for autonomous vehicles, comprising:

[0006] Within a preset geographical area, determine the vehicle's driving information;

[0007] Based on the vehicle driving information and the obtained static environment information within the preset geographical range, the reference driving information of the target vehicle is determined, wherein the static environment information includes road information of multiple lanes, and the reference driving information includes at least one candidate path of the target vehicle.

[0008] Based on the obtained surrounding vehicle driving information of the target vehicle, the reference driving information is used to determine the following vehicle corresponding to the target vehicle on each candidate path.

[0009] For the candidate path and the car following ahead, a finite-time optimal control problem is constructed in real time to determine the target driving trajectory of the target vehicle and the corresponding target car following ahead. The finite-time optimal control problem includes constraint information that constrains the target vehicle.

[0010] Optionally, determining the reference driving information of the target vehicle based on the vehicle driving information and the obtained static environmental information within the preset geographical range includes:

[0011] Determine the current location of the target vehicle, as well as the number of lanes and road structure information of the lanes that are at least partially passable;

[0012] The candidate path is determined based on the current location, the number of lanes, and / or the road structure information.

[0013] Optionally, determining the candidate path based on the current location, the lane quantity information, and / or the road structure information includes:

[0014] The centerline of each lane is selected as the pre-driving line, and the pre-driving line is discretized in equal space.

[0015] The discretized pre-driving line is set as τ. i Based on the vehicle's traffic regulations and driving habits, a desired reference speed is determined for each candidate route.

[0016] Based on the desired reference speed, the set of reference trajectories corresponding to the candidate paths is defined as follows: Where, N lane This represents the total number of lanes that are partially passable.

[0017] Optionally, the step of determining the following vehicle corresponding to the target vehicle on each candidate path based on the obtained surrounding vehicle driving information and the reference driving information includes:

[0018] The vehicle driving information of the surrounding vehicles in each of the candidate paths is obtained through the network platform.

[0019] On each candidate path, determine the distance between the projected position of the target vehicle and the projected positions of the surrounding vehicles;

[0020] If the distance satisfies the first distance condition, the nearest surrounding vehicle on each candidate path to the target vehicle is identified as the following vehicle corresponding to the target vehicle.

[0021] Optionally, the step of determining the following vehicle corresponding to the target vehicle on each candidate path based on the obtained surrounding vehicle driving information and the reference driving information specifically includes:

[0022] Based on the current location of the target vehicle, extract a set of reference trajectories Π corresponding to the candidate path from the preset geographical range;

[0023] Based on the reference trajectory set Π, all the candidate paths τ i Projection is performed to determine the projection points o of the reference trajectory set Π on each of the candidate paths. i，ego i = 1: N lane ;

[0024] The set of surrounding vehicle path projections received by the target vehicle is defined as follows: Where N surr O is the number of surrounding vehicles. j ={τ i o i，j} is the projection set of the candidate paths transmitted by the network platform for the j-th surrounding vehicle, τ i For the candidate path where the j-th surrounding vehicle is currently located, o i，j For the target vehicle on the candidate path τ i Projection points on;

[0025] For each candidate path τ of the target vehicle i If the surrounding vehicle j is located on the candidate path τ i If the target vehicle and the surrounding vehicles are compared on the candidate path τ, then... i Projection point o on i，ego and o i，j ,like:

[0026] d safe ≤o i，j -o i，ego ≤d bound

[0027] Then the surrounding vehicles j are considered as the target vehicle on the candidate path τ. i The candidate for the lead vehicle, of which d bound It is the vehicle-to-the-front (D) tracking threshold. safe Determine the safe distance that needs to be maintained between the target vehicle and the surrounding vehicles; analyze all vehicles on the same candidate path τ. i The vehicle closest to the target vehicle is identified as the vehicle on the candidate path τ. i The target vehicle mentioned above corresponds to the following vehicle p. i The corresponding following distance is

[0028]

[0029]

[0030] If there are no surrounding vehicles within the preceding vehicle tracking threshold, then the vehicle following the preceding vehicle p... i Set it to -1, and define the set of cars following the leader as...

[0031] Optionally, the real-time construction of the finite-time optimal control problem includes:

[0032] Construct an objective function for the reference trajectory corresponding to the candidate path, wherein the objective function is used to determine the reference trajectory, and the objective function includes at least one of the following: reference trajectory tracking content, vehicle control content, and distance control content of the following vehicle;

[0033] The real-time construction of the finite-time optimal control problem also includes:

[0034] The constraint information constrains the target vehicle, wherein the constraint information includes at least one of the following: dynamic model conditions of the target vehicle, control quantity constraint conditions, following distance conditions, safety constraint conditions between the target vehicle and surrounding vehicles, and safety constraint conditions between the target vehicle and the road.

[0035] As an option, where,

[0036] The i-th reference trajectory corresponding to the constructed candidate path The objective function is:

[0037]

[0038] in, The control sequence is N, which is the optimization variable of the objective function. p Let k be the number of prediction time-domain steps in the finite-time optimal control problem, k∈{0,…,N}. p -1}, For the i-th reference trajectory From o i The reference trajectory at the initial k-th step, x k To predict the target vehicle state at the k-th step in the time domain, let x0 be the current state of the target vehicle, Q be the state weight matrix, and u be the weight matrix. k To predict the target vehicle control quantity at the k-th step in the time domain, R is the control quantity weight matrix. To predict the target vehicle and candidate path τ at the k-th step in the time domain i The distance between the car in front and the car on the road, d des Let S be the desired following distance, and J be the scaling factor of the target vehicle. bias A fixed bias is used to reward the fuel economy and safety benefits of following the vehicle in front. This is an indication function used to indicate the car-following target and the fixed bias;

[0039] Accordingly, the dynamic model conditions of the target vehicle are:

[0040] x k+1 =f ego (x k u k )

[0041] Among them, f ego It is a nonlinear mapping function that reflects the transition relationship of the target vehicle from the state quantity at step k to the state quantity at step k+1 under the control quantity at step k;

[0042] The control quantity constraint condition is as follows:

[0043]

[0044] in, The feasible set of control variables includes the upper and lower bounds of each output control variable, reflecting the physical constraints in the real world;

[0045] The following distance condition is:

[0046]

[0047] in, p represents the preceding car on the i-th candidate path. i The state quantity at step k, It is a non-linear mapping, representing the distance between the target vehicle and the following vehicle at the k-th step in the prediction time domain;

[0048] The safety constraints between the target vehicle and the surrounding vehicles are as follows:

[0049] h(x k x surr，k )≥d safe

[0050] Where, x surr，k Let X be the state quantity of the surrounding vehicles at the k-th step in the prediction time domain, which is determined by the predicted trajectory X of the surrounding vehicles. surr As determined, h(x) k x surr，k () is a nonlinear mapping that reflects the distance between the target vehicle and the surrounding vehicles at the k-th step in the prediction time domain;

[0051] The safety constraints between the target vehicle and the road are as follows:

[0052] h(x k x road，k )≥d safe

[0053] Where, x road，k To predict the road information at the k-th step in the time domain, h(x) k x road，k ) is a nonlinear mapping that reflects the distance between the target vehicle and the road at the k-th step in the prediction time domain.

[0054] Optionally, the step of constructing a finite-time optimal control problem in real time for the candidate path and the following vehicle, and using the finite-time optimal control problem to determine the target driving trajectory of the target vehicle and the corresponding target following vehicle, includes:

[0055] The rolling time-domain optimization method is used to solve the objective function, dynamic model conditions of the target vehicle, control quantity constraints, following distance conditions, safety constraints with surrounding vehicles, and safety constraints with the road in the finite-time domain optimal control problem, thereby obtaining the optimal objective function corresponding to the i-th reference trajectory. Optimal control sequence and optimal state sequence

[0056] The objective function for constructing the optimal tracking trajectory is solved by:

[0057]

[0058] Select the optimal path τ * and the corresponding optimal control sequence and optimal state sequence

[0059] Define the target vehicle path projection set as O = {τ} * o *}, where o * The nearest projection point on the optimal path;

[0060] The target vehicle path projection set O and the optimal state sequence X are transmitted to the surrounding vehicles through the network platform.

[0061] Optionally, the method further includes:

[0062] Based on the target vehicle's travel target, determine whether the target vehicle has reached its destination;

[0063] If it is determined that the target vehicle has not reached the destination, the static environment information is reacquired to redetermine the target vehicle's target driving trajectory and the corresponding target following vehicle.

[0064] This application also provides a cooperative platooning control device for autonomous vehicles, including:

[0065] The acquisition module is configured to determine the vehicle driving information of the target vehicle within a preset geographical range;

[0066] The processing module is configured to determine the reference driving information of the target vehicle based on the vehicle driving information and the static environment information obtained within the preset geographical range, wherein the static environment information includes road information of multiple lanes and the reference driving information includes at least one candidate path of the target vehicle.

[0067] Based on the obtained surrounding vehicle driving information of the target vehicle, the reference driving information is used to determine the following vehicle corresponding to the target vehicle on each candidate path.

[0068] For the candidate path and the car following ahead, a finite-time optimal control problem is constructed in real time to determine the target driving trajectory of the target vehicle and the corresponding target car following ahead. The finite-time optimal control problem includes constraint information that constrains the target vehicle.

[0069] This application also provides a storage medium carrying one or more computer programs, which, when executed by a processor, implement the steps of the method described above.

[0070] The method in this embodiment relies on the communication architecture of the Internet of Vehicles to obtain the status and intentions of surrounding traffic participants in a timely manner, and constructs a distributed optimal control problem in combination with reference trajectory. Through steps such as static trajectory planning, vehicle platooning decision-making, and dynamic trajectory optimization and tracking, it realizes multi-vehicle collaborative formation decision-making control, thereby realizing various autonomous driving functions such as autonomous formation, formation maintenance and switching of intelligent connected vehicle groups.

[0071] Furthermore, the method described in this application employs a distributed control approach during the control process, allowing each vehicle to autonomously control itself, thus reducing the computational resource requirements of intelligent connected vehicles. Each vehicle can autonomously select the vehicle in front to follow, resulting in a flexible and adaptable overall platoon configuration. The control algorithm considers surrounding vehicles as dynamic traffic participants, enabling high real-time dynamic response. Attached Figure Description

[0072] Figure 1 This is a flowchart of an autonomous vehicle cooperative formation control method according to an embodiment of this application;

[0073] Figure 2 Examples of embodiments of this application Figure 1A flowchart of one embodiment of step S200;

[0074] Figure 3 Examples of embodiments of this application Figure 1 A flowchart of one embodiment of step S300;

[0075] Figure 4 This is a flowchart of one embodiment of the cooperative formation control method for autonomous vehicles according to this application;

[0076] Figure 5 This is a schematic diagram of the overall structure of the control system of the target vehicle according to an embodiment of this application;

[0077] Figure 6 This is a schematic diagram of static trajectory planning according to an embodiment of the present disclosure;

[0078] Figure 7 This is a schematic diagram of workshop queueing decision-making according to an embodiment of this disclosure;

[0079] Figure 8 This is a schematic diagram of preferred dynamic trajectory tracking according to an embodiment of the present disclosure;

[0080] Figure 9 This is a flowchart of another specific embodiment of the autonomous vehicle cooperative formation control method of this disclosure;

[0081] Figure 10 This is a structural block diagram of an autonomous vehicle cooperative formation control device according to an embodiment of the present disclosure. Detailed Implementation

[0082] Various embodiments and features of this application are described herein with reference to the accompanying drawings.

[0083] It should be understood that various modifications can be made to the embodiments described herein. Therefore, the above description should not be considered as limiting, but merely as an example of embodiments. Other modifications within the scope and spirit of this application will be apparent to those skilled in the art.

[0084] The accompanying drawings, which are included in and form part of this specification, illustrate embodiments of the present application and, together with the general description of the present application given above and the detailed description of the embodiments given below, serve to explain the principles of the present application.

[0085] These and other features of this application will become apparent from the following description of preferred forms of embodiments given as non-limiting examples, with reference to the accompanying drawings.

[0086] It should also be understood that although this application has been described with reference to some specific examples, those skilled in the art can certainly implement many other equivalent forms of this application.

[0087] The above and other aspects, features and advantages of this application will become more apparent when taken in conjunction with the accompanying drawings and in view of the following detailed description.

[0088] Specific embodiments of this application are described thereafter with reference to the accompanying drawings; however, it should be understood that the claimed embodiments are merely examples of this application, which can be implemented in various ways. Well-known and / or repeated functions and structures are not described in detail to avoid unnecessary or redundant details that could obscure the application. Therefore, the specific structural and functional details claimed herein are not intended to be limiting, but merely serve as the basis and representative basis for the claims to teach those skilled in the art to use this application in a variety of substantially any suitable detailed structures.

[0089] This specification may use the phrases “in one embodiment,” “in another embodiment,” “in yet another embodiment,” or “in other embodiments,” all of which may refer to one or more of the same or different embodiments according to this application.

[0090] This application discloses an autonomous vehicle cooperative platooning control method. This method can be applied to target vehicles such as automobiles. The target vehicle can interact with surrounding vehicles through the Internet of Vehicles (IoV) platform. According to the multi-vehicle cooperative control method, multi-vehicle cooperation between the target vehicle and its surrounding vehicles can be realized, which greatly leverages the advantages of the IoV and enables multi-vehicle cooperative autonomous driving.

[0091] The method mainly includes the following steps: For a target vehicle traveling within a preset geographical area (such as a pre-stored map), the target vehicle's control system can determine the vehicle's driving information, including its initial position, travel target, and approximate path. This information can be obtained through the target vehicle's positioning system or through a vehicle-to-everything (V2X) network platform. Furthermore, static environmental information within the preset geographical area can be obtained through the V2X, including road information for multiple lanes. Based on the approximate path and static environmental information, the target vehicle can determine its reference driving information, which includes at least one candidate path. Additionally, the target vehicle can send a request to the network platform to obtain surrounding vehicle driving information. The network platform can determine the surrounding vehicle driving information within the preset geographical area based on the target vehicle's driving information and information from other vehicles, and send this information to the target vehicle. Based on the received surrounding vehicle driving information and the reference driving information, the target vehicle uses this information to determine the corresponding following vehicle on each candidate path. Then, using a real-time constructed finite-time optimal control problem, the target vehicle's target trajectory and the corresponding following vehicle are determined. This allows the target vehicle to travel based on the target trajectory and follow the vehicle in front. Furthermore, the connected platform can transmit the target vehicle's driving information to surrounding vehicles so they can drive accordingly. This method enables multi-vehicle cooperation, improves lane utilization, and ensures the safety and driving efficiency of all vehicles.

[0092] The method of this application will now be described in detail with reference to the accompanying drawings, such as... Figure 1 As shown and combined Figure 5 The method includes the following steps:

[0093] S100 determines the vehicle driving information of the target vehicle within a preset geographical range.

[0094] For example, the preset geographical range can be a pre-defined geographic location, which can exist in the form of a pre-stored map, such as pre-stored in the control unit (or control system) of the target vehicle. The pre-stored map can be set according to the actual application scenario of the target vehicle, or it can be set based on empirical data. For example, the preset geographical range can be set according to the current location of the target vehicle and its travel target.

[0095] On the one hand, within a preset geographical area, the target vehicle can determine its relevant driving information, including its initial location, destination, and approximate path. This information can be obtained using appropriate positioning devices, such as multiple vehicle-mounted positioning devices to collect the information separately. On the other hand, the target vehicle can also obtain its driving information from other auxiliary devices.

[0096] S200, based on the vehicle driving information and the obtained static environment information within the preset geographical range, determine the reference driving information of the target vehicle, wherein the static environment information includes road information of multiple lanes, and the reference driving information includes at least one candidate path of the target vehicle.

[0097] For example, static environmental information within a preset geographical area may include road information for multiple lanes within that area. The target vehicle and surrounding vehicles can travel collaboratively on these multiple roads, including going straight and changing lanes. Road information may include parameters such as the number, shape, distance, location, and size of roads. Based on its initial position, travel target, approximate path, and other vehicle driving information, as well as this static environmental information, the target vehicle can determine its reference driving information. In other words, reference driving information can be determined based on the target vehicle's current state and road information, thus providing a reference for the target vehicle's driving.

[0098] In this embodiment, the reference driving information includes at least one candidate path for the target vehicle and a desired reference speed. The candidate path corresponds to a reference trajectory, from which a trajectory can be selected for the target vehicle to drive. In one embodiment, the data for each candidate path may be a series of neighboring points. In another embodiment, the reference trajectory may include the candidate path and driving speed parameters.

[0099] In one embodiment, a static trajectory planning layer (or static trajectory planning layer) can be constructed in the control system of the target vehicle. This static trajectory planning layer can implement the above steps, combined with... Figure 9 This includes determining the reference driving information of the target vehicle based on vehicle driving information and the obtained static environmental information.

[0100] S300, based on the obtained surrounding vehicle driving information of the target vehicle, and using the reference driving information, determine the following vehicle corresponding to the target vehicle on each of the candidate paths.

[0101] For example, the surrounding vehicle driving information refers to the driving-related information of the surrounding vehicles, such as their driving trajectories and position information. This surrounding vehicle driving information can be obtained by the vehicle-to-everything (V2X) network platform from the surrounding vehicles and then sent to the target vehicle.

[0102] The target vehicle accurately identifies the preceding vehicle on each candidate path based on the driving trajectories and positional information of surrounding vehicles. Each preceding vehicle can be considered as a candidate vehicle that the target vehicle can follow.

[0103] Specifically, when surrounding vehicles are traveling within a preset geographical area, they can move at a specific speed and switch between a single road or multiple roads. The surrounding vehicle travel information includes various relevant information about the surrounding vehicles in that state. The target vehicle can combine its own candidate paths and speed parameters with the acquired surrounding vehicle travel information to determine the vehicles following it on each candidate path, thus providing the data basis for selecting the target vehicle to follow.

[0104] In one embodiment, a workshop queueing decision layer (or workshop queueing decision) can be constructed in the control system of the target vehicle. This workshop queueing decision layer can implement the above steps, combined with... Figure 9 This includes determining the preceding vehicle corresponding to the target vehicle on each candidate path based on the acquired surrounding vehicle driving information and using reference driving information.

[0105] S400, for the candidate path and the car following ahead, a finite-time optimal control problem is constructed in real time, so as to use the finite-time optimal control problem to determine the target driving trajectory corresponding to the target vehicle and the corresponding target car following ahead, wherein the finite-time optimal control problem includes constraint information for constraining the target vehicle.

[0106] For example, the role of the finite-time optimal control problem is to derive a specific optimal system control law by constructing a performance index function in the finite time domain and taking its extreme value for an existing system model.

[0107] The finite-time optimal control problem includes constraint information (or constraint conditions) that constrain a target vehicle. In one embodiment, the constraint information may include at least one of the following: dynamic model conditions of the target vehicle, control quantity constraints, following distance conditions, safety constraints with surrounding vehicles, and safety constraints with the road.

[0108] Finite-time optimal control problems can also include an objective function for a reference trajectory, which includes the following performance metrics: reference trajectory tracking performance, control output performance, and following performance of the vehicle ahead.

[0109] The target vehicle can use a real-time constructed finite-time optimal control problem to constrain candidate paths and the following vehicles on each candidate path. This includes: constraining the candidate paths and the following vehicles on each candidate path using the target vehicle's dynamic model conditions, control quantity constraints, following distance conditions, safety constraints with surrounding vehicles, and safety constraints with the road; and using an objective function to define the target vehicle's target trajectory and the corresponding target following vehicles. The target trajectory and the corresponding target following vehicles are then obtained.

[0110] Thus, the target vehicle can travel along the target trajectory and follow the vehicle in front of it, achieving multi-vehicle cooperative driving while ensuring safety and meeting the requirements of multi-lane driving. Furthermore, the target vehicle can send its driving information to the network platform, which can then share this information with other vehicles (including those around the target vehicle) for their use.

[0111] In one embodiment, the target vehicle's control system can construct a dynamic trajectory optimization tracking layer (or dynamic trajectory optimization tracking). This dynamic trajectory optimization tracking layer can implement the above steps, combined with... Figure 9 This includes determining the target driving trajectory and the corresponding target following vehicle by using a real-time constructed finite-time domain optimal control problem based on candidate paths and preceding vehicles.

[0112] The autonomous vehicle cooperative platooning control method of this embodiment relies on the communication architecture of the Internet of Vehicles to obtain the status and intentions of surrounding traffic participants in a timely manner, and constructs a distributed optimal control problem in combination with reference trajectory. Through steps such as static trajectory planning, vehicle platooning decision-making, and dynamic trajectory optimization and tracking, it realizes multi-vehicle cooperative platooning decision-making control, thereby realizing a variety of autonomous driving functions such as autonomous platooning, formation maintenance and switching of intelligent connected vehicle groups.

[0113] Furthermore, the method described in this application employs a distributed control approach during the control process, allowing each vehicle to control itself autonomously, thus reducing the computational resource requirements of intelligent connected vehicles. Each vehicle can autonomously select the vehicle in front to follow, resulting in a flexible and adaptable overall platoon configuration. Each vehicle can be considered a dynamic traffic participant, enabling high real-time dynamic response.

[0114] In one embodiment of this application, the reference driving information of the target vehicle is determined based on the vehicle driving information and the obtained static environmental information within the preset geographical range, such as... Figure 2 As shown, it includes:

[0115] S210, determine the current location of the target vehicle, as well as the lane number information and road structure information of the lanes that are at least partially passable;

[0116] S220, determine the candidate path based on the current location, the number of lanes information and / or the road structure information.

[0117] For example, within a preset geographical area, the target vehicle and surrounding vehicles can travel in multiple lanes. Road structure information includes static information such as the shape and structure of the road.

[0118] In this embodiment, the static trajectory planning layer in the target vehicle's control system can determine the target vehicle's current location, lane quantity information, and road structure information through positioning and / or networking, thereby determining candidate paths.

[0119] In one specific embodiment, determining the candidate path based on the current location, the lane quantity information, and / or the road structure information includes the following steps:

[0120] The centerline of each lane is selected as the pre-driving line, and the pre-driving line is discretized in equal space.

[0121] The discretized pre-driving line is set as τ. i Based on the vehicle's traffic regulations and driving habits, a desired reference speed is determined for each candidate route.

[0122] Based on the desired reference speed, the set of reference trajectories corresponding to the candidate paths is defined as follows: Where, N lane This represents the total number of lanes that are partially passable.

[0123] Specific combination Figure 6 The static trajectory planning layer needs to determine the current position of the target vehicle, consider all locally passable lanes in the pre-stored map, and for each lane L... i i = 1: N lane N lane Given the total number of locally passable lanes, and based on static information such as road structure, the centerline of each lane is selected as a candidate path. The candidate paths are then spatially discretized to form numerous discrete points. The discretized candidate paths are denoted as τ. i Based on traffic regulations such as road speed limits and human driving habits, a desired reference speed is determined for each candidate path. Define the reference trajectory set as It can be predefined in a preset map and then called up online in real time.

[0124] In one embodiment of this application, based on the obtained surrounding vehicle driving information of the target vehicle, the reference driving information is used to determine the following vehicle corresponding to the target vehicle on each of the candidate paths, such as... Figure 3 As shown, it includes:

[0125] S310, obtain the surrounding vehicle driving information in each of the candidate paths through the network platform;

[0126] S320, on each candidate path, determine the distance between the projected position of the target vehicle and the projected position of the surrounding vehicles;

[0127] S330, if the distance satisfies the first distance condition, the nearest surrounding vehicle on each candidate path to the target vehicle is determined as the following vehicle corresponding to the target vehicle.

[0128] For example, the surrounding vehicle driving information refers to the driving information of the vehicles around the target vehicle, such as their driving trajectories and speeds. This surrounding vehicle driving information can be obtained by the vehicle-to-everything (V2X) network platform from the surrounding vehicles, and then the target vehicle can obtain this information from the network platform.

[0129] The projection position can be the projection of the target vehicle on the candidate path. The target vehicle and the surrounding vehicles each have their own corresponding projection positions, and there is a corresponding distance between them. The content recorded in this distance needs to meet the first distance condition in order to ensure that the target vehicle maintains a safe distance from the surrounding vehicles. This distance meeting the first distance condition can be used as a prerequisite for determining the vehicle to follow.

[0130] In one embodiment, the vehicle queueing decision layer in the target vehicle's control system can, if the distance meets a first distance condition, identify the nearest surrounding vehicle on each candidate path as the vehicle to follow the target vehicle. The distance between the target vehicle and the vehicle in front can be set according to actual driving needs, and can also be adjusted according to changes in the usage scenario.

[0131] In one specific embodiment, the step of determining the following vehicle corresponding to the target vehicle on each candidate path based on the obtained surrounding vehicle driving information and the reference driving information specifically includes the following steps, combined with... Figure 7 :

[0132] Based on the current location of the target vehicle, extract a set of reference trajectories Π corresponding to the candidate path from the preset geographical range;

[0133] Based on the reference trajectory set Π, all the candidate paths τ i Projection is performed to determine the projection points o of the reference trajectory set Π on each of the candidate paths. i，ego i = 1: N lane ;

[0134] The set of surrounding vehicle path projections received by the target vehicle is defined as follows: Where N surr O is the number of surrounding vehicles. j ={τ i o i，j} is the projection set of the candidate paths transmitted by the network platform for the j-th surrounding vehicle, τ i For the candidate path where the j-th surrounding vehicle is currently located, o i，j For the target vehicle on the candidate path τ i The projection points on the surrounding vehicle paths are used to filter the projection set, thereby determining the vehicles following the target vehicle on each candidate path. The filtering process is as follows:

[0135] For each candidate path τ of the target vehicle i If the surrounding vehicle j is located on the candidate path τ i If the target vehicle and the surrounding vehicles are compared on the candidate path τ, then... i Projection point o on i，ego and o i，j If the following formula (1) holds true:

[0136] d safe ≤o i，j -o i，ego ≤d bound (1)

[0137] Then the surrounding vehicles j are considered as the target vehicle on the candidate path τ. i The candidate for the lead vehicle, of which d bound It is the vehicle-to-the-front (D) tracking threshold. safe The safe distance that needs to be maintained between the target vehicle and the surrounding vehicles is determined by the above formula (1). If the distance between the target vehicle and the candidate vehicle in front satisfies the first distance condition, then the distance between the target vehicle and the candidate vehicle in front can be considered to satisfy the first distance condition.

[0138] Analyze all candidate paths τ in the same context i The vehicle closest to the target vehicle is identified as the vehicle on the candidate path τ. i The target vehicle mentioned above corresponds to the following vehicle.

[0139] Specifically, for all vehicles on the same candidate path, the vehicle closest to the target vehicle is considered as the following vehicle on that candidate path, as shown in formulas (2) and (3):

[0140]

[0141]

[0142] For each candidate path τ i Determine the car following the preceding vehicle p on this path. i and the corresponding front vehicle distance. If there are no surrounding vehicles within the preceding vehicle tracking threshold, i.e., there is no preceding vehicle on the candidate path, then the vehicle will follow the preceding vehicle p.i Set it to -1. Define the set of cars following the car as...

[0143] In one embodiment of this application, the real-time construction of the finite-time optimal control problem includes:

[0144] Construct an objective function for the reference trajectory corresponding to the candidate path, wherein the objective function is used to determine the reference trajectory, and the objective function includes at least one of the following: reference trajectory tracking content, vehicle control content, and distance control content of the following vehicle;

[0145] The real-time construction of the finite-time optimal control problem also includes:

[0146] The constraint information constrains the target vehicle, wherein the constraint information includes at least one of the following: dynamic model conditions of the target vehicle, control quantity constraint conditions, following distance conditions, safety constraint conditions between the target vehicle and surrounding vehicles, and safety constraint conditions between the target vehicle and the road.

[0147] Specifically, the reference trajectory tracking content in the objective function can specify the target vehicle's tracking performance on the reference trajectory; the vehicle control content can specify the control magnitude of the target vehicle; and the control distance content for following the preceding vehicle can specify the distance between the target vehicle and the following vehicle.

[0148] The dynamic model conditions of the target vehicle can be the state equations of the target vehicle transitioning from the current state to the next state under the current control input. Control input constraints represent physical constraints in the real world. The following distance condition is the distance between the target vehicle and the vehicle in front. The safety constraints with surrounding vehicles are the safety conditions under which the target vehicle will not collide with surrounding vehicles under control commands. The safety constraints with the road are the constraints on the distance between the target vehicle and the road.

[0149] In one embodiment of this application, wherein,

[0150] The i-th reference trajectory corresponding to the constructed candidate path The objective function is given by the following formula (5):

[0151]

[0152] Combination Figure 8 ,in, The control sequence is N, which is the optimization variable of the objective function. p Let k be the total number of steps in the prediction time domain for the control command in the finite-time optimal control problem, k∈{0,...,N}. p -1}, For the i-th reference trajectory From o i The reference trajectory for the initial k-th step (the k-th step of the control command), x k To predict the target vehicle's state variables at the k-th step in the time domain, including the target vehicle's x-coordinate, yaw angle, longitudinal velocity, lateral velocity, and yaw rate, x0 represents the target vehicle's current state variable, Q is the state variable weight matrix, and u... k To predict the target vehicle control input at the k-th step in the time domain, which includes the longitudinal acceleration and front wheel steering angle of the target vehicle, R is the control input weight matrix. To predict the target vehicle and candidate path τ at the k-th step in the time domain i The distance between the car in front and the car on the road, d des Let S be the desired following distance, and J be the scaling factor of the target vehicle. bias A fixed bias is used to reward the fuel economy and safety benefits of following the vehicle in front. This is an indicator function used to indicate the car-following target and the fixed bias.

[0153] Accordingly, the dynamic model conditions of the target vehicle are as follows: (6)

[0154] x k+1 =f ego (x k u k (6)

[0155] Among them, f ego It is a nonlinear mapping function that reflects the transition relationship of the target vehicle from the state variable at step k to the state variable at step k+1 under the control variable at step k.

[0156] The control quantity constraint condition is as follows: (7)

[0157]

[0158] in, The feasible set of control variables includes the upper and lower bounds of each output control variable, reflecting the physical constraints in the real world.

[0159] The following car distance condition is given by the following formula (8):

[0160]

[0161] in, p represents the preceding car on the i-th candidate path. i The state quantity at step k, This is a non-linear mapping, representing the distance between the target vehicle and the preceding vehicle at step k. If there is no preceding vehicle, the following distance condition is ignored in the objective function.

[0162] The safety constraints between the target vehicle and the surrounding vehicles are given by the following formula (9):

[0163] h(x k x surr，k )≥d safe (9)

[0164] Where, x surr，k Let X be the state quantity of the surrounding vehicles at the k-th step in the prediction time domain, which is determined by the predicted trajectory X of the surrounding vehicles. surr As determined, h(x) k x surr，k ) is a nonlinear mapping that reflects the distance between the target vehicle and the surrounding vehicles at the k-th step in the prediction time domain.

[0165] The safety constraint between the target vehicle and the road is given by the following formula (10):

[0166] h(x k x road，k )≥d safe (10)

[0167] Where, x road，k To predict the road information at the k-th step in the time domain, h(x) k，xroad，k ) is a nonlinear mapping that reflects the distance between the target vehicle and the road at the k-th step in the prediction time domain. It can be regarded as the shortest distance between the centroid of the target vehicle and the edge curve of the road.

[0168] In one embodiment of this application, the step of constructing a finite-time optimal control problem in real time for the candidate path and the car-following vehicle, and using the finite-time optimal control problem to determine the target driving trajectory corresponding to the target vehicle and the corresponding target car-following vehicle, includes the following steps:

[0169] The rolling time-domain optimization method is used to solve the objective function, dynamic model conditions of the target vehicle, control quantity constraints, following distance conditions, safety constraints with surrounding vehicles, and safety constraints with the road in the finite-time domain optimal control problem, thereby obtaining the optimal objective function corresponding to the i-th reference trajectory. Optimal control sequence and optimal state sequence The rolling time-domain optimization method can utilize the prediction model to perform real-time optimization based on the data at the current moment.

[0170] The objective function for the optimal tracking trajectory obtained by solving is shown in formula (4):

[0171]

[0172] Select the optimal path τ * and the corresponding optimal control sequence and optimal state sequence

[0173] Define the target vehicle path projection set as O = {τ} * o *}, where o * It is the nearest projection point on the optimal path.

[0174] The target vehicle's path projection set O and optimal state sequence X are transmitted to surrounding vehicles through the network platform. This allows surrounding vehicles to control their own driving based on the target vehicle's path projection set O and optimal state sequence X, thereby maintaining overall coordination among multiple vehicles in multiple lanes. (See also...) Figure 9 .

[0175] In one embodiment of this application, such as Figure 4 As shown and combined Figure 9 The method further includes the following steps:

[0176] S500, based on the travel target of the target vehicle, determine whether the target vehicle has reached the destination;

[0177] S600, if it is determined that the target vehicle has not reached the destination, the static environment information is reacquired to redetermine the target driving trajectory corresponding to the target vehicle and the corresponding target following vehicle.

[0178] Specifically, if the target vehicle determines that it has reached the destination, it can terminate autonomous driving. If it determines that it has not reached the destination, it can autonomously form up and switch formations, and repeat the control steps of the autonomous vehicle cooperative formation control method in this embodiment, including reacquiring the static environment information, and then redetermining the target vehicle's target driving trajectory and the corresponding target following vehicle.

[0179] The following section provides an overall description of the entire cooperative platooning control method for autonomous vehicles, using a specific embodiment as an example. Figure 9 The steps are as follows: Starting from the initial position, the target vehicle needs to have a map pre-stored in the control system to determine the vehicle's driving target and obtain the initial state of the target vehicle. First, static trajectory planning is performed, and a reference trajectory set Π is obtained based on the map and the target vehicle's state. Second, the workshop queue order is determined, and the surrounding vehicle path projection set O is obtained from the network platform. surrFirst, determine the set P of vehicles following each candidate path. Finally, perform dynamic tracking optimization based on the predicted trajectory X of the surrounding vehicles. surr Constructing an optimal tracking control problem for each reference trajectory Obtain the optimal control sequence U and the optimal state sequence X. The first step in implementing the optimal control sequence U is... The target vehicle's path projection set O and optimal state sequence X are transmitted to the connected platform. If the target vehicle reaches the global endpoint, the control process ends; otherwise, it restarts from the static path planning process. During the journey, the target vehicle can freely choose its tracking trajectory and follow the vehicle in front, thus enabling autonomous platooning and formation switching.

[0180] Based on the same inventive concept, embodiments of this application also provide a cooperative platooning control device for autonomous vehicles, such as... Figure 10 As shown, it includes:

[0181] The acquisition module is configured to determine the vehicle driving information of the target vehicle within a preset geographical range.

[0182] For example, the preset geographical range can be a pre-defined geographic location, which can exist in the form of a pre-stored map, such as pre-stored in the control unit (or control system) of the target vehicle. The pre-stored map can be set according to the actual application scenario of the target vehicle, or it can be set based on empirical data. For example, the preset geographical range can be set according to the current location of the target vehicle and its travel target.

[0183] On one hand, within a preset geographical area, the acquisition module can determine the relevant vehicle driving information, including initial location, destination, and approximate path. This information can be obtained using appropriate positioning devices, such as multiple vehicle-mounted positioning devices. On the other hand, the acquisition module can also obtain the vehicle driving information from other auxiliary devices.

[0184] The processing module is configured to determine the reference driving information of the target vehicle based on the vehicle driving information and the static environment information obtained within the preset geographical range, wherein the static environment information includes road information of multiple lanes and the reference driving information includes at least one candidate path of the target vehicle.

[0185] Based on the obtained surrounding vehicle driving information of the target vehicle, the reference driving information is used to determine the following vehicle corresponding to the target vehicle on each candidate path.

[0186] For the candidate path and the car following ahead, a finite-time optimal control problem is constructed in real time to determine the target driving trajectory of the target vehicle and the corresponding target car following ahead. The finite-time optimal control problem includes constraint information that constrains the target vehicle.

[0187] For example, the static environmental information within a preset geographical area may include road information for multiple lanes within that area. The target vehicle and surrounding vehicles can travel collaboratively on these multiple roads, including going straight and changing lanes. The road information may include parameters such as the number, shape, distance, location, and size of roads. The processing module can determine the target vehicle's reference driving information based on its initial position, travel target, approximate path, and other vehicle driving information, as well as this static environmental information. In other words, reference driving information can be determined based on the target vehicle's current state and road information, thereby providing a reference for the target vehicle's travel.

[0188] In this embodiment, the reference driving information includes at least one candidate path for the target vehicle and a desired reference speed. The candidate path corresponds to a reference trajectory, and the processing module can select a trajectory from the reference trajectory for the target vehicle to drive. In one embodiment, the data for each candidate path can be a series of neighboring points. In another embodiment, the reference trajectory can include the candidate path and driving speed parameters.

[0189] In one embodiment, a static trajectory planning layer can be constructed in the control system of the target vehicle. The processing module can implement the above steps through the static trajectory planning layer, including determining the reference driving information of the target vehicle based on the vehicle driving information and the obtained static environment information.

[0190] In this embodiment, the surrounding vehicle driving information refers to the driving-related information of the surrounding vehicles, such as their driving trajectories and speeds. This surrounding vehicle driving information can be obtained from the surrounding vehicles by the vehicle-to-everything (V2X) network platform and then sent to the target vehicle.

[0191] The processing module accurately determines the leading vehicle on each candidate path based on information such as the driving trajectory and speed of surrounding vehicles. Each leading vehicle can be considered as a candidate vehicle that the target vehicle can follow.

[0192] Specifically, when surrounding vehicles are traveling within a preset geographical area, they can move at a specific speed and switch between a single road or multiple roads. The surrounding vehicle travel information includes various relevant information about the surrounding vehicles in that state. The processing module can combine its own candidate paths and travel speed parameters with the acquired surrounding vehicle travel information to determine the following vehicle on each candidate path, thus providing the data basis for selecting the target following vehicle.

[0193] In one embodiment, a workshop queue order decision layer can be constructed in the control system of the target vehicle. The processing module can implement the above steps through the workshop queue order decision layer, including determining the following vehicle corresponding to the target vehicle on each candidate path based on the acquired surrounding vehicle driving information and using reference driving information.

[0194] In this embodiment, for example, the role of the finite-time optimal control problem is to derive a specific optimal system control law by constructing a performance index function in the finite-time domain and taking its extreme value for an existing system model.

[0195] The finite-time optimal control problem includes constraint information (or constraint conditions) that constrain a target vehicle. In one embodiment, the constraint information may include at least one of the following: dynamic model conditions of the target vehicle, control quantity constraints, following distance conditions, safety constraints with surrounding vehicles, and safety constraints with the road.

[0196] Finite-time optimal control problems can also include an objective function for a reference trajectory, which includes the following performance metrics: reference trajectory tracking performance, control output performance, and following performance of the vehicle ahead.

[0197] The processing module can use a real-time constructed finite-time optimal control problem to constrain candidate paths and the following vehicles on each candidate path. This includes: constraining the candidate paths and the following vehicles on each candidate path using the target vehicle's dynamic model conditions, control quantity constraints, following distance conditions, safety constraints with surrounding vehicles, and safety constraints with the road; and using an objective function to define the target vehicle's target trajectory and the corresponding target following vehicles. This allows for the determination of the target vehicle's target trajectory and the corresponding target following vehicles.

[0198] Thus, the target vehicle can travel along the target trajectory and follow the vehicle in front of it, achieving multi-vehicle cooperative driving while ensuring safety and meeting the requirements of multi-lane driving. Furthermore, the target vehicle can send its driving information to the network platform, which can then share this information with other vehicles (including those around the target vehicle) for their use.

[0199] In one embodiment, the control system of the target vehicle can construct a dynamic trajectory optimization and tracking layer, which can implement the above steps, including determining the target driving trajectory and the corresponding target following vehicle based on the candidate path and the following vehicle, using a finite-time-domain optimal control problem constructed in real time.

[0200] In one embodiment of this application, the processing module is further configured as follows:

[0201] Determine the current location of the target vehicle, as well as the number of lanes and road structure information of the lanes that are at least partially passable;

[0202] The candidate path is determined based on the current location, the number of lanes, and / or the road structure information.

[0203] In one embodiment of this application, the processing module is further configured as follows:

[0204] The centerline of each lane is selected as the pre-driving line, and the pre-driving line is discretized in equal space.

[0205] The discretized pre-driving line is set as τ. i Based on the vehicle's traffic regulations and driving habits, a desired reference speed is determined for each candidate route.

[0206] Based on the desired reference speed, the set of reference trajectories corresponding to the candidate paths is defined as follows: Where, N lane This represents the total number of lanes that are partially passable.

[0207] In one embodiment of this application, the processing module is further configured as follows:

[0208] The vehicle driving information of the surrounding vehicles in each of the candidate paths is obtained through the network platform.

[0209] On each candidate path, determine the distance between the projected position of the target vehicle and the projected positions of the surrounding vehicles;

[0210] If the distance satisfies the first distance condition, the nearest surrounding vehicle on each candidate path to the target vehicle is identified as the following vehicle corresponding to the target vehicle.

[0211] In one embodiment of this application, the processing module is specifically configured as follows:

[0212] Based on the current location of the target vehicle, extract a set of reference trajectories Π corresponding to the candidate path from the preset geographical range;

[0213] Based on the reference trajectory set Π, all the candidate paths τ i Projection is performed to determine the projection points o of the reference trajectory set Π on each of the candidate paths. i，ego i = 1: N lane ;

[0214] The set of surrounding vehicle path projections received by the target vehicle is defined as follows: Where N surr O is the number of surrounding vehicles. j ={τ i o i，j} is the projection set of the candidate paths transmitted by the network platform for the j-th surrounding vehicle, τ i For the candidate path where the j-th surrounding vehicle is currently located, o i，j For the target vehicle on the candidate path τ i Projection points on;

[0215] For each candidate path τ of the target vehicle i If the surrounding vehicle j is located on the candidate path τ i If the target vehicle and the surrounding vehicles are compared on the candidate path τ, then... i Projection point o on i，ego and o i，j ,like:

[0216] d safe ≤o i，j -o i，ego ≤d bound

[0217] Then the surrounding vehicles j are considered as the target vehicle on the candidate path τ. i The candidate for the lead vehicle, of which d bound It is the vehicle-to-the-front (D) tracking threshold. safe The required safe distance between the target vehicle and the surrounding vehicles;

[0218] Analyze all candidate paths τ in the same context i The vehicle closest to the target vehicle is identified as the vehicle on the candidate path τ. i The target vehicle mentioned above corresponds to the following vehicle p. i The corresponding following distance is

[0219]

[0220]

[0221] If there are no surrounding vehicles within the preceding vehicle tracking threshold, then the vehicle following the preceding vehicle p... i Set it to -1, and define the set of cars following the leader as...

[0222] In one embodiment of this application, the device further includes a construction module configured to:

[0223] Construct an objective function for the reference trajectory corresponding to the candidate path, wherein the objective function is used to determine the reference trajectory, and the objective function includes at least one of the following: reference trajectory tracking content, vehicle control content, and distance control content of the following vehicle;

[0224] The build module is also configured as follows:

[0225] The constraint information constrains the target vehicle, wherein the constraint information includes at least one of the following: dynamic model conditions of the target vehicle, control quantity constraint conditions, following distance conditions, safety constraint conditions between the target vehicle and surrounding vehicles, and safety constraint conditions between the target vehicle and the road.

[0226] In one embodiment of this application, wherein,

[0227] The i-th reference trajectory corresponding to the constructed candidate path The objective function is:

[0228]

[0229] in, The control sequence is N, which is the optimization variable of the objective function. p Let k be the number of prediction time-domain steps in the finite-time optimal control problem, k∈{0,...,N}. p -1}, For the i-th reference trajectory From o i The reference trajectory at the initial k-th step, x k To predict the target vehicle state at the k-th step in the time domain, let x0 be the current state of the target vehicle, Q be the state weight matrix, and u be the weight matrix. k To predict the target vehicle control quantity at the k-th step in the time domain, R is the control quantity weight matrix. To predict the target vehicle and candidate path τ at the k-th step in the time domain i The distance between the car in front and the car on the road, d des Let S be the desired following distance, and J be the scaling factor of the target vehicle. biasThis is a fixed bias used to reward the fuel economy and safety benefits of following the vehicle in front. This is an indication function used to indicate the car-following target and the fixed bias;

[0230] Accordingly, the dynamic model conditions of the target vehicle are:

[0231] x k+1 =f ego (x k u k )

[0232] Among them, f ego It is a nonlinear mapping function that reflects the transition relationship of the target vehicle from the state quantity at step k to the state quantity at step k+1 under the control quantity at step k;

[0233] The control quantity constraint condition is as follows:

[0234]

[0235] in, The feasible set of control variables includes the upper and lower bounds of each output control variable, reflecting the physical constraints in the real world;

[0236] The following distance condition is:

[0237]

[0238] in, p represents the preceding car on the i-th candidate path. i The state at step k, It is a nonlinear mapping, representing the distance between the target vehicle and the following vehicle at the k-th step in the prediction time domain;

[0239] The safety constraints between the target vehicle and the surrounding vehicles are as follows:

[0240] h(x k x surr，k )≥d safe

[0241] Where, x surr，k Let X be the state quantity of the surrounding vehicles at the k-th step in the prediction time domain, which is determined by the predicted trajectory X of the surrounding vehicles. surr As determined, h(x) k x surr，k () is a nonlinear mapping that reflects the distance between the target vehicle and the surrounding vehicles at the k-th step in the prediction time domain;

[0242] The safety constraints between the target vehicle and the road are as follows:

[0243] h(xk x road，k )≥d safe

[0244] Where, x road，k To predict the road information at the k-th step in the time domain, h(x) k x road，k ) is a nonlinear mapping that reflects the distance between the target vehicle and the road at the k-th step in the prediction time domain. It can be regarded as the shortest distance between the centroid of the target vehicle and the edge curve of the road.

[0245] In one embodiment of this application, the processing module is further configured as follows:

[0246] The rolling time-domain optimization method is used to solve the objective function, dynamic model conditions of the target vehicle, control quantity constraints, following distance conditions, safety constraints with surrounding vehicles, and safety constraints with the road in the finite-time domain optimal control problem, thereby obtaining the optimal objective function corresponding to the i-th reference trajectory. Optimal control sequence and optimal state sequence

[0247] The objective function for constructing the optimal tracking trajectory is solved by:

[0248]

[0249] Select the optimal path τ * and the corresponding optimal control sequence and optimal state sequence

[0250] Define the target vehicle path projection set as O = {τ} * o *}, where o * The nearest projection point on the optimal path;

[0251] The target vehicle path projection set O and the optimal state sequence X are transmitted to the surrounding vehicles through the network platform.

[0252] In one embodiment of this application, the processing module is further configured as follows:

[0253] Based on the target vehicle's travel target, determine whether the target vehicle has reached its destination;

[0254] If it is determined that the target vehicle has not reached the destination, the static environment information is reacquired to redetermine the target vehicle's target driving trajectory and the corresponding target following vehicle.

[0255] Based on the same inventive concept, embodiments of this application also provide an electronic device having a memory and a processor, wherein the memory stores an executable program, and the processor executes the executable program to implement the steps of the method described in any embodiment.

[0256] Based on the same inventive concept, embodiments of this application also provide a storage medium carrying one or more computer programs, which, when executed by a processor, implement the steps of the method described in any embodiment.

[0257] The storage medium in this embodiment may be included in an electronic device / system; or it may exist independently and not assembled into an electronic device / system. The storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0258] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0259] The above embodiments are merely exemplary embodiments of this application and are not intended to limit this application. The scope of protection of this application is defined by the claims. Those skilled in the art can make various modifications or equivalent substitutions to this application within its substance and scope of protection, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of this application.

Claims

1. A cooperative formation control method for autonomous vehicles, characterized in that, include: Within a preset geographical area, determine the vehicle's driving information; Based on the vehicle driving information and the obtained static environment information within the preset geographical range, the reference driving information of the target vehicle is determined, wherein the static environment information includes road information of multiple lanes, and the reference driving information includes at least one candidate path of the target vehicle. Based on the obtained surrounding vehicle driving information of the target vehicle, and using the reference driving information, the following vehicle corresponding to the target vehicle on each candidate path is determined, including: The vehicle driving information of the surrounding vehicles in each of the candidate paths is obtained through the network platform. On each candidate path, determine the distance between the projected position of the target vehicle and the projected positions of the surrounding vehicles; If the distance satisfies the first distance condition, the nearest surrounding vehicle on each candidate path to the target vehicle is identified as the following vehicle corresponding to the target vehicle. Specifically, determining the following vehicle corresponding to the target vehicle on each candidate path based on the obtained surrounding vehicle driving information and the reference driving information includes: Based on the current location of the target vehicle, a set of reference trajectories corresponding to the candidate paths is extracted from the preset geographical area. ; Based on the reference trajectory set to all the candidate paths Projection is performed to determine the reference trajectory set. Projection points on each of the candidate paths , ; The set of surrounding vehicle path projections received by the target vehicle is defined as follows: ,in The number of surrounding vehicles. For the first The projection set of the candidate paths transmitted by surrounding vehicles through the network platform. For the first The candidate paths currently occupied by the surrounding vehicles. For the target vehicle on the candidate path Projection points on; For each candidate path of the target vehicle If surrounding vehicles Located in the candidate path Then, the target vehicle and the surrounding vehicles are compared on the candidate path. Projection point on and ,like: Then the surrounding vehicles The target vehicle is considered to be on the candidate path Among the candidate precedents, It is the forward vehicle tracking threshold. The required safe distance between the target vehicle and the surrounding vehicles; Analyze all candidate paths in the same context The vehicle closest to the target vehicle is identified as the vehicle on the candidate path. The target vehicle mentioned above corresponds to the following vehicle. The corresponding following distance is ; If there are no surrounding vehicles within the preceding vehicle tracking threshold, then follow the preceding vehicle. Set as Define the set of cars following ahead as ; For the candidate path and the car following ahead, a finite-time optimal control problem is constructed in real time to determine the target driving trajectory of the target vehicle and the corresponding target car following ahead. The finite-time optimal control problem includes constraint information that constrains the target vehicle.

2. The method according to claim 1, characterized in that, The step of determining the reference driving information of the target vehicle based on the vehicle driving information and the obtained static environmental information within the preset geographical range includes: Determine the current location of the target vehicle, as well as the number of lanes and road structure information of the lanes that are at least partially passable; The candidate path is determined based on the current location, the number of lanes, and / or the road structure information.

3. The method according to claim 2, characterized in that, Determining the candidate path based on the current location, the number of lanes, and / or the road structure information includes: The centerline of each lane is selected as the pre-driving line, and the pre-driving line is discretized in equal space. Set the discretized pre-driving line as Based on the vehicle's traffic regulations and driving habits, a desired reference speed is determined for each candidate route. ; Based on the desired reference speed, the set of reference trajectories corresponding to the candidate paths is defined as follows: ,in, This represents the total number of lanes that are partially passable.

4. The method according to claim 1, characterized in that, The real-time construction of the finite-time optimal control problem includes: Construct an objective function for the reference trajectory corresponding to the candidate path, wherein the objective function is used to determine the reference trajectory, and the objective function includes at least one of the following: reference trajectory tracking content, vehicle control content, and distance control content of the following vehicle; The real-time construction of the finite-time optimal control problem also includes: The constraint information constrains the target vehicle, wherein the constraint information includes at least one of the following: dynamic model conditions of the target vehicle, control quantity constraint conditions, following distance conditions, safety constraint conditions between the target vehicle and surrounding vehicles, and safety constraint conditions between the target vehicle and the road.

5. The method according to claim 4, characterized in that, in, The candidate path constructed corresponds to the first Reference trajectory The objective function is: in, The control sequence is the optimization variable of the objective function. Let be the number of steps in the prediction time domain in the finite-time optimal control problem. , For the first Reference trajectory Zhong Cong The beginning of the first The reference trajectory of the step, To predict the first in the time domain The target vehicle state quantity of the step. The current state quantity of the target vehicle. The state variable weight matrix, To predict the first in the time domain The target vehicle control quantity for the step. For the control quantity weight matrix, To predict the first in the time domain Target vehicle and candidate path The distance between the car and the car in front. To achieve the desired following distance, This is the scaling factor for the car-following target. This is a fixed bias used to reward the fuel economy and safety benefits of following the vehicle in front. This is an indication function used to indicate the car-following target and the fixed bias; Accordingly, the dynamic model conditions of the target vehicle are: in, It is a nonlinear mapping function, reflecting the target vehicle in the first... Step control quantity from the first Step state quantity to the first The transition relationship of step state variables; The control quantity constraint condition is as follows: in, The feasible set of control variables includes the upper and lower bounds of each output control variable, reflecting the physical constraints in the real world; The following distance condition is: in, Representing the Following the lead of the candidate path In the State variables at step time This is a nonlinear mapping, representing the target vehicle and the following vehicle in the prediction time domain at the th... The distance between steps; The safety constraints between the target vehicle and the surrounding vehicles are as follows: in, For the surrounding vehicles in the prediction time domain, the first The state variables of the step are predicted by the cyclic vehicle trajectory. It has been determined that It is a nonlinear mapping that reflects the relationship between the target vehicle and the surrounding vehicles in the prediction time domain. The distance between steps; The safety constraints between the target vehicle and the road are as follows: in, To predict the first in the time domain Road information for walking It is a nonlinear mapping that reflects the relationship between the target vehicle and the road in the prediction time domain. The distance between steps.

6. The method according to claim 5, characterized in that, The step involves constructing a finite-time optimal control problem in real time for the candidate path and the car-following vehicle, and using this finite-time optimal control problem to determine the target driving trajectory of the target vehicle and the corresponding target car-following vehicle, including: The objective function, dynamic model conditions of the target vehicle, control quantity constraints, following distance conditions, safety constraints with surrounding vehicles, and safety constraints with the road are solved using the rolling time-domain optimization method in the finite-time-domain optimal control problem, to obtain the corresponding... The optimal objective function of the reference trajectory Optimal control sequence and the optimal state sequence ; The objective function for constructing the optimal tracking trajectory is solved by: Select the optimal path and the corresponding optimal control sequence and the optimal state sequence ; Define the target vehicle path projection set as ,in, The nearest projection point on the optimal path; The target vehicle path projection set is obtained through the network platform. With the optimal state sequence Transmitted to the surrounding vehicles.

7. The method according to claim 1, characterized in that, The method further includes: Based on the target vehicle's travel target, determine whether the target vehicle has reached its destination; If it is determined that the target vehicle has not reached the destination, the static environment information is reacquired to redetermine the reference driving trajectory of the target vehicle and the corresponding candidate following vehicles.

8. A cooperative platooning control device for autonomous vehicles, characterized in that, include: The acquisition module is configured to determine the vehicle driving information of the target vehicle within a preset geographical range; The processing module is configured to determine the reference driving information of the target vehicle based on the vehicle driving information and the static environment information obtained within the preset geographical range, wherein the static environment information includes road information of multiple lanes and the reference driving information includes at least one candidate path of the target vehicle. Based on the obtained surrounding vehicle driving information of the target vehicle, and using the reference driving information, the following vehicle corresponding to the target vehicle on each candidate path is determined, including: The vehicle driving information of the surrounding vehicles in each of the candidate paths is obtained through the network platform. On each candidate path, determine the distance between the projected position of the target vehicle and the projected positions of the surrounding vehicles; If the distance satisfies the first distance condition, the nearest surrounding vehicle on each candidate path to the target vehicle is identified as the following vehicle corresponding to the target vehicle. Specifically, determining the following vehicle corresponding to the target vehicle on each candidate path based on the obtained surrounding vehicle driving information and the reference driving information includes: Based on the current location of the target vehicle, a set of reference trajectories corresponding to the candidate paths is extracted from the preset geographical area. ; Based on the reference trajectory set to all the candidate paths Projection is performed to determine the reference trajectory set. Projection points on each of the candidate paths , ; The set of surrounding vehicle path projections received by the target vehicle is defined as follows: ,in The number of surrounding vehicles. For the first The projection set of the candidate paths transmitted by surrounding vehicles through the network platform. For the first The candidate paths currently occupied by the surrounding vehicles. For the target vehicle on the candidate path Projection points on; For each candidate path of the target vehicle If surrounding vehicles Located in the candidate path Then, the target vehicle and the surrounding vehicles are compared on the candidate path. Projection point on and ,like: Then the surrounding vehicles The target vehicle is considered to be on the candidate path Among the candidate precedents, It is the forward vehicle tracking threshold. The required safe distance between the target vehicle and the surrounding vehicles; Analyze all candidate paths in the same context The vehicle closest to the target vehicle is identified as the vehicle on the candidate path. The target vehicle mentioned above corresponds to the following vehicle. The corresponding following distance is ; If there are no surrounding vehicles within the preceding vehicle tracking threshold, then follow the preceding vehicle. Set as Define the set of cars following ahead as ; For the candidate path and the car following ahead, a finite-time optimal control problem is constructed in real time to determine the target driving trajectory of the target vehicle and the corresponding target car following ahead. The finite-time optimal control problem includes constraint information that constrains the target vehicle.

9. A storage medium, characterized in that, The storage medium carries one or more computer programs, which, when executed by a processor, implement the steps of the method as described in any one of claims 1 to 7.

Images