WO2022078067A1 - 面向智能车辆的区域协同驾驶意图调度方法、系统和介质 - Google Patents

面向智能车辆的区域协同驾驶意图调度方法、系统和介质 Download PDF

Info

Publication number
WO2022078067A1
WO2022078067A1 PCT/CN2021/113984 CN2021113984W WO2022078067A1 WO 2022078067 A1 WO2022078067 A1 WO 2022078067A1 CN 2021113984 W CN2021113984 W CN 2021113984W WO 2022078067 A1 WO2022078067 A1 WO 2022078067A1
Authority
WO
WIPO (PCT)
Prior art keywords
vehicle
driving
driving intention
vehicles
dispatching
Prior art date
Application number
PCT/CN2021/113984
Other languages
English (en)
French (fr)
Inventor
綦科
李文康
Original Assignee
广州大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 广州大学 filed Critical 广州大学
Priority to US18/549,989 priority Critical patent/US20240217559A1/en
Publication of WO2022078067A1 publication Critical patent/WO2022078067A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
    • G05D1/0223—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving speed control of the vehicle
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/0464—Convolutional networks [CNN, ConvNet]
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W60/00—Drive control systems specially adapted for autonomous road vehicles
    • B60W60/001—Planning or execution of driving tasks
    • B60W60/0015—Planning or execution of driving tasks specially adapted for safety
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W60/00—Drive control systems specially adapted for autonomous road vehicles
    • B60W60/001—Planning or execution of driving tasks
    • B60W60/0027—Planning or execution of driving tasks using trajectory prediction for other traffic participants
    • B60W60/00276—Planning or execution of driving tasks using trajectory prediction for other traffic participants for two or more other traffic participants
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
    • G05D1/0214—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory in accordance with safety or protection criteria, e.g. avoiding hazardous areas
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0276—Control of position or course in two dimensions specially adapted to land vehicles using signals provided by a source external to the vehicle
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0276—Control of position or course in two dimensions specially adapted to land vehicles using signals provided by a source external to the vehicle
    • G05D1/0278—Control of position or course in two dimensions specially adapted to land vehicles using signals provided by a source external to the vehicle using satellite positioning signals, e.g. GPS
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods

Definitions

  • the invention relates to the field of intelligent vehicle control, in particular to a method, system and medium for dispatching regional cooperative driving intentions for intelligent vehicles.
  • intelligent driving in the mode of human-vehicle driving is the mainstream mode of intelligent transportation.
  • the vehicle control can be transferred between the person and the vehicle, and the driver can take over the vehicle control and issue vehicle control commands at any time.
  • an intelligent vehicle drives autonomously
  • the intelligent vehicle collects and recognizes environmental information in real time through sensing equipment, and then makes driving decisions based on the environmental information.
  • drivers will change their driving intentions at any time, take over vehicle control, and control the vehicle to perform actions such as acceleration, deceleration, uniform speed, and lane change.
  • Such sudden changes in driving intention will seriously affect driving safety.
  • multiple drivers changing their driving intentions at the same time will easily lead to conflicts between the driving intentions of drivers between vehicles in adjacent areas, reducing the overall driving efficiency in the area.
  • the vehicle under test and the vehicles in the adjacent area are autonomously driven, a safe distance and a safe speed are maintained between each vehicle, and the horizontal and vertical motion states (position, speed) of each vehicle are matched.
  • the driver of the vehicle under test If a sudden lane change strategy is adopted, the driver behind the target lane adopts a sudden acceleration driving strategy, and the driver in front of the target lane adopts a sudden deceleration driving strategy, the test vehicle may collide with the front and rear vehicles in the target lane.
  • intelligent driving mostly monitors the driving intention of vehicles in the vicinity of the vehicle through vehicle-to-vehicle communication and V2X technology.
  • the vehicle-to-vehicle communication and V2X technology cannot realize the global perception and coordinated scheduling of the driving intentions of all vehicles within a certain area. Therefore, for intelligent driving in the human-vehicle co-driving mode scenario, it is also necessary to coordinate the scheduling of various vehicles within a certain area.
  • the driving intention of the vehicle and then guide the driving decision of the vehicle according to the driving intention of each vehicle after the coordinated scheduling, so that the vehicle can drive in a safe and orderly manner, and improve the intelligent driving safety and traffic efficiency in the mode of human-vehicle co-driving.
  • the first purpose of the present invention is to overcome the shortcomings and deficiencies of the prior art, and to provide a method for dispatching regional cooperative driving intentions for intelligent vehicles, which can coordinately dispatch the driving intentions of all vehicles within an area, and generate a global dispatching result to guide the vehicles.
  • Driving decision-making on the one hand, avoids the risk of vehicle conflict and improves driving safety, on the other hand, it also improves the overall traffic efficiency within the region.
  • the second object of the present invention is to provide an intelligent vehicle-oriented regional cooperative driving intention scheduling device.
  • the third object of the present invention is to provide a regional cooperative driving intention scheduling system for intelligent vehicles.
  • a fourth object of the present invention is to provide a storage medium.
  • a fifth object of the present invention is to provide a computing device.
  • an intelligent vehicle-oriented regional cooperative driving intention scheduling method comprising:
  • the driving intention information of each vehicle within the dispatch area the status information of the vehicle and the position information of the vehicle, a global map of the driving intention of all vehicles within the dispatch area is generated;
  • the process of identifying the driving intention of the vehicle through the state information and location information of the vehicle includes:
  • the position information P i of each vehicle i within the dispatch area is represented by point coordinates, 1 ⁇ i ⁇ N, and mapped to the corresponding grid of the occupancy grid, and the mapping has The grid of vehicle location information is marked as occupied, and the grid that is not mapped with vehicle location information is marked as idle;
  • the predicted state of the grid is occupied by a vehicle, that is, when the grid is predicted to be occupied by a vehicle, it is predicted that there is no conflict of vehicle driving intentions in the grid. the vehicle is driven in accordance with its driving intent;
  • step Se When the predicted state of the grid is occupied by multiple vehicles, that is, when the grid is predicted to be occupied by multiple vehicles, it is predicted that there is a conflict of vehicle driving intentions in the grid; at this time, step Se is entered;
  • an intelligent vehicle-oriented regional cooperative driving intention scheduling device comprising:
  • the information acquisition module is used to acquire the state information and location information of the vehicle, and is used to acquire the driving intention of the vehicle identified by the state information and location information of the vehicle;
  • the driving intention global map generation module is used to generate the driving intention global map of all vehicles within the scheduling area according to the driving intention information, vehicle status information and vehicle position information of the vehicles within the scheduling area;
  • the map model building module is used to build a grid map model of scheduling area occupation
  • the global scheduling result generation module is used to coordinate the global driving intention of vehicles within the scheduling area according to the constructed scheduling area occupied grid map model, and generate the global scheduling result of vehicle driving intention in the scheduling area; so that the global scheduling result of vehicle driving intention through the scheduling area, Guide the driving decisions of vehicles within the dispatch area.
  • the third object of the present invention is achieved by the following technical solutions: a regional cooperative driving intention dispatching system for intelligent vehicles, including a cloud dispatching system, and an on-board driving intention perception system and a vehicle-mounted driving intention control system arranged on the vehicle;
  • the in-vehicle driving intention perception system is connected to the in-vehicle driving intention control system, and is used to obtain the state information and position information of the vehicle, and sends the acquired state information and position information of the vehicle to the in-vehicle driving intention control system;
  • the vehicle driving intention control system is wirelessly connected to the cloud dispatching system, used for identifying the driving intention of the vehicle according to the state information of the vehicle, and sending the driving intention information of the vehicle and the state information and position information of the vehicle to the cloud dispatching system ;
  • the cloud scheduling system is used for executing the intelligent vehicle-oriented regional cooperative driving intention scheduling method according to the first object of the present invention.
  • the vehicle-mounted driving intention perception system includes a vehicle state acquisition unit and a positioning unit;
  • the vehicle state acquisition unit is used to acquire the state information of the vehicle, including the state of the accelerator pedal of the vehicle, the steering wheel angle, the state of the brake pedal and the absolute speed of the vehicle;
  • the positioning unit is used for acquiring the position information of the vehicle, including the GPS longitude information and GPS latitude information of the vehicle.
  • the vehicle-mounted driving intention control system includes a driving intention recognition unit, a communication unit and an output unit;
  • the driving intention recognition unit is used for recognizing the driving intention of the vehicle according to the state information of the vehicle;
  • the communication unit is used to send the driving intention information of the vehicle and the state information and position information of the vehicle to the cloud dispatching system;
  • the output unit is configured to receive the global dispatching result of the vehicle driving intention in the driving dispatching area sent by the cloud dispatching system, so that the vehicle driving decision-making system guides the driving decision of the vehicle according to the dispatching result;
  • the cloud scheduling system includes a cloud communication server and a cloud cooperative scheduling server;
  • the cloud cooperative scheduling server is used for executing the intelligent vehicle-oriented regional cooperative driving intention scheduling method according to the first object of the present invention
  • the cloud communication server is used to communicate with the communication unit in the vehicle-mounted driving intention control system, receive the vehicle's driving intention information and the vehicle's status information and position information sent by the vehicle-mounted driving intention control system, and dispatch the vehicle in the driving dispatch area to drive the vehicle.
  • the intention global scheduling result is sent to the vehicle driving intention control system.
  • the fourth object of the present invention is achieved by the following technical solutions: a storage medium, characterized in that a program is stored, and when the program is executed by a processor, the regional coordination for intelligent vehicles described in the first object of the present invention is realized Driving intent scheduling method.
  • the fifth object of the present invention is achieved by the following technical solutions: a computing device, comprising a processor and a memory for storing a program executable by the processor, when the processor executes the program stored in the memory, the first object of the present invention is achieved.
  • a computing device comprising a processor and a memory for storing a program executable by the processor, when the processor executes the program stored in the memory, the first object of the present invention is achieved.
  • the present invention has the following advantages and effects:
  • the regional cooperative driving intention scheduling method for intelligent vehicles of the present invention firstly obtains the status information, position information and driving intention information of each vehicle within the scheduling area, and generates a global driving intention of all vehicles within the scheduling area based on the above information. Then build a grid map model for the occupancy of the scheduling area; according to the constructed grid map model for the occupancy of the scheduling area, coordinate the global driving intention of the vehicles within the scheduling area to generate the global scheduling result of the vehicle driving intention in the scheduling area; make the driving intention of the vehicle in the scheduling area
  • the global scheduling results guide the driving decisions of vehicles within the scheduling area.
  • the present invention is based on the grid map model occupied by the dispatching area, and can coordinate the global driving intention of the vehicles within the dispatching area, and can comprehensively dispatch and control the driving intentions of each vehicle within the dispatching area, generate a global dispatching result, and guide the vehicle to drive. Decision-making, on the one hand, avoids the risk of vehicle conflict and improves driving safety, on the other hand, it will also improve the overall traffic efficiency within the region.
  • a driving intention recognition model is constructed based on the convolutional neural network model, and then the driving intention recognition model is used to recognize the driving intention of the vehicle based on the vehicle state information, which can accurately Recognize vehicle driving intent.
  • a map model of the scheduling area occupancy grid (occupancy grid) is constructed, and the prediction of each vehicle is determined according to the driving intention of each vehicle in the scheduling area.
  • Position information based on the predicted position information of each vehicle, the predicted state of each grid can be determined, so that the vehicles that will occupy the grid can be predicted.
  • the driving intention of the vehicle can be controlled by controlling the vehicle. The driving of the vehicle is scheduled.
  • the method of the present invention is based on the occupancy grid map model of the dispatching area, which can realize the unified dispatching of the driving intentions of vehicles within the dispatching area, control the driving intentions of all vehicles in the dispatching area, and can effectively avoid the occurrence of multiple vehicles arriving at the same grid at the same time.
  • the phenomenon of vehicle collision is based on the occupancy grid map model of the dispatching area, which can realize the unified dispatching of the driving intentions of vehicles within the dispatching area, control the driving intentions of all vehicles in the dispatching area, and can effectively avoid the occurrence of multiple vehicles arriving at the same grid at the same time.
  • the regional cooperative driving intention dispatching system for intelligent vehicles of the present invention includes a cloud dispatching system, an on-board driving intention perception system and an on-board driving intention control system set on the vehicle;
  • the driving intention perception system collects the state information and position information of the vehicle, and then the on-board driving intention control system on each vehicle identifies the driving intention of the vehicle based on the state of the vehicle, and finally sends the driving intention information of the vehicle as well as the state information and position information of the vehicle.
  • the cloud dispatching system is implemented by the cloud dispatching system to generate the global dispatching results of vehicle driving intentions in the dispatching area, and send the dispatching results to the vehicle driving intention control system, so that the vehicles can travel based on the dispatching results sent by the cloud dispatching system.
  • the present invention can realize the global control of vehicles in the area based on the cloud dispatching system, and can effectively avoid collision between vehicles in the intelligent driving mode in the human-vehicle co-driving mode scenario, and improve the human-vehicle co-driving mode. intelligent driving safety and traffic efficiency.
  • FIG. 1 is a flow chart of a method for scheduling regional cooperative driving intentions for intelligent vehicles according to the present invention.
  • Fig. 2 is a straight line area occupancy grid map model established in the method of the present invention.
  • Fig. 3 is a curved area occupancy grid map model established in the method of the present invention.
  • FIG. 4 is a structural block diagram of an intelligent vehicle-oriented regional cooperative driving intention scheduling device according to the present invention.
  • FIG. 5 is a structural block diagram of the regional cooperative driving intention scheduling system for intelligent vehicles according to the present invention.
  • This embodiment discloses a method for dispatching regional cooperative driving intentions for intelligent vehicles.
  • the method is used in intelligent driving in a human-vehicle co-driving mode, which can avoid the risk of vehicle conflict, improve driving safety, and on the other hand, also improve regional
  • the overall traffic efficiency within the range, the specific process of the method is shown in Figure 1, including:
  • the acquired vehicle state information specifically includes the accelerator pedal state of the vehicle, the steering wheel angle, the brake pedal state, and the absolute speed of the vehicle.
  • the process of identifying the driving intention information of the vehicle through the state information of the vehicle may include:
  • the certainty threshold of the left lane-changing and right-lane changing driving intention categories is set to 80%
  • the certainty threshold of the remaining unchanged driving intention category is set to 70%
  • the certainty threshold of the acceleration and deceleration driving intention categories is set to be 80%. 80%.
  • the scheduling area occupancy grid (occupancy grid) map model is to divide the area obtained by dividing the lane lines into equal parts, and each obtained grid is called a driving unit (cell).
  • the vertical length of the cell is set to 5m, and the horizontal width is set to the width of a single lane by default.
  • each cell is approximated as a rectangle; as shown in Figure 3, for a curved lane, each cell can be approximated as a convex quadrilateral, and the coordinates of a cell are determined by the coordinates of every four vertices, Use the four line segments connecting the four vertices as the extent of a cell. If the vehicle position is within the range of a certain cell, the state of the cell is marked as occupied, as shown in the filled part in FIG. 2 and FIG. 3 , and the status of other unoccupied cells is marked as idle.
  • step S4 the process of generating the global scheduling result of vehicle driving intention in the scheduling area in step S4 is as follows:
  • the position information P i of each vehicle i in the dispatching area is represented by point coordinates, 1 ⁇ i ⁇ N, and mapped to the corresponding cell of the occupancygrid map model, and the vehicles are mapped
  • the cell with location information is marked as occupied, and the cell with no vehicle location information mapped is marked as idle.
  • the current position of each vehicle in the occupied grid model is determined. Therefore, based on the above-mentioned longitudinal displacement S i ′ and lateral displacement S i ′′ information, the predicted position of the vehicle can be obtained, such as shown in Figure 2.
  • the predicted state of the cell is identified as being occupied by each vehicle among multiple vehicles; for example, for a certain cell, when it is predicted that the vehicle a1 will arrive at the cell according to the predicted position information of each vehicle, the predicted state of the cell is It is marked as occupied by vehicle a1; and when it is predicted that vehicles a1, a2 and a3 will arrive at the cell according to the predicted position information of each vehicle, the predicted state of the cell is marked as occupied by vehicles a1, a2 and a3. ;
  • For each cell determine whether its predicted state is occupied by multiple vehicles; among them:
  • the vehicle that is predicted to occupy the cell is controlled to drive according to its driving intention. intent to drive;
  • step Se When the predicted state of the cell is occupied by multiple vehicles, that is, when it is predicted that the cell will be occupied by multiple vehicles, it is predicted that the cell has a conflict of vehicle driving intentions; at this time, step Se is entered;
  • This embodiment is based on the operations of steps Sa to Sf, so that among the vehicles predicted to arrive at the same grid at the next moment, only one vehicle can finally reach the grid, which can effectively prevent two or more vehicles from arriving at the same grid at the same moment. A cell, resulting in a collision traffic accident.
  • This embodiment discloses an intelligent vehicle-oriented regional cooperative driving intention scheduling device. As shown in FIG. 4 , it includes an information acquisition module, a driving intention global map generation module, a map model construction module and a global scheduling result generation module: each module realizes The functions are as follows:
  • the information acquisition module is used to acquire the state information and location information of the vehicle, and is used to acquire the driving intention of the vehicle identified by the state information and location information of the vehicle;
  • the driving intention global map generation module is used to generate the driving intention global map of all vehicles within the scheduling area according to the driving intention information, vehicle status information and vehicle position information of the vehicles within the scheduling area;
  • the map model building module is used to build a grid map model of scheduling area occupation
  • the global scheduling result generation module is used to coordinate the global driving intention of vehicles within the scheduling area according to the constructed scheduling area occupied grid map model, and generate the global scheduling result of vehicle driving intention in the scheduling area; so that the global scheduling result of vehicle driving intention through the scheduling area, Guide the driving decisions of vehicles within the dispatch area.
  • the smart vehicle-oriented regional cooperative driving intention dispatching system in this embodiment includes a cloud dispatching system 30 , an on-board driving intention sensing system 10 and an on-board driving intention control system 20 provided on the vehicle. in:
  • the in-vehicle driving intention perception system is connected to the in-vehicle driving intention control system, and is used to obtain the status information and position information of the vehicle, and send the obtained vehicle status information and position information to the in-vehicle driving intention control system;
  • the vehicle-mounted driving intention perception system 10 includes a vehicle state acquisition unit 11 and a positioning unit 12 . in:
  • the vehicle state obtaining unit 11 is used for obtaining state information of the vehicle, including the state of the accelerator pedal of the vehicle, the steering wheel angle, the state of the brake pedal and the absolute speed of the vehicle.
  • the positioning unit 12 is configured to acquire the position information of the vehicle, including the GPS longitude information and the GPS latitude information of the vehicle.
  • the vehicle driving intention control system is wirelessly connected to the cloud dispatching system, used to identify the vehicle's driving intention according to the vehicle's state information, and send the vehicle's driving intention information as well as the vehicle's state information and location information to the cloud dispatching system.
  • the vehicle-mounted driving intention control system 20 includes a driving intention recognition unit 21 , a communication unit 22 and an output unit 23 . in:
  • the driving intention recognition unit 21 is used for recognizing the driving intention of the vehicle according to the state information of the vehicle.
  • the process of identifying the driving intention of the vehicle by the driving intention recognition unit according to the state information of the vehicle may be as follows:
  • the certainty threshold of the left lane change and right lane change driving intention categories may be set to 80%
  • the certainty threshold of the remaining unchanged driving intention category may be set to 70%
  • the certainty threshold of the acceleration and deceleration driving intention categories may be set. is 80%.
  • the communication unit 22 is used to send the driving intention information of the vehicle and the state information and position information of the vehicle to the cloud dispatching system; in this embodiment, the communication unit may be a wireless communication module arranged on the vehicle, including a 4G communication module, 5G communication module, etc., the communication unit can upload the corresponding information of the vehicle driving intention control system to the cloud dispatching system through the mobile communication base station.
  • the output unit 23 is used to receive the global dispatching result of the vehicle driving intention in the driving dispatching area sent by the cloud dispatching system, so that the vehicle driving decision-making system guides the driving decision of the vehicle according to the dispatching result; in this embodiment, the output unit is connected to the communication unit, through The communication unit obtains the global scheduling result of the vehicle driving intention in the scheduling area from the cloud scheduling system.
  • the vehicle driving decision system is a system installed on the vehicle to control the driving of the vehicle. The vehicle decision system can control the vehicle to perform corresponding movements according to the driving intention of the vehicle.
  • the cloud scheduling system 30 is configured to execute the intelligent vehicle-oriented regional cooperative driving intention scheduling method described in Embodiment 1.
  • the cloud scheduling system 30 includes a cloud communication server 31 and a cloud cooperative scheduling server 32; wherein,
  • the cloud collaborative scheduling server is configured to execute the intelligent vehicle-oriented regional collaborative driving intention scheduling method described in Embodiment 1, as follows:
  • the driving intention information of each vehicle within the dispatch area the status information of the vehicle and the position information of the vehicle, a global map of the driving intention of all vehicles within the dispatch area is generated;
  • the global driving intention of vehicles in the scheduling area is coordinated to generate the global scheduling result of vehicle driving intention in the scheduling area.
  • the cloud communication server is used to communicate with the communication unit in the in-vehicle driving intention control system, receive the driving intention information of the vehicle and the status information and position information of the vehicle sent by the in-vehicle driving intention control system, and globalize the driving intention of the vehicle in the driving dispatch area.
  • the scheduling result is sent to the vehicle driving intention control system.
  • This embodiment discloses a storage medium that stores a program.
  • the program is executed by a processor, the method for dispatching regional cooperative driving intentions for intelligent vehicles described in Embodiment 1 is implemented, as follows:
  • the driving intention information of each vehicle within the dispatch area the status information of the vehicle and the position information of the vehicle, a global map of the driving intention of all vehicles within the dispatch area is generated;
  • the global driving intention of vehicles in the scheduling area is coordinated to generate the global scheduling result of vehicle driving intention in the scheduling area.
  • the storage medium may be a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a U disk, a removable hard disk, and other media.
  • This embodiment discloses a computing device, which includes a processor and a memory for storing a program executable by the processor. It is characterized in that, when the processor executes the program stored in the memory, the intelligent vehicle-oriented intelligent vehicle described in Embodiment 1 is implemented.
  • the regional collaborative driving intent scheduling method is as follows:
  • a global map of the driving intention of all vehicles within the dispatch area is generated;
  • the global driving intention of vehicles in the scheduling area is coordinated to generate the global scheduling result of vehicle driving intention in the scheduling area.
  • the computing device may be a terminal device such as a desktop computer, a notebook computer, a smart phone, a PDA handheld terminal, and a tablet computer.
  • the present invention is aimed at scenarios where multiple drivers change their driving intentions at the same time in the co-driving mode, which may easily lead to conflicts between vehicles in adjacent areas, and coordinately dispatch the driving intentions of all vehicles in the area to generate the driving intentions.
  • the global scheduling sequence guides vehicle driving decisions. On the one hand, it avoids the risk of vehicle conflict and improves driving safety. On the other hand, it also improves the overall traffic efficiency in the area.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Biophysics (AREA)
  • Software Systems (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Mathematical Physics (AREA)
  • Biomedical Technology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Human Computer Interaction (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Traffic Control Systems (AREA)

Abstract

一种面向智能车辆的区域协同驾驶意图调度方法、系统和介质，首先获取到调度区域范围内各车辆的状态信息、位置信息以及驾驶意图信息，基于调度区域范围内各车辆的状态信息、位置信息以及驾驶意图信息生成调度区域范围内全部车辆的驾驶意图全局图；然后构建调度区域占据栅格地图模型；根据构建的调度区域占据栅格地图模型，协同调度区域范围内车辆全局驾驶意图，生成调度区域车辆驾驶意图全局调度结果；使得通过调度区域车辆驾驶意图全局调度结果，指导调度区域范围内车辆的行车决策。由此能够对调度区域范围内的各车辆驾驶意图进行全面调度和控制，生成全局调度结果，指导车辆行车决策，一方面避免车辆冲突风险，提高行车安全性，另一方面也会提升区域范围内的整体通行效率。

Description

面向智能车辆的区域协同驾驶意图调度方法、系统和介质 技术领域
本发明涉及智能车辆控制领域，特别涉及一种面向智能车辆的区域协同驾驶意图调度方法、系统和介质。
背景技术
智能驾驶的发展有助于提升道路交通智能化水平、推动交通运输行业的转型升级。在无人驾驶完全安全可用之前，人车共驾模式的智能驾驶是智能交通的主流模式。在该种模式下，车辆控制权可以在人和车之间转移，驾驶人可以随时接管车辆控制并发出车辆控制指令。
在智能车辆自主驾驶时，智能车辆通过感知设备实时采集和识别环境信息，然后根据环境信息进行行驶决策。但由于驾驶场景的多样性，以及驾驶人驾驶习惯的多样性，驾驶人会随时改变驾驶意图，接管车辆控制，控制车辆实施加速、减速、匀速、换道等动作。这类驾驶意图突变的行为，将会严重影响行车安全。尤其在车流量较大的行车环境下，多个驾驶人同时改变驾驶意图，将易导致相邻区域范围内车辆间驾驶人的行车意图发生冲突，降低区域范围的整体行车效率。例如：被试车与相邻区域范围内的车辆都执行自主驾驶，各车之间保持安全距离和安全速度，各车横向和纵向的运动状态(位置，速度)匹配，此时，被试车驾驶人采取突然换道策略，目标车道后车驾驶人采取突然加速行驶策略，目标车道前车驾驶人采取突然减速行驶策略，则被试车与目标车道前后车都可能发生碰撞事故。
当前，智能驾驶大多通过车车通信和V2X技术，监控本车邻近范围内车辆的行车意图。但是，车车通信和V2X技术却无法实现对一定区域范围内全部车辆驾驶意图的全局感知和协同调度，因此，对于人车共驾模式场景下的智能驾驶，也有必要协同调度一定区域范围内各车辆的驾驶意图，再根据协同调度后的各车辆驾驶意图顺序指导车辆行驶决策，使车辆安全的按序行驶，提升人车共驾模式下的智能驾驶安全性和通行效率。
发明内容
本发明的第一目的在于克服现有技术的缺点与不足，提供一种面向智能车辆的区域协同驾驶意图调度方法，该方法能够协同调度区域范围内全部车辆的驾驶意图，生成全局调度结果指导车辆行车决策，一方面避免车辆冲突风险，提高行车安全性，另一方面也会提升区域 范围内的整体通行效率。
本发明的第二目的在于提供一种面向智能车辆的区域协同驾驶意图调度装置。
本发明的第三目的在于提供一种面向智能车辆的区域协同驾驶意图调度系统。
本发明的第四目的在于提供一种存储介质。
本发明的第五目的在于提供一种计算设备。
本发明的第一目的通过下述技术方案实现：一种面向智能车辆的区域协同驾驶意图调度方法，包括：
获取车辆的状态信息和位置信息；
获取通过车辆的状态信息所识别出的车辆的驾驶意图信息；
根据调度区域范围内各车辆的驾驶意图信息、车辆的状态信息和车辆的位置信息，生成调度区域范围内全部车辆的驾驶意图全局图；
构建调度区域占据栅格地图模型；
根据构建的调度区域占据栅格地图模型，协同调度区域范围内车辆全局驾驶意图，生成调度区域车辆驾驶意图全局调度结果；使得通过调度区域车辆驾驶意图全局调度结果，指导调度区域范围内车辆的行车决策。
优选的，通过车辆的状态信息和位置信息识别出车辆驾驶意图的过程包括：
基于卷积神经网络构建驾驶意图识别模型，其中将车辆状态信息作为驾驶意图识别模型的输入量I，由驾驶意图识别模型Softmax层输出对驾驶意图的识别向量w＝(w1,w2,w3,w4,w5)，其中：w1、w2、w3、w4、w5分别为左换道、保持不变、右换道、加速、减速的驾驶意图类别的概率；
设定各种驾驶意图类别的确信阈值；当某一驾驶意图类别的输出概率大于对应的确信阈值时，则判断车辆存在该类别的驾驶意图C；其中：
C∈{Ca:左换道，Cb:保持不变，Cc:右换道，Cd:加速，Cf:减速}。
优选的，生成调度区域范围内全部车辆的驾驶意图全局图为：G＝[g i]，1≤i≤N，N为调度区域范围内车辆总数，其中：g i＝(C i,V i,W i,P i)，C i,V i,W i,P i分别为调度区域范围内第i辆车的驾驶意图、绝对速度、方向盘转角和车辆位置信息。
更进一步的，生成调度区域车辆驾驶意图全局调度结果的过程如下：
Sa、基于调度区域占据栅格模型，将调度区域范围内每一车辆i的位置信息P i用点坐标表示，1≤i≤N，并映射到占据栅格的对应网格中，将映射有车辆位置信息的网格标识为占用，将未映射有车辆位置信息的网格标识为空闲；
Sb、根据调度区域内每一车辆i的驾驶意图C i、绝对速度V i和方向盘转角W i，计算每一 车辆i在安全加速度a s和安全时间t s内的纵向位移
Figure PCTCN2021113984-appb-000001
和横向位移
Figure PCTCN2021113984-appb-000002
并依据纵向位移S i'和横向位移S i”确定车辆i的预测位置信息
Figure PCTCN2021113984-appb-000003
1≤i≤N；
Sc、针对调度区域中的每一个网格，根据调度区域内每一车辆i的预测位置信息
Figure PCTCN2021113984-appb-000004
获取预测状态；
Sd、根据每个网格的预测状态，预测该网格是否存在车辆驾驶意图冲突的现象，具体为：
针对每个网格，判定其预测状态是否为被多辆车占用；其中：
当网格的预测状态为被一辆车占用时，即预测该网格将被一辆车占用时，则预测该网格不存在车辆驾驶意图冲突的现象，此时控制预测占用该网格的车辆按照其驾驶意图进行行驶；
当网格的预测状态为被多辆车占用时，即预测网格将被多辆车占用时，则预测该网格存在车辆驾驶意图冲突的现象；此时进入步骤Se；
Se、判定预测占用网格的多辆车中，是否有驾驶意图为：保持不变；
若有，则设置预测占用该网格的所有车辆驾驶意图均为：保持不变；
若否，则从预测占用该网格的多辆车中，随机选取一辆车辆按照其驾驶意图进行行驶，其他车辆驾驶意图全部设置为：保持不变；
Sf、基于上述操作确定出调度区域内各车辆的驾驶意图，生成调度结果。
本发明的第二目的通过下述技术方案实现：一种面向智能车辆的区域协同驾驶意图调度装置，包括：
信息获取模块，用于获取车辆的状态信息和位置信息，用于获取通过车辆的状态信息和位置信息所识别出的车辆驾驶意图；
驾驶意图全局图生成模块，用于根据调度区域范围内车辆的驾驶意图信息、车辆状态信息和车辆位置信息，生成调度区域范围内全部车辆的驾驶意图全局图；
地图模型构建模块，用于构建调度区域占据栅格地图模型；
全局调度结果生成模块，用于根据构建的调度区域占据栅格地图模型，协同调度区域范围内车辆全局驾驶意图，生成调度区域车辆驾驶意图全局调度结果；使得通过调度区域车辆驾驶意图全局调度结果，指导调度区域范围内车辆的行车决策。
本发明的第三目的通过下述技术方案实现：一种面向智能车辆的区域协同驾驶意图调度系统，包括云端调度系统，以及设置在车辆上的车载驾驶意图感知系统和车载驾驶意图控制系统；
所述车载驾驶意图感知系统连接车载驾驶意图控制系统，用于获取车辆的状态信息和位 置信息，并将获取到的车辆的状态信息和位置信息发送给车载驾驶意图控制系统；
所述车辆驾驶意图控制系统通过无线的方式连接云端调度系统，用于根据车辆的状态信息识别出车辆的驾驶意图，并且将车辆的驾驶意图信息以及车辆的状态信息和位置信息发送给云端调度系统；
所述云端调度系统，用于执行本发明第一目的所述的面向智能车辆的区域协同驾驶意图调度方法。
优选的，所述车载驾驶意图感知系统包括车辆状态获取单元和定位单元；
所述车辆状态获取单元，用于获取车辆的状态信息，包括车辆的油门踏板状态、方向盘转角、制动踏板状态和车辆绝对速度；
所述定位单元，用于获取车辆的位置信息，包括车辆GPS经度信息和GPS维度信息。
优选的，所述车载驾驶意图控制系统包括驾驶意图识别单元、通信单元和输出单元；
所述驾驶意图识别单元，用于根据车辆的状态信息识别出车辆的驾驶意图；
所述通信单元，用于将车辆的驾驶意图信息以及车辆的状态信息和位置信息发送给云端调度系统；
所述输出单元，用于接收云端调度系统发送的行车调度区域车辆驾驶意图全局调度结果，使得车辆行车决策系统根据该调度结果指导车辆的行车决策；
所述云端调度系统包括云端通信服务器和云端协同调度服务器；
所述云端协同调度服务器，用于执行本发明第一目的所述的面向智能车辆的区域协同驾驶意图调度方法；
所述云端通信服务器，用于与车载驾驶意图控制系统中的通信单元进行通信，接收车载驾驶意图控制系统发送的车辆的驾驶意图信息以及车辆的状态信息和位置信息，并且将行车调度区域车辆驾驶意图全局调度结果发送给车载驾驶意图控制系统。
本发明的第四目的通过下述技术方案实现：一种存储介质，其特征在于，存储有程序，所述程序被处理器执行时，实现本发明第一目的所述的面向智能车辆的区域协同驾驶意图调度方法。
本发明的第五目的通过下述技术方案实现：一种计算设备，包括处理器以及用于存储处理器可执行程序的存储器，所述处理器执行存储器存储的程序时，实现本发明第一目的所述的面向智能车辆的区域协同驾驶意图调度方法。
本发明相对于现有技术具有如下的优点及效果：
(1)本发明面向智能车辆的区域协同驾驶意图调度方法，首先获取到调度区域范围内各车辆的状态信息、位置信息以及驾驶意图信息，基于上述信息生成调度区域范围内全部车辆 的驾驶意图全局图；然后构建调度区域占据栅格地图模型；根据构建的调度区域占据栅格地图模型，协同调度区域范围内车辆全局驾驶意图，生成调度区域车辆驾驶意图全局调度结果；使得通过调度区域车辆驾驶意图全局调度结果，指导调度区域范围内车辆的行车决策。由上述可知，本发明基于调度区域占据栅格地图模型，协同调度区域范围内车辆全局驾驶意图，能够对调度区域范围内的各车辆驾驶意图进行全面调度和控制，生成全局调度结果，指导车辆行车决策，一方面避免车辆冲突风险，提高行车安全性，另一方面也会提升区域范围内的整体通行效率。
(2)本发明面向智能车辆的区域协同驾驶意图调度方法中，基于卷积神经网络模型构建驾驶意图识别模型，然后由驾驶意图识别模型基于车辆状态信息对车辆的驾驶意图进行识别，能够准确的识别出车辆驾驶意图。
(3)本发明面向智能车辆的区域协同驾驶意图调度方法中，在生成调度结果时，构建调度区域occupancy grid(占据栅格)地图模型，并且根据调度区域各车辆的驾驶意图确定各车辆的预测位置信息，基于各车辆的预测位置信息能够确定各网格的预测状态，从而预测出将占用网格的车辆，在占用网格的车辆为两辆及以上时，通过控制车辆的驾驶意图能够对车辆的驾驶进行调度。因此本发明方法基于调度区域occupancy grid地图模型，能够实现对调度区域范围内车辆驾驶意图的统一调度，控制调度范围内所有车辆的驾驶意图，可以有效避免多辆车辆同时到达同一个网格而出现车辆碰撞的现象。
(4)本发明面向智能车辆的区域协同驾驶意图调度系统中，包括云端调度系统以及设置在车辆上的车载驾驶意图感知系统和车载驾驶意图控制系统；在本发明中，由各车辆上的车载驾驶意图感知系统采集车辆的状态信息和位置信息，然后由各车辆上的车载驾驶意图控制系统基于车辆的状态识别出车辆的驾驶意图，最终将车辆驾驶意图信息以及车辆的状态信息和位置信息发送云端调度系统，由云端调度系统实现生成调度区域车辆驾驶意图全局调度结果，并且将调度结果发送到车载驾驶意图控制系统中，使得车辆基于云端调度系统发送的调度结果进行行驶。由上述可见，本发明基于云端调度系统能够实现区域范围内车辆的全局控制，对于人车共驾模式场景下的智能驾驶模式下，能够有效避免车辆之间的碰撞，提升人车共驾模式下的智能驾驶安全性和通行效率。
附图说明
图1是本发明面向智能车辆的区域协同驾驶意图调度方法流程图。
图2是本发明方法中建立的直线区域occupancy grid地图模型。
图3是本发明方法中建立的弯曲区域occupancy grid地图模型。
图4是本发明面向智能车辆的区域协同驾驶意图调度装置结构框图。
图5是本发明面向智能车辆的区域协同驾驶意图调度系统结构框图。
具体实施方式
下面结合实施例及附图对本发明作进一步详细的描述，但本发明的实施方式不限于此。
实施例1
本实施例公开了一种面向智能车辆的区域协同驾驶意图调度方法，该方法在人车共驾模式的智能驾驶中使用，能够避免车辆冲突风险，提高行车安全性，另一方面也会提升区域范围内的整体通行效率，该方法的具体过程如图1所示，包括：
S1、获取车辆的状态信息和位置信息；同时获取通过车辆的状态信息所识别出的车辆的驾驶意图信息。在本实施例中，获取的车辆状态信息具体包括车辆的油门踏板状态、方向盘转角、制动踏板状态和车辆绝对速度。
其中，通过车辆的状态信息识别出车辆的驾驶意图信息的过程可以包括：
S11、基于CNN卷积神经网络构建驾驶意图识别模型，其中将车辆状态信息作为驾驶意图识别模型的输入量I，由驾驶意图识别模型Softmax层输出对驾驶意图的识别向量w＝(w1,w2,w3,w4,w5)，其中：w1、w2、w3、w4、w5分别为左换道、保持不变、右换道、加速、减速的驾驶意图类别的概率。
S12、设定各种驾驶意图类别的确信阈值；将获取到的车辆当前的状态信息输入到驾驶意图识别模型中，当驾驶意图识别模型的某一驾驶意图类别的输出概率大于对应的确信阈值时，则判断车辆存在该类别的驾驶意图C；其中：
C∈{Ca:左换道，Cb:保持不变，Cc:右换道，Cd:加速，Cf:减速}。
在本实施例中，设定左换道、右换道驾驶意图类别的确信阈值为80％，设置保持不变驾驶意图类别的确信阈值为70％，设置加速、减速驾驶意图类别的确信阈值为80％。
S2、根据调度区域范围内各车辆的驾驶意图信息、车辆的状态信息和车辆的位置信息，生成调度区域范围内全部车辆的驾驶意图全局图。在本实施例中，生成调度区域范围内全部车辆的驾驶意图全局图为：G＝[g i]，1≤i≤N，N为调度区域范围内车辆总数，其中：g i＝(C i,V i,W i,P i)，C i,V i,W i,P i分别为调度区域范围内第i辆车的驾驶意图、绝对速度、方向盘转角和车辆位置信息。
S3、构建调度区域occupancy grid(占据栅格)地图模型。在本实施例中，调度区域occupancy grid(占据栅格)地图模型，是对车道线划分得到的区域进行等分，得到的每一个网格称为一 个行车单元(cell)。在本实施例中，cell的纵向长度设为5m，横向宽度默认为单个车道宽度。如图2所示，对于直车道，每个cell近似为一个长方形；如图3所示，对于弯曲车道，每个cell可近似为一个凸四边形，用每四个顶点坐标确定一个cell的坐标，用连接四个顶点的四条线段作为一个cell的范围。如果车辆位置在某个cell范围内，则该cell的状态标识为占用，如图2和图3中被填充的部分所示，其他未被占用的cell状态标识为空闲。
S4、根据构建的调度区域occupancy grid地图模型，协同调度区域范围内车辆全局驾驶意图，生成调度区域车辆驾驶意图全局调度结果；使得通过调度区域车辆驾驶意图全局调度结果R＝[C i]，1≤i≤N，指导调度区域范围内车辆的行车决策。
在本实施例中，步骤S4中生成调度区域车辆驾驶意图全局调度结果的过程如下：
Sa、基于调度区域占据栅格模型，将调度区域范围内每一车辆i的位置信息P i用点坐标表示，1≤i≤N，并映射到occupancygrid地图模型的对应cell中，将映射有车辆位置信息的cell标识为占用，将未映射有车辆位置信息的cell标识为空闲。
Sb、根据调度区域内每一车辆i的驾驶意图C i、绝对速度V i和方向盘转角W i，计算每一车辆i在安全加速度a s和安全时间t s内的纵向位移
Figure PCTCN2021113984-appb-000005
和横向位移
Figure PCTCN2021113984-appb-000006
并依据纵向位移S i'和横向位移S i”确定车辆i的预测位置信息
Figure PCTCN2021113984-appb-000007
1≤i≤N。
在本实施例中，基于位置信息，各车辆在占据栅格模型中当前的位置是确定的，因此基于上述纵向位移S i'和横向位移S i”信息，就能够得到车辆的预测位置，如图2中所示。
Sc、针对调度区域中的每一个cell，根据调度区域内每一车辆i的预测位置信息
Figure PCTCN2021113984-appb-000008
获取预测状态；其中针对于各cell，当根据车辆i的预测位置信息
Figure PCTCN2021113984-appb-000009
预测到车辆i将到达该cell时，则将该cell的预测状态标识为被车辆i所占用，而当根据各车辆的预测位置信息，预测到将有多个车辆到达该cell时，则将该cell的预测状态标识为被多辆车中的各车辆所占用；例如针对于某一cell，当根据各车辆的预测位置信息，预测到车辆a1将到达该cell时，则将该cell的预测状态标识为被车辆a1所占用；而当根据各车辆的预测位置信息，预测到车辆a1、a2和a3将到达该cell时，则将该cell的预测状态标识为被车辆a1、a2和a3所占用；
Sd、根据每个cell的预测状态，预测该cell是否存在车辆驾驶意图冲突的现象，具体为：
针对每个cell，判定其预测状态是否为被多辆车占用；其中：
当cell的预测状态为被一辆车占用时，即预测该cell将被一辆车占用时，则预测该cell不存在车辆驾驶意图冲突的现象，此时控制预测占用该cell的车辆按照其驾驶意图进行行驶；
当cell的预测状态为被多辆车占用时，即预测cell将被多辆车占用时，则预测该cell存 在车辆驾驶意图冲突的现象；此时进入步骤Se；
Se、判定预测占用cell的多辆车中，是否有驾驶意图为：保持不变；
若有，则设置预测占用该cell的所有车辆驾驶意图均为：保持不变；
若否，则从预测占用该cell的多辆车中，随机选取一辆车辆按照其驾驶意图进行行驶，其他车辆驾驶意图全部设置为：保持不变；其中车辆驾驶意图为保持不变，指的是车辆按照原有的驾驶状态进行驾驶。
Sf、基于上述操作确定出调度区域内各车辆的驾驶意图C i，生成调度结果R＝[C i]，1≤i≤N。
本实施例基于步骤Sa至Sf的操作，使得下一时刻预测到达同一个网格的车辆中，只有一辆车最终能够到达该网格，能够有效避免同一时刻，两辆及以上车辆同时到达同一个cell，从而出现碰撞的交通事故。
本领域技术人员可以理解，实现本实施例方法中的全部或部分步骤可以通过程序来指令相关的硬件来完成，相应的程序可以存储于计算机可读存储介质中。应当注意，尽管在附图中以特定顺序描述了本实施例1的方法操作，但是这并非要求或者暗示必须按照该特定顺序来执行这些操作，或是必须执行全部所示的操作才能实现期望的结果。相反，描绘的步骤可以改变执行顺序，有些步骤也可以同时执行。附加地或备选地，可以省略某些步骤，将多个步骤合并为一个步骤执行，和/或将一个步骤分解为多个步骤执行。
实施例2
本实施例公开了一种面向智能车辆的区域协同驾驶意图调度装置，如图4所示，包括信息获取模块、驾驶意图全局图生成模块、地图模型构建模块和全局调度结果生成模块：各个模块实现的功能分别如下：
信息获取模块，用于获取车辆的状态信息和位置信息，用于获取通过车辆的状态信息和位置信息所识别出的车辆驾驶意图；
驾驶意图全局图生成模块，用于根据调度区域范围内车辆的驾驶意图信息、车辆状态信息和车辆位置信息，生成调度区域范围内全部车辆的驾驶意图全局图；
地图模型构建模块，用于构建调度区域占据栅格地图模型；
全局调度结果生成模块，用于根据构建的调度区域占据栅格地图模型，协同调度区域范围内车辆全局驾驶意图，生成调度区域车辆驾驶意图全局调度结果；使得通过调度区域车辆驾驶意图全局调度结果，指导调度区域范围内车辆的行车决策。
本实施例上述各个模块的具体实现可以参见上述实施例1，在此不再一一赘述。需要说明 的是，本实施例提供的装置仅以上述各功能模块的划分进行举例说明，在实际应用中，可以根据需要而将上述功能分配由不同的功能模块完成，即将内部结构划分成不同的功能模块，以完成以上描述的全部或者部分功能。
实施例3
本实施例公开了一种面向智能车辆的区域协同驾驶意图调度系统，基于本实施例系统，能够实现实施例1所述的面向智能车辆的区域协同驾驶意图调度方法。如图5中所示，本实施例面向智能车辆的区域协同驾驶意图调度系统包括云端调度系统30，以及设置在车辆上的车载驾驶意图感知系统10和车载驾驶意图控制系统20。其中：
车载驾驶意图感知系统连接车载驾驶意图控制系统，用于获取车辆的状态信息和位置信息，并将获取到的车辆的状态信息和位置信息发送给车载驾驶意图控制系统；
在本实施例中，车载驾驶意图感知系统10包括车辆状态获取单元11和定位单元12。其中：
车辆状态获取单元11，用于获取车辆的状态信息，包括车辆的油门踏板状态、方向盘转角、制动踏板状态和车辆绝对速度。
定位单元12，用于获取车辆的位置信息，包括车辆GPS经度信息和GPS维度信息。
车辆驾驶意图控制系统通过无线的方式连接云端调度系统，用于根据车辆的状态信息识别出车辆的驾驶意图，并且将车辆的驾驶意图信息以及车辆的状态信息和位置信息发送给云端调度系统。
在本实施例中，车载驾驶意图控制系统20包括驾驶意图识别单元21、通信单元22和输出单元23。其中：
驾驶意图识别单元21，用于根据车辆的状态信息识别出车辆的驾驶意图。本实施例中驾驶意图识别单元根据车辆的状态信息识别出车辆的驾驶意图的过程可以如下：
基于CNN卷积神经网络构建驾驶意图识别模型，其中将车辆状态信息作为驾驶意图识别模型的输入量I，由驾驶意图识别模型Softmax层输出对驾驶意图的识别向量w＝(w1,w2,w3,w4,w5)，其中：w1、w2、w3、w4、w5分别为左换道、保持不变、右换道、加速、减速的驾驶意图类别的概率。
设定各种驾驶意图类别的确信阈值；将获取到的车辆当前的状态信息输入到驾驶意图识别模型中，当驾驶意图识别模型的某一驾驶意图类别的输出概率大于对应的确信阈值时，则判断车辆存在该类别的驾驶意图C；其中：
C∈{Ca:左换道，Cb:保持不变，Cc:右换道，Cd:加速，Cf:减速}。
在本实施例中，可以设定左换道、右换道驾驶意图类别的确信阈值为80％，设置保持不 变驾驶意图类别的确信阈值为70％，设置加速、减速驾驶意图类别的确信阈值为80％。
通信单元22，用于将车辆的驾驶意图信息以及车辆的状态信息和位置信息发送给云端调度系统；在本实施例中，通信单元可以是设置在车辆上的无线通信模块，包括4G通信模块、5G通信模块等，通信单元通过移动通信基站可以将车载驾驶意图控制系统相应的信息上传到云端调度系统。在本实施例中车载驾驶意图控制系统通过通信单元向云端调度系统上传O＝[C,V,W,P]信息，其中，C为识别的驾驶意图，V为车辆状态信息S中的车辆绝对速度，W为车辆状态信息S中的方向盘转角，P为车辆位置信息。
输出单元23，用于接收云端调度系统发送的行车调度区域车辆驾驶意图全局调度结果，使得车辆行车决策系统根据该调度结果指导车辆的行车决策；在本实施例中，输出单元连接通信单元，通过通信单元从云端调度系统获取到调度区域车辆驾驶意图全局调度结果，车辆行车决策系统为安装在车辆上控制车辆行驶的系统，车辆决策系统能够根据车辆的驾驶意图控制车辆进行相应的运动。
云端调度系统30，用于执行实施例1所述的面向智能车辆的区域协同驾驶意图调度方法。
在本实施例中，云端调度系统30包括云端通信服务器31和云端协同调度服务器32；其中，
云端协同调度服务器，用于执行实施例1所述的面向智能车辆的区域协同驾驶意图调度方法，如下：
获取车辆的状态信息和位置信息；
获取通过车辆的状态信息所识别出的车辆的驾驶意图信息；
根据调度区域范围内各车辆的驾驶意图信息、车辆的状态信息和车辆的位置信息，生成调度区域范围内全部车辆的驾驶意图全局图；
构建调度区域占据栅格地图模型；
根据构建的调度区域占据栅格地图模型，协同调度区域范围内车辆全局驾驶意图，生成调度区域车辆驾驶意图全局调度结果。
上述内容具体的操作过程见实施例1中描述，此处不再赘述。
云端通信服务器，用于与车载驾驶意图控制系统中的通信单元进行通信，接收车载驾驶意图控制系统发送的车辆的驾驶意图信息以及车辆的状态信息和位置信息，并且将行车调度区域车辆驾驶意图全局调度结果发送给车载驾驶意图控制系统。
实施例4
本实施例公开了一种存储介质，存储有程序，所述程序被处理器执行时，实现实施例1所述的面向智能车辆的区域协同驾驶意图调度方法，如下：
获取车辆的状态信息和位置信息；
获取通过车辆的状态信息所识别出的车辆的驾驶意图信息；
根据调度区域范围内各车辆的驾驶意图信息、车辆的状态信息和车辆的位置信息，生成调度区域范围内全部车辆的驾驶意图全局图；
构建调度区域占据栅格地图模型；
根据构建的调度区域占据栅格地图模型，协同调度区域范围内车辆全局驾驶意图，生成调度区域车辆驾驶意图全局调度结果。
上述内容具体的操作过程见实施例1中描述，此处不再赘述。
在本实施例中，存储介质可以是磁盘、光盘、计算机存储器、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、U盘、移动硬盘等介质。
实施例5
本实施例公开一种计算设备，包括处理器以及用于存储处理器可执行程序的存储器，其特征在于，所述处理器执行存储器存储的程序时，实现实施例1所述的面向智能车辆的区域协同驾驶意图调度方法，如下：
获取车辆的状态信息和位置信息；
获取通过车辆的状态信息所识别出的车辆的驾驶意图信息；
根据调度区域范围内各车辆的驾驶意图信息、车辆的状态信息和车辆的位置信息，生成调度区域范围内全部车辆的驾驶意图全局图；
构建调度区域占据栅格地图模型；
根据构建的调度区域占据栅格地图模型，协同调度区域范围内车辆全局驾驶意图，生成调度区域车辆驾驶意图全局调度结果。
上述内容具体的操作过程见实施例1中描述，此处不再赘述。
本实施例中，计算设备可以是台式电脑、笔记本电脑、智能手机、PDA手持终端、平板电脑等终端设备。
综上，本发明针对人车共驾模式下，多个驾驶人同时改变驾驶意图，易导致相邻区域范围内车辆间行车意图发生冲突的场景，协同调度区域范围内全部车辆的驾驶意图，生成全局调度顺序指导车辆行车决策，一方面避免车辆冲突风险，提高行车安全性，另一方面也会提升区域范围内的整体通行效率。
上述实施例为本发明较佳的实施方式，但本发明的实施方式并不受上述实施例的限制，其他的任何未背离本发明的精神实质与原理下所作的改变、修饰、替代、组合、简化，均应 为等效的置换方式，都包含在本发明的保护范围之内。

Claims (10)

  1. 一种面向智能车辆的区域协同驾驶意图调度方法，其特征在于，包括：
    获取车辆的状态信息和位置信息；
    获取通过车辆的状态信息所识别出的车辆的驾驶意图信息；
    根据调度区域范围内各车辆的驾驶意图信息、车辆的状态信息和车辆的位置信息，生成调度区域范围内全部车辆的驾驶意图全局图；
    构建调度区域占据栅格地图模型；
    根据构建的调度区域占据栅格地图模型，协同调度区域范围内车辆全局驾驶意图，生成调度区域车辆驾驶意图全局调度结果；使得通过调度区域车辆驾驶意图全局调度结果，指导调度区域范围内车辆的行车决策。
  2. 根据权利要求1所述的面向智能车辆的区域协同驾驶意图调度方法，其特征在于，通过车辆的状态信息和位置信息识别出车辆驾驶意图的过程包括：
    基于卷积神经网络构建驾驶意图识别模型，其中将车辆状态信息作为驾驶意图识别模型的输入量I，由驾驶意图识别模型Softmax层输出对驾驶意图的识别向量w＝(w1,w2,w3,w4,w5)，其中：w1、w2、w3、w4、w5分别为左换道、保持不变、右换道、加速、减速的驾驶意图类别的概率；
    设定各种驾驶意图类别的确信阈值；当某一驾驶意图类别的输出概率大于对应的确信阈值时，则判断车辆存在该类别的驾驶意图C；其中：
    C∈{Ca:左换道，Cb:保持不变，Cc:右换道，Cd:加速，Cf:减速}。
  3. 根据权利要求1所述的面向智能车辆的区域协同驾驶意图调度方法，其特征在于，生成调度区域范围内全部车辆的驾驶意图全局图为：G＝[g i]，1≤i≤N，N为调度区域范围内车辆总数，其中：g i＝(C i,V i,W i,P i)，C i,V i,W i,P i分别为调度区域范围内第i辆车的驾驶意图、绝对速度、方向盘转角和车辆位置信息。
  4. 根据权利要求3所述的面向智能车辆的区域协同驾驶意图调度方法，其特征在于，生成调度区域车辆驾驶意图全局调度结果的过程如下：
    Sa、基于调度区域占据栅格模型，将调度区域范围内每一车辆i的位置信息P i用点坐标表示，1≤i≤N，并映射到占据栅格的对应网格中，将映射有车辆位置信息的网格标识为占用，将未映射有车辆位置信息的网格标识为空闲；
    Sb、根据调度区域内每一车辆i的驾驶意图C i、绝对速度V i和方向盘转角W i，计算每一 车辆i在安全加速度a s和安全时间t s内的纵向位移
    Figure PCTCN2021113984-appb-100001
    和横向位移
    Figure PCTCN2021113984-appb-100002
    并依据纵向位移S i'和横向位移S i”确定车辆i的预测位置信息
    Figure PCTCN2021113984-appb-100003
    1≤i≤N；
    Sc、针对调度区域中的每一个网格，根据调度区域内每一车辆i的预测位置信息
    Figure PCTCN2021113984-appb-100004
    获取预测状态；
    Sd、根据每个网格的预测状态，预测该网格是否存在车辆驾驶意图冲突的现象，具体为：
    针对每个网格，判定其预测状态是否为被多辆车占用；其中：
    当网格的预测状态为被一辆车占用时，即预测该网格将被一辆车占用时，则预测该网格不存在车辆驾驶意图冲突的现象，此时控制预测占用该网格的车辆按照其驾驶意图进行行驶；
    当网格的预测状态为被多辆车占用时，即预测网格将被多辆车占用时，则预测该网格存在车辆驾驶意图冲突的现象；此时进入步骤Se；
    Se、判定预测占用网格的多辆车中，是否有驾驶意图为：保持不变；
    若有，则设置预测占用该网格的所有车辆驾驶意图均为：保持不变；
    若否，则从预测占用该网格的多辆车中，随机选取一辆车辆按照其驾驶意图进行行驶，其他车辆驾驶意图全部设置为：保持不变；
    Sf、基于上述操作确定出调度区域内各车辆的驾驶意图，生成调度结果。
  5. 一种面向智能车辆的区域协同驾驶意图调度装置，其特征在于，包括：
    信息获取模块，用于获取车辆的状态信息和位置信息，用于获取通过车辆的状态信息和位置信息所识别出的车辆驾驶意图；
    驾驶意图全局图生成模块，用于根据调度区域范围内车辆的驾驶意图信息、车辆状态信息和车辆位置信息，生成调度区域范围内全部车辆的驾驶意图全局图；
    地图模型构建模块，用于构建调度区域占据栅格地图模型；
    全局调度结果生成模块，用于根据构建的调度区域占据栅格地图模型，协同调度区域范围内车辆全局驾驶意图，生成调度区域车辆驾驶意图全局调度结果；使得通过调度区域车辆驾驶意图全局调度结果，指导调度区域范围内车辆的行车决策。
  6. 一种面向智能车辆的区域协同驾驶意图调度系统，其特征在于，包括云端调度系统，以及设置在车辆上的车载驾驶意图感知系统和车载驾驶意图控制系统；
    所述车载驾驶意图感知系统连接车载驾驶意图控制系统，用于获取车辆的状态信息和位置信息，并将获取到的车辆的状态信息和位置信息发送给车载驾驶意图控制系统；
    所述车辆驾驶意图控制系统通过无线的方式连接云端调度系统，用于根据车辆的状态信 息识别出车辆的驾驶意图，并且将车辆的驾驶意图信息以及车辆的状态信息和位置信息发送给云端调度系统；
    所述云端调度系统，用于执行权利要求1～4中任一项所述的面向智能车辆的区域协同驾驶意图调度方法。
  7. 根据权利要求6所述的面向智能车辆的区域协同驾驶意图调度系统，其特征在于，所述车载驾驶意图感知系统包括车辆状态获取单元和定位单元；
    所述车辆状态获取单元，用于获取车辆的状态信息，包括车辆的油门踏板状态、方向盘转角、制动踏板状态和车辆绝对速度；
    所述定位单元，用于获取车辆的位置信息，包括车辆GPS经度信息和GPS维度信息。
  8. 根据权利要求6所述的面向智能车辆的区域协同驾驶意图调度系统，其特征在于，所述车载驾驶意图控制系统包括驾驶意图识别单元、通信单元和输出单元；
    所述驾驶意图识别单元，用于根据车辆的状态信息识别出车辆的驾驶意图；
    所述通信单元，用于将车辆的驾驶意图信息以及车辆的状态信息和位置信息发送给云端调度系统；
    所述输出单元，用于接收云端调度系统发送的行车调度区域车辆驾驶意图全局调度结果，使得车辆行车决策系统根据该调度结果指导车辆的行车决策；
    所述云端调度系统包括云端通信服务器和云端协同调度服务器；
    所述云端协同调度服务器，用于执行权利要求1～4中任一项所述的面向智能车辆的区域协同驾驶意图调度方法；
    所述云端通信服务器，用于与车载驾驶意图控制系统中的通信单元进行通信，接收车载驾驶意图控制系统发送的车辆的驾驶意图信息以及车辆的状态信息和位置信息，并且将行车调度区域车辆驾驶意图全局调度结果发送给车载驾驶意图控制系统。
  9. 一种存储介质，其特征在于，存储有程序，其特征在于，所述程序被处理器执行时，实现权利要求1～4中任一项所述的面向智能车辆的区域协同驾驶意图调度方法。
  10. 一种计算设备，包括处理器以及用于存储处理器可执行程序的存储器，其特征在于，所述处理器执行存储器存储的程序时，实现权利要求1～4中任一项所述的面向智能车辆的区域协同驾驶意图调度方法。
PCT/CN2021/113984 2020-10-15 2021-08-23 面向智能车辆的区域协同驾驶意图调度方法、系统和介质 WO2022078067A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US18/549,989 US20240217559A1 (en) 2020-10-15 2021-08-23 Smart vehicle-oriented method and system for collaborative dispatching of driving intents in area

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011102809.7A CN112230657B (zh) 2020-10-15 2020-10-15 面向智能车辆的区域协同驾驶意图调度方法、系统和介质
CN202011102809.7 2020-10-15

Publications (1)

Publication Number Publication Date
WO2022078067A1 true WO2022078067A1 (zh) 2022-04-21

Family

ID=74113735

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/113984 WO2022078067A1 (zh) 2020-10-15 2021-08-23 面向智能车辆的区域协同驾驶意图调度方法、系统和介质

Country Status (3)

Country Link
US (1) US20240217559A1 (zh)
CN (1) CN112230657B (zh)
WO (1) WO2022078067A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114822048A (zh) * 2022-05-25 2022-07-29 云控智行科技有限公司 一种网联车的云端规划服务系统及方法
CN115840404A (zh) * 2022-12-21 2023-03-24 浙江大学 一种基于自动驾驶专用路网和数字孪生地图的云控自动驾驶系统

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112230657B (zh) * 2020-10-15 2024-03-29 广州大学 面向智能车辆的区域协同驾驶意图调度方法、系统和介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2013087067A1 (de) * 2011-12-14 2013-06-20 Continental Teves Ag & Co. Ohg Freirauminformationen in einem belegungsgitter als grundlage für die bestimmung eines manöverraums für ein fahrzeug
CN108919803A (zh) * 2018-07-04 2018-11-30 北京踏歌智行科技有限公司 一种矿用无人驾驶车辆的协同控制方法及装置
US10235885B2 (en) * 2016-01-29 2019-03-19 Electronics And Telecommunications Research Institute Autonomous vehicle driving system and method
CN110164183A (zh) * 2019-05-17 2019-08-23 武汉理工大学 一种在车车通信条件下考虑他车驾驶意图的安全辅助驾驶预警方法
CN111487975A (zh) * 2020-04-30 2020-08-04 畅加风行(苏州)智能科技有限公司 一种基于智能网联系统的港口卡车自动编队方法及系统
CN112230657A (zh) * 2020-10-15 2021-01-15 广州大学 面向智能车辆的区域协同驾驶意图调度方法、系统和介质

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108297864A (zh) * 2018-01-25 2018-07-20 广州大学 驾驶员与车辆主动安全技术联动的控制方法及控制系统
CN108447308A (zh) * 2018-05-17 2018-08-24 中交信息技术国家工程实验室有限公司 一种基于车路协同的交叉口车辆碰撞风险预测方法及系统
CN110136254B (zh) * 2019-06-13 2019-12-13 吉林大学 基于动态概率驾驶地图的驾驶辅助信息显示方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2013087067A1 (de) * 2011-12-14 2013-06-20 Continental Teves Ag & Co. Ohg Freirauminformationen in einem belegungsgitter als grundlage für die bestimmung eines manöverraums für ein fahrzeug
US10235885B2 (en) * 2016-01-29 2019-03-19 Electronics And Telecommunications Research Institute Autonomous vehicle driving system and method
CN108919803A (zh) * 2018-07-04 2018-11-30 北京踏歌智行科技有限公司 一种矿用无人驾驶车辆的协同控制方法及装置
CN110164183A (zh) * 2019-05-17 2019-08-23 武汉理工大学 一种在车车通信条件下考虑他车驾驶意图的安全辅助驾驶预警方法
CN111487975A (zh) * 2020-04-30 2020-08-04 畅加风行(苏州)智能科技有限公司 一种基于智能网联系统的港口卡车自动编队方法及系统
CN112230657A (zh) * 2020-10-15 2021-01-15 广州大学 面向智能车辆的区域协同驾驶意图调度方法、系统和介质

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
JI XUEWU, FEI CONG;HE XIANG-KUN;LIU YU-LONG;LIU YA-HUI: "Intention Recognition and Trajectory Prediction for Vehicles Using LSTM Network", CHINA JOURNAL OF HIGHWAY AND TRANSPORT, CN, vol. 32, no. 6, 30 June 2019 (2019-06-30), CN , pages 34 - 42, XP055922095, ISSN: 1001-7372, DOI: 10.19721/j.cnki.1001-7372.2019.06.003 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114822048A (zh) * 2022-05-25 2022-07-29 云控智行科技有限公司 一种网联车的云端规划服务系统及方法
CN115840404A (zh) * 2022-12-21 2023-03-24 浙江大学 一种基于自动驾驶专用路网和数字孪生地图的云控自动驾驶系统
CN115840404B (zh) * 2022-12-21 2023-11-03 浙江大学 一种基于自动驾驶专用路网和数字孪生地图的云控自动驾驶系统

Also Published As

Publication number Publication date
CN112230657B (zh) 2024-03-29
CN112230657A (zh) 2021-01-15
US20240217559A1 (en) 2024-07-04

Similar Documents

Publication Publication Date Title
WO2022078067A1 (zh) 面向智能车辆的区域协同驾驶意图调度方法、系统和介质
EP3965004B1 (en) Automatic lane changing method and device, and storage medium
CN111898211A (zh) 基于深度强化学习的智能车速度决策方法及其仿真方法
WO2021244207A1 (zh) 训练驾驶行为决策模型的方法及装置
CN108292474A (zh) 派遣和维护自主车辆的车队的协调
JP2023506099A (ja) 車両交通の適応制御
Wang et al. Research on eco-driving optimization of hybrid electric vehicle queue considering the driving style
CN113311817B (zh) 车辆协同控制测试方法、装置、电子设备及存储介质
Cai et al. Task-efficiency oriented V2X communications: Digital twin meets mobile edge computing
WO2023174080A1 (zh) 一种路径规划方法和装置以及一种云平台
US11377120B1 (en) Autonomous vehicle control based on risk-based interactions
US20240132109A1 (en) Vehicle Control Method, Apparatus, and System
CN114644016A (zh) 车辆自动驾驶决策方法、装置、车载终端和存储介质
US20230256999A1 (en) Simulation of imminent crash to minimize damage involving an autonomous vehicle
CN113879182A (zh) 车辆能量管理控制方法、系统、设备及介质
US20230339519A1 (en) Safety measurement for a counterfactual risk in autonomous vehicle driving
CN116629114A (zh) 多智能体的模型训练方法、系统、计算机设备和存储介质
US20230192077A1 (en) Adjustment of object trajectory uncertainty by an autonomous vehicle
Xia et al. Lane scheduling around crossroads for edge computing based autonomous driving
CN117104272A (zh) 一种智能驾驶方法、系统、交通工具及存储介质
US11904905B2 (en) Dynamic adjustment of autonomous vehicle system based on deep learning optimizations
WO2021229671A1 (ja) 走行支援装置および走行支援方法
US20230025222A1 (en) Method and system for ctrolling intelligent network vehicle
JP7503618B2 (ja) 車両のスケジューリング方法、装置、電子デバイス、記憶媒体、及びプログラム
Esmaeili et al. A platoon formation strategy for heterogeneous vehicle types at a signalized intersection

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21879114

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 18549989

Country of ref document: US

122 Ep: pct application non-entry in european phase

Ref document number: 21879114

Country of ref document: EP

Kind code of ref document: A1