Centralized vehicle cooperative decision ramp merging and separating method and device and cloud

Abstract

The application discloses a centralized vehicle cooperative decision ramp merging and separating method and device and a cloud, and belongs to the technical field of automatic driving, and comprises the following steps: planning a global path of a target vehicle by using a rapid expansion random tree according to current position information and a destination address of the target vehicle; obtaining positions of the target vehicle at each time point on the global path by combining the global path with an expected driving speed, so that other vehicles plan a global path by combining the positions of the target vehicle at each time point on the global path; and if the global path planning time of the target vehicle is later than the global path planning time of other vehicles when the target vehicle reaches a target section, and a planned ramp merging or separating route has been occupied by other vehicles, then a cooperative control strategy is generated based on sensing data of a roadside device, running data of the vehicle and sensing data of a vehicle end sensor. The application maximizes the driving efficiency of a single intelligent vehicle and the traffic efficiency of a road under the premise of ensuring the driving safety of the intelligent vehicle.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of automatic driving, and more particularly relates to a centralized vehicle cooperative decision ramp merging and separating method and device and a cloud. BACKGROUND

[0002] With the rapid development of economy, the proportion of private car ownership is rapidly rising, the number of motor vehicles is increasing year by year and the increment is increasing; in addition, the urbanization process is accelerated, and the concentration of vehicles is increased. The increasing number of vehicles has made the road traffic system of many places increasingly heavy, while people's demand for more convenient transportation environment is increasing, and this contradiction has led to many traffic problems and even larger social contradictions. The increasingly serious traffic problems have attracted widespread attention from all sectors of society and have become a focus of public concern.

[0003] Path planning has a wide range of applications in the field of transportation, such as GPS navigation, path planning in GIS systems, and urban road network planning. Since the birth of path planning algorithms, they have gradually evolved from traditional algorithms to intelligent algorithms. Path planning algorithms are diverse, each with its own characteristics, and should be selected according to the actual application situation when applied. At present, the overall intelligence of vehicles has been greatly improved, and the intelligent road transportation has become more and more advanced. In order to provide users with a more comfortable and safe driving experience and to alleviate traffic pressure and reduce traffic accidents, it is necessary to combine intelligent transportation and intelligent vehicles.

[0004] Referring to the patent application with publication number CN115061466A, a vehicle-road cooperation automatic driving method, roadside equipment, cloud control platform and system are disclosed, the method comprising: in response to detecting a first vehicle, collecting first driving information and road condition information of the first vehicle, the first driving information including positioning information of the first vehicle; transmitting the first driving information and the road condition information to a server; receiving first vehicle control information for the first vehicle, the first vehicle control information being generated based on the first driving information and the road condition information; transmitting the first vehicle control information to the first vehicle, the first driving information being collected by the roadside equipment, and the first vehicle control information including at least one of the following: behavior decision information to indicate the action of the first vehicle; motion planning information to indicate the motion state and / or motion trajectory of the first vehicle; control instruction information for controlling the actuators of the first vehicle.

[0005] With reference to the patent application with publication number CN113734202A, the method can include the following steps performed in the cloud: obtaining planning path information and real-time state information of each vehicle in the Internet of Vehicles; determining a coordination zone between the subject vehicle and other vehicles based on the planning path information of each vehicle; determining a vehicle set that needs to be coordinated with the subject vehicle based on the coordination zone and the real-time state information; and determining the passing right of each vehicle based on the vehicle set. The present application improves the low traffic efficiency caused by mutual avoidance of vehicles and the driving safety problem caused by mutual rush of vehicles through multi-vehicle coordination processing in the cloud, thereby improving traffic efficiency and improving vehicle driving safety.

[0006] However, the above scheme has the following technical problems: only the working method and idea of the collaborative decision system are given, and the system collaborative decision process and the global path planning method of the intelligent vehicle are not given; the data interaction relationship among the vehicle, the road, and the cloud is given, but not specifically applied to the specific application scenario and working condition of the collaborative decision, and the method for judging whether the global path exists collision safety.

[0007] Therefore, how to improve the driving efficiency of a single intelligent vehicle and the traffic efficiency of a road to the greatest extent under the premise of ensuring the driving safety of the intelligent vehicle is a technical problem that needs to be solved at present. SUMMARY

[0008] In view of the above defects or improvement needs of the prior art, the present application proposes a centralized multi-agent vehicle collaborative decision ramp merging and separation method, device and cloud, which maximizes the driving efficiency of a single intelligent vehicle and the traffic efficiency of a road under the premise of ensuring the driving safety of the intelligent vehicle in the collaborative decision mode.

[0009] To achieve the above-mentioned purpose, according to one aspect of the present application, a centralized vehicle collaborative decision ramp merging and separation method is provided, comprising:

[0010] Planning a global path of the target vehicle using a rapid expansion random tree based on the position information and destination address of the target vehicle at the current time;

[0011] Deriving the position of the target vehicle at each time point on the global path by combining the generated global path with the expected driving speed, so that other vehicles plan a global path in combination with the position of the target vehicle at each time point on the global path;

[0012] If the global path planning time of the target vehicle is later than the global path planning time of other vehicles when the target vehicle reaches the target road section, and the planned ramp merging or separation route has been occupied by other vehicles, a collaborative control strategy is generated based on the perception data of the roadside equipment in the control road section, the operation data of the vehicle, and the perception data of the vehicle-end sensor.

[0013] In some optional implementations, the step of planning the global path of the target vehicle using a fast expanding random tree based on the target vehicle's current location information and destination address includes:

[0014] Define the global path planning task space n represents the spatial dimension, dividing the global path planning task space into spaces containing obstacles. and blank areas Let P be the initial state point corresponding to the current position information of the target vehicle. start ∈X free The destination address corresponds to the target point P. end ∈X free ;

[0015] Point P start As the root node of the entire tree structure, a random point P is generated at a random location in the blank area. rand , with P rand Centered on P, traverse and search all nodes in the tree structure, calculating the relationship between these nodes and P. rand The Euclidean distance between them is used to select the node P with the smallest distance. near As the nearest node;

[0016] Let P near Point to P rand The direction of the point is the growth direction of the tree structure, causing the search tree to start from P. near A point grows a certain distance along the growth direction to obtain the next node in the tree structure, denoted as P. new Determine the target vehicle at point P. near To P new On the represented path, whether it collides with obstacles around the path; if it collides with an obstacle, then remove node P. new Then, generate a random point P again at a random location in the blank area. rand Perform a sampling search; if no collision occurs, then set P... new Points are added to the tree until a new node P is generated. new Reaching target point P end , or with the target point P end The distance between them is less than one unit growth step, from the target point P. end By tracing back a series of parent nodes in sequence, the final path is obtained.

[0017] In some alternative implementations, the determination of the target vehicle at point P... near To P new Whether the path collided with obstacles surrounding the path, including:

[0018] Consider the target vehicle as a circle with a diameter of length L; the path traversed by the center of this circle is point P. near To P new Given a path, determine if the path traversed by the center of the circle intersects with P. near To P new The target vehicle collides with obstacles around its path, where the value of L is the distance between the two points furthest from the target vehicle.

[0019] In some optional implementations, the generation of a collaborative control strategy based on sensing data from roadside equipment within the controlled road segment, vehicle operation data, and vehicle-mounted sensor sensing data includes:

[0020] If the target vehicle has a path conflict with the first vehicle, but not with other vehicles, and the target vehicle's deceleration will not affect the driving of other vehicles, while the first vehicle's deceleration will affect the driving of other vehicles, then the target vehicle's driving process will change to a process of first decelerating and then accelerating. When the first vehicle releases the right-of-way, the target vehicle will change lanes to the target lane and then leave or enter the main road.

[0021] In some optional implementations, the generation of a collaborative control strategy based on sensing data from roadside equipment within the controlled road segment, vehicle operation data, and vehicle-mounted sensor sensing data includes:

[0022] If the target vehicle and the first vehicle have a path conflict, but there is no path conflict with other vehicles, and the target vehicle's deceleration will affect the driving of other vehicles, while the first vehicle's deceleration will not affect the driving of other vehicles, then the first vehicle's driving process will change to first decelerate and then accelerate. The first vehicle actively releases the right-of-way, and the target vehicle accelerates or changes lanes at a constant speed to the target lane and then leaves or enters the main road.

[0023] In some optional implementations, the generation of a collaborative control strategy based on sensing data from roadside equipment within the controlled road segment, vehicle operation data, and vehicle-mounted sensor sensing data includes:

[0024] If the target vehicle has path conflicts with multiple vehicles, the target vehicle's speed is gradually increased. Then, global path planning is performed again based on the increased speed until a suitable global driving path is found. The final determined driving speed cannot exceed the maximum allowed driving speed.

[0025] In some optional implementations, the generation of a collaborative control strategy based on sensing data from roadside equipment within the controlled road segment, vehicle operation data, and vehicle-mounted sensor sensing data includes:

[0026] If the target vehicle and multiple vehicles have path conflicts, and increasing the target vehicle's desired speed to its maximum permissible speed still results in a conflict, then the target vehicle's speed is reduced. A second global travel path is then planned based on the desired speed, and the corresponding vehicle travel time T is calculated. limit ;

[0027] The target vehicle gradually reduces its speed, according to T limit The value is used to calculate the minimum driving speed v of the target vehicle. limit If the target vehicle decelerates to v limit If a suitable global path has already been found, then follow v limit If the target vehicle does not proceed, it will proceed according to the second global driving path.

[0028] According to another aspect of the present invention, a centralized vehicle cooperative decision-making ramp merging and separating device is provided, comprising:

[0029] The global path planning module is used to plan the global path of the target vehicle based on its current location information and destination address using a fast expanding random tree. The generated global path is combined with the expected driving speed to determine the position of the target vehicle on the global path at each time point, so that other vehicles can plan their global paths based on the position of the target vehicle on the global path at each time point.

[0030] The collaborative control module is used to generate a collaborative control strategy based on the perception data of roadside equipment within the controlled road segment, the vehicle's operation data, and the perception data of vehicle-side sensors when the global path planning time of the target vehicle is later than the global path planning time of other vehicles when it arrives at the target road segment, and when the planned ramp merging or separating routes are occupied by other vehicles.

[0031] According to another aspect of the invention, a cloud-based system including a centralized vehicle collaborative decision-making ramp merging and separating device is provided.

[0032] According to another aspect of the present invention, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0033] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects:

[0034] The centralized multi-agent vehicle collaborative decision-making ramp merging and separation method provided by this invention maximizes the driving efficiency of individual intelligent vehicles and the traffic efficiency of roads under the premise of ensuring the driving safety of intelligent vehicles in a collaborative decision-making mode. Attached Figure Description

[0035] Figure 1 This is a hardware system architecture diagram of a centralized multi-agent vehicle dynamic collaborative decision-making ramp merging and separation method provided in an embodiment of the present invention;

[0036] Figure 2 This is a functional schematic diagram of a centralized multi-agent vehicle dynamic collaborative decision-making and vehicle control system provided in an embodiment of the present invention;

[0037] Figure 3 This is a flowchart illustrating a centralized vehicle collaborative decision-making ramp merging and separation method provided in an embodiment of the present invention;

[0038] Figure 4 This is a schematic diagram of the tree structure generation process of the RRT algorithm provided in an embodiment of the present invention;

[0039] Figure 5 This is a schematic diagram illustrating the possibility of path collision provided by an embodiment of the present invention;

[0040] Figure 6 This is a schematic diagram of the road conditions for a target vehicle provided in an embodiment of the present invention;

[0041] Figure 7 This is a schematic diagram of a working condition provided in an embodiment of the present invention;

[0042] Figure 8 This is a schematic diagram of a second working condition provided in an embodiment of the present invention;

[0043] Figure 9 This is a flowchart of a global path generation process for an intelligent vehicle provided in an embodiment of the present invention;

[0044] Figure 10 This is a schematic diagram of an implementation process under a specific working condition provided in an embodiment of the present invention;

[0045] Figure 11 This is a schematic diagram of an implementation process for working condition two provided in an embodiment of the present invention;

[0046] Figure 12 This is a schematic diagram of a three-condition implementation process provided by an embodiment of the present invention;

[0047] Figure 13 This is a schematic diagram of a four-condition implementation process provided by an embodiment of the present invention;

[0048] Figure 14 This is a schematic diagram of a centralized vehicle collaborative decision-making ramp merging and separating device provided in an embodiment of the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0050] In the embodiments of the present invention, "first," "second," etc., are used to distinguish different objects, rather than to describe a specific order or sequence.

[0051] The hardware system required for the centralized vehicle group collaborative decision-making ramp merging and separation technology provided by this invention includes, for example: Figure 1 As shown: the cloud subsystem, roadside subsystem, and vehicle-to-vehicle (V2V) subsystem. The cloud subsystem, as the central hub for collaborative decision-making and control, makes decisions and controls vehicle behavior within the controlled road segment based on various data sources. The roadside subsystem, on the one hand, can perceive vehicle status and traffic conditions within the controlled road segment in real time, and on the other hand, acts as an information bridge between the cloud subsystem and the V2V subsystem. The V2V subsystem serves individual autonomous vehicles, possesses necessary sensing equipment, and controls the vehicle based on decisions made by the cloud subsystem.

[0052] Specifically, in this embodiment of the invention, the cloud subsystem mainly needs to provide the following functions:

[0053] ① Real-time data fusion and processing. The input data sources of the cloud subsystem are extensive, including not only real-time data such as the perception data of roadside equipment within its controlled road section, vehicle operation data and vehicle-end sensor perception data, and the operating status data of each subsystem and unit, but also basic data, historical data, external data source data (such as meteorological data), and manual control command data.

[0054] ② Real-time decision generation. Based on the aforementioned fused data, the cloud subsystem generates real-time control schemes, i.e., collaborative control strategies, that apply to all vehicles within its control range. This aims to achieve the highest possible traffic efficiency while ensuring the driving safety of intelligent connected vehicles, fully leveraging the advantages of the vehicle-road integrated system in the intelligent connected environment.

[0055] ③ Basic and Historical Data Storage. Generating vehicle collaborative decision-making schemes requires not only real-time input data but also a large amount of basic data such as road alignment, elevation, number of lanes, lane width, and shoulder information. Simultaneously, both input and output data need to be stored in real-time for subsequent scheme optimization and other purposes.

[0056] ④ Collaboration between cloud subsystems. On the one hand, to ensure the security of centralized vehicle group collaborative decision-making control systems, it is necessary to build mirrored cloud subsystems to avoid major traffic accidents caused by the failure of cloud subsystems performing tasks. When intelligent connected vehicles switch between various cloud subsystems, the handover of data and control between the preceding and following cloud subsystems needs to be done well. On the other hand, collaboration between cloud subsystems can also be applied to vehicle route selection, implementation of collaborative management and control measures, and rational allocation of traffic volume on various road segments in the road network.

[0057] Specifically, in this embodiment of the invention, the roadside subsystem mainly needs to provide the following functions:

[0058] ① In terms of perception, the roadside subsystem includes various sensing devices, including lidar, millimeter-wave radar, and video detectors. Within a set of roadside subsystems, different sensing devices can be deployed at different locations along the road segment at different densities to achieve multi-source overall perception of the road segment.

[0059] ② Regarding communication with the vehicle, the roadside subsystem and the vehicle can communicate in real time through C-V2X wireless connection technology to receive data from on-board sensing devices, vehicle operating status data, etc., and issue control decisions made by the cloud subsystem.

[0060] Specifically, in this embodiment of the invention, the vehicle terminal system mainly needs to provide the following functions:

[0061] The vehicle subsystem is the final executor of decisions in intelligent connected transportation systems. It takes the vehicle as its carrier and includes four main functional units: Electronic Control Unit (ECU), Sensing Unit, Network Communication Unit, and Chassis Electronic Control System.

[0062] like Figure 2 The diagram shown is a schematic diagram of a centralized multi-agent vehicle dynamic cooperative decision-making control system provided by an embodiment of the present invention. The multi-agent vehicle dynamic cooperative decision-making part is the focus, as shown below. Figure 3 As shown, it includes the following steps:

[0063] S1: Based on the target vehicle's current location information and destination address, use a fast expanding random tree to plan the target vehicle's global path;

[0064] S2: Combine the generated global path with the desired driving speed to determine the position of the target vehicle on the global path at each time point, so that other vehicles can plan their global paths based on the position of the target vehicle on the global path at each time point.

[0065] S3: If the global path planning time of the target vehicle is later than the global path planning time of other vehicles when it arrives at the target road segment, and the planned ramp merging or separating routes have been occupied by other vehicles, then a collaborative control strategy is generated based on the perception data of the roadside equipment in the controlled road segment, the vehicle's operation data, and the perception data of the vehicle-side sensors.

[0066] Specifically, in the embodiments of the present invention, each step can be broken down or combined into the following stages for implementation:

[0067] (1) In the first stage, the passenger starts the target vehicle. The passenger inputs the destination address through the target vehicle's HMI (Human Machine Interface). The cloud uses the Rapidly-exploring Random Tree (RRT) algorithm to plan the global path of the target vehicle based on the target vehicle's current location information and destination address.

[0068] The RRT algorithm continuously constructs a tree-like search structure in a multidimensional space, incrementally expanding into unexplored areas. Each vertex of the tree structure is a state node, and the line segment between two adjacent nodes represents the connection between the current state node and the previous state node. The tree structure generation process of the RT algorithm is as follows: Figure 4 As shown, the global path planning task space is first defined. n represents the spatial dimension; the task space can be divided into spaces containing obstacles. and blank areas The current location information of the target vehicle within the controlled road section is used as the initial state point P. start The initial state point is P. start ∈X free The destination address is the target point P. end The target point is P. end ∈X free There are obstacles in the controlled section of the road. Figure 4 (Middle black area).

[0069] (2) In the second stage, point P will be... start As the root node of the entire tree structure, a random point P is generated at a random location in the blank area. rand , with P rand Centered on P, traverse and search all nodes in the tree structure, calculating the relationship between these nodes and P. rand Sort the nodes by Euclidean distance and select the node P with the smallest distance. near As the nearest node, let P be the node from which the node originates. near Point to P rand The direction of the point is the growth direction of the tree structure. By setting the growth step size (denoted as l), the search tree can grow from P...near The point grows a certain distance along the growth direction (the value of l needs to be selected through experimentation; too large a growth step size will cause the algorithm to "jump" over obstacles with small cross-sections, while too short a step size will often weaken the algorithm's traversal and search capabilities, reducing the algorithm's planning efficiency), to obtain the next node in the tree structure, denoted as P. new Determine where the vehicle is traveling. near To P new On the represented path, whether it collides with obstacles around the path; if it collides with an obstacle, then remove the node P. new Then, generate a random point P again at a random location in the blank area. rand Perform a sampling search; if no collision occurs, then set P... new A node is added to the tree, and the algorithm iterates until a new node P is generated. new Reaching the target location, or when the distance to the target point is less than one unit growth step size l, indicates the end of the search process. The algorithm then backtracks from the target point to a series of parent nodes in sequence to obtain the final path.

[0070] (3) Third stage judgment point P near To P new To determine whether the path collides with nearby obstacles, the target vehicle can be viewed as a circle with a diameter of length L, where L is the distance between the two furthest points on the target vehicle, obtained through calibration. The path traversed by the center of this circle is point P. near To P new The path, such as Figure 5 The diagram illustrates whether there is a possibility of a collision along the path.

[0071] (4) In the fourth stage, the generated global path is combined with the passenger's expected travel speed to determine the position of the target vehicle on the global path at each time point. The state information of the target vehicle is recorded as V. n =(T n Q n ), Q n For T n Target vehicle V at any time n The location on a high-precision map can also be considered as T. n Time position Q m The obstacle at the location is the target vehicle V. n This recording method facilitates the planning of global paths for other intelligent vehicles, ensuring that there are no path conflicts between the intelligent vehicles.

[0072] (5) The fifth stage assumes that the global path planning time of target vehicle A is later than when it arrives at the destination vehicle. Figure 6The section shown represents the global path planning time for other vehicles. At this time, target vehicle A needs to exit the main road and enter the ramp. However, the planned route to exit the main road and enter the ramp has been occupied by other vehicles. If other vehicles insist on traveling along their originally planned routes, and target vehicle A also insists on leaving the main road along its globally planned route, a collision with other vehicles will occur. If vehicles slow down and wait for the right-of-way to be released, it may result in excessively long waiting times for their own vehicles, leading to an excessive reduction in their own travel efficiency and also an excessive reduction in the traffic efficiency of this section of road.

[0073] Regarding the situation that occurs in the fifth stage, there are several specific working conditions, and solutions are provided for each of them:

[0074] Scenario 1: Target vehicle A and vehicle B have a path conflict, but no path conflict with other vehicles. Target vehicle A's deceleration will not affect other vehicles; however, vehicle B's deceleration will affect other vehicles. In this scenario, target vehicle A's driving process changes to a deceleration-acceleration sequence. Once vehicle B releases its right-of-way, target vehicle A changes lanes to lane 1 and then leaves the main road. Figure 7 The path replanning and obstacle avoidance methods are the same as those in the second and third stages.

[0075] Scenario 2: Target vehicle A and vehicle B have a path conflict, but no path conflict with other vehicles. Target vehicle A's deceleration will affect the driving of other vehicles; vehicle B's deceleration will not affect the driving of other vehicles. In this scenario, vehicle B's driving process changes to first decelerating and then accelerating. Vehicle B actively relinquishes its right-of-way, and target vehicle A accelerates or changes lanes at a constant speed to lane 1 and then leaves the main road. Figure 8 The path replanning and obstacle avoidance methods are the same as those in the second and third stages.

[0076] Scenario 3: Target vehicle A has path conflicts with multiple vehicles. Target vehicle A gradually increases its speed by Δv (system set value), and then performs global planning again based on the increased speed until a suitable global driving path is found. However, the final determined driving speed cannot exceed the maximum value allowed by the road and the system.

[0077] Condition 4: Target vehicle A has path conflicts with multiple vehicles. Increasing the expected speed of target vehicle A to the maximum allowed by both the road and the system results in a conflict. Therefore, the speed of target vehicle A is reduced for global path planning.

[0078] Specifically, firstly, a second global driving path is planned based on the system's desired driving speed, and the corresponding vehicle travel time T is calculated. limitThe target vehicle A gradually reduces its speed by Δv1 (the system's set value), and then performs global planning again based on the reduced speed until a suitable global driving path is found. The system first determines the path based on T. limit The value is used to calculate the vehicle's minimum speed v. limit When the vehicle decelerates to v limit If a suitable global path has been found beforehand, the vehicle will travel at that speed; otherwise, the vehicle will travel according to the second global path.

[0079] In this embodiment of the invention, the above four working conditions and corresponding solutions are also applicable to the scenario where vehicles merge into the main road via ramp collaborative decision-making.

[0080] like Figure 9 The diagram shown is a flowchart of a global path generation process for an intelligent vehicle provided by an embodiment of the present invention; Figure 10 The diagram shown is a flowchart of a working condition provided by an embodiment of the present invention; Figure 11 The diagram shown is a flowchart of a working condition two according to an embodiment of the present invention; Figure 12 The diagram shown is a three-process flow chart provided by an embodiment of the present invention; Figure 13 The diagram shown is a flowchart of a four-condition operation provided by an embodiment of the present invention.

[0081] like Figure 14 The image shows a centralized vehicle cooperative decision-making ramp merging and separating device provided in an embodiment of the present invention, comprising:

[0082] The global path planning module 1401 is used to plan the global path of the target vehicle using a fast expanding random tree based on the target vehicle's current location information and destination address; the generated global path is combined with the expected driving speed to obtain the target vehicle's position on the global path at each time point, so that other vehicles can plan their global paths based on the target vehicle's position on the global path at each time point.

[0083] The collaborative control module 1402 is used to generate a collaborative control strategy based on the perception data of roadside equipment in the controlled road segment, the vehicle's operation data, and the perception data of vehicle-end sensors when the global path planning time of the target vehicle is later than the global path planning time of other vehicles when it arrives at the target road segment, and the planned ramp merging or separating routes have been occupied by other vehicles.

[0084] The specific implementation methods of each module can be referred to the description of the above method embodiments, and the embodiments of the present invention will not be repeated.

[0085] It should be noted that, depending on the implementation needs, the various steps / components described in this application can be broken down into more steps / components, or two or more steps / components or parts of the operation of steps / components can be combined into new steps / components to achieve the purpose of this invention.

[0086] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A centralized vehicle collaborative decision-making method for ramp merging and separation, characterized in that, include: Based on the target vehicle's current location and destination address, a global path for the target vehicle is planned using a fast expanding random tree. The generated global path is combined with the expected driving speed to determine the position of the target vehicle on the global path at each time point, so that other vehicles can plan their global paths based on the position of the target vehicle on the global path at each time point. If the global path planning time of the target vehicle is later than the global path planning time of other vehicles when it arrives at the target road segment, and the planned ramp merging or separating routes have been occupied by other vehicles, then a collaborative control strategy is generated based on the perception data of the roadside equipment in the controlled road segment, the vehicle's operation data, and the perception data of the vehicle-side sensors. The collaborative control strategy is generated based on the sensing data from roadside equipment within the controlled road section, vehicle operation data, and vehicle-mounted sensor sensing data, including: If the target vehicle and multiple vehicles have path conflicts, and increasing the target vehicle's desired speed to its maximum permissible speed still results in a conflict, then the target vehicle's speed is reduced. A second global travel path is then planned based on the desired speed, and the corresponding travel time is calculated. ; The target vehicle gradually reduces its speed, according to The value is used to calculate the minimum speed of the target vehicle. If the target vehicle slows down to If a suitable global path has already been found, then proceed accordingly. If the target vehicle does not proceed, it will proceed according to the second global driving path.

2. The method according to claim 1, characterized in that, The step of planning the global path of the target vehicle using a fast expanding random tree based on the target vehicle's current location information and destination address includes: Define the global path planning task space , Representing the spatial dimension, the global path planning task space is divided into spaces containing obstacles. and blank areas Let the initial state point corresponding to the current position information of the target vehicle be . The destination address corresponds to the target point. ; Point As the root node of the entire tree structure, a random point is generated at a random position in the blank area. ,by Centered on the tree structure, traverse and search all nodes to calculate the relationship between these nodes and the tree structure. The node with the smallest Euclidean distance between them is selected. As the nearest node; Let from Click The direction of the point is the growth direction of the tree structure, causing the search tree to grow from... A point grows a certain distance along the growth direction to obtain the next node in the tree structure, denoted as . Determine the location of the target vehicle. arrive On the represented path, whether it collides with obstacles around the path; if it collides with an obstacle, the node is removed. Then generate a random point again at a random location in the blank area. Perform a sampling search; if no collision occurs, then... Points are added to the tree until a new node is generated. Reach the target point or with the target point The distance between them is less than one unit growth step, from the target point By tracing back a series of parent nodes in sequence, the final path is obtained.

3. The method according to claim 2, characterized in that, The determination of the target vehicle at the point arrive Whether the path collided with obstacles surrounding the path, including: The target vehicle can be considered as a circle with a diameter of length L. The path traversed by the center of this circle is the point. arrive Given a path, determine if the path traversed by the center of the circle is the same as the path traversed by the center of the circle. arrive The target vehicle collides with obstacles around its path, where the value of L is the distance between the two points furthest from the target vehicle.

4. The method according to any one of claims 1 to 3, characterized in that, The collaborative control strategy is generated based on the sensing data from roadside equipment within the controlled road section, vehicle operation data, and vehicle-mounted sensor sensing data, including: If the target vehicle has a path conflict with the first vehicle, but not with other vehicles, and the target vehicle's deceleration will not affect the driving of other vehicles, while the first vehicle's deceleration will affect the driving of other vehicles, then the target vehicle's driving process will change to a process of first decelerating and then accelerating. When the first vehicle releases the right-of-way, the target vehicle will change lanes to the target lane and then leave or enter the main road.

5. The method according to any one of claims 1 to 3, characterized in that, The collaborative control strategy is generated based on the sensing data from roadside equipment within the controlled road section, vehicle operation data, and vehicle-mounted sensor sensing data, including: If the target vehicle and the first vehicle have a path conflict, but there is no path conflict with other vehicles, and the target vehicle's deceleration will affect the driving of other vehicles, while the first vehicle's deceleration will not affect the driving of other vehicles, then the first vehicle's driving process will change to first decelerate and then accelerate. The first vehicle actively releases the right-of-way, and the target vehicle accelerates or changes lanes at a constant speed to the target lane and then leaves or enters the main road.

6. The method according to any one of claims 1 to 3, characterized in that, The collaborative control strategy is generated based on the sensing data from roadside equipment within the controlled road section, vehicle operation data, and vehicle-mounted sensor sensing data, including: If the target vehicle has path conflicts with multiple vehicles, the target vehicle's speed is gradually increased. Then, global path planning is performed again based on the increased speed until a suitable global driving path is found. The final determined driving speed cannot exceed the maximum allowed driving speed.

7. A centralized vehicle collaborative decision-making ramp merging and separating device, characterized in that, include: The global path planning module is used to plan the global path of the target vehicle using a fast expanding random tree based on the target vehicle's current location information and destination address. The generated global path is combined with the expected driving speed to determine the position of the target vehicle on the global path at each time point, so that other vehicles can plan their global paths based on the position of the target vehicle on the global path at each time point. The collaborative control module is used to generate a collaborative control strategy based on the perception data of roadside equipment in the controlled road segment, the vehicle's operation data, and the perception data of vehicle-end sensors when the global path planning time of the target vehicle is later than the global path planning time of other vehicles when it arrives at the target road segment, and the planned ramp merging or separating routes have been occupied by other vehicles. The collaborative control strategy is generated based on the sensing data from roadside equipment within the controlled road section, vehicle operation data, and vehicle-mounted sensor sensing data, including: If the target vehicle and multiple vehicles have path conflicts, and increasing the target vehicle's desired speed to its maximum permissible speed still results in a conflict, then the target vehicle's speed is reduced. A second global travel path is then planned based on the desired speed, and the corresponding travel time is calculated. ; The target vehicle gradually reduces its speed, according to The value is used to calculate the minimum speed of the target vehicle. If the target vehicle slows down to If a suitable global path has already been found, then proceed accordingly. If the target vehicle does not proceed, it will proceed according to the second global driving path.

8. A cloud-based system including the centralized vehicle collaborative decision-making ramp merging and separating device as described in claim 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images