WO2020143288A1 - 一种复杂工况下自动驾驶车辆决策***及其轨迹规划方法 - Google Patents

一种复杂工况下自动驾驶车辆决策***及其轨迹规划方法 Download PDF

Info

Publication number
WO2020143288A1
WO2020143288A1 PCT/CN2019/115628 CN2019115628W WO2020143288A1 WO 2020143288 A1 WO2020143288 A1 WO 2020143288A1 CN 2019115628 W CN2019115628 W CN 2019115628W WO 2020143288 A1 WO2020143288 A1 WO 2020143288A1
Authority
WO
WIPO (PCT)
Prior art keywords
vehicle
trajectory
decision
speed
unit
Prior art date
Application number
PCT/CN2019/115628
Other languages
English (en)
French (fr)
Inventor
赵万忠
徐灿
王春燕
陈青云
Original Assignee
南京航空航天大学
南京航空航天大学秦淮创新研究院
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 南京航空航天大学, 南京航空航天大学秦淮创新研究院 filed Critical 南京航空航天大学
Publication of WO2020143288A1 publication Critical patent/WO2020143288A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/0088Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots characterized by the autonomous decision making process, e.g. artificial intelligence, predefined behaviours
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02Control of position or course in two dimensions
    • G05D1/021Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0212Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
    • G05D1/0223Control 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
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02Control of position or course in two dimensions
    • G05D1/021Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0231Control of position or course in two dimensions specially adapted to land vehicles using optical position detecting means
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02Control of position or course in two dimensions
    • G05D1/021Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0255Control of position or course in two dimensions specially adapted to land vehicles using acoustic signals, e.g. ultra-sonic singals
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02Control of position or course in two dimensions
    • G05D1/021Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0257Control of position or course in two dimensions specially adapted to land vehicles using a radar
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02Control of position or course in two dimensions
    • G05D1/021Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0276Control of position or course in two dimensions specially adapted to land vehicles using signals provided by a source external to the vehicle
    • G05D1/0278Control 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

Definitions

  • the invention belongs to the technical field of vehicle automatic driving, and in particular relates to an automatic driving vehicle decision system and a trajectory planning method under complex working conditions.
  • the system is the brain of autonomous driving. It mainly plans a safe trajectory for autonomous vehicles based on traffic information sensed by vehicle sensors, such as surrounding vehicle speed, yaw rate, lane line and road boundary information. After the corresponding trajectory is obtained, the corresponding control quantity is obtained through the vehicle dynamics model and passed to the lower-level execution system, so as to control the steering wheel, brakes and throttle to realize the automatic driving of the vehicle.
  • the current research on vehicle decision-making systems is mainly for decision-making under static conditions, that is, traffic conditions determined by the movement of surrounding vehicles. For actual complex traffic, the future movement of surrounding vehicles may change and needs to be predicted. It’s not the same as following the car or changing lanes. In this situation, the decision-making system needs to ensure that the vehicle avoids the surrounding obstacles in real time and improve the driving safety of the vehicle.
  • the trajectory planning technology is a key technology for the decision-making system of autonomous vehicles.
  • the current research mainly focuses on the single planning of speed or path, such as planning only the speed when following the car, and keeping the speed unchanged when changing lanes. path.
  • speed or path In order to realize automatic driving in the true sense, especially in the face of overtaking conditions, it is necessary to plan the speed and path at the same time when planning the trajectory. Therefore, how to plan a collision-free trajectory by controlling the speed and movement direction of vehicles under actual complicated traffic conditions is very important.
  • the object of the present invention is to provide an automatic driving vehicle decision-making system and its trajectory planning method under complex working conditions, in order to solve the existing technology in the multi-lane, high-speed scenes and other complex working conditions
  • a safe and stable trajectory can be planned in real time through the technical solution of the present invention to realize automatic driving of vehicles.
  • the invention discloses a decision-making system for autonomous driving vehicles under complex working conditions, including: an environment perception unit, a real-time decision unit, a trajectory planning unit and a control unit; wherein,
  • Environment awareness unit used to sense the motion information of autonomous vehicles and surrounding vehicles in real time
  • Real-time decision-making unit including the on-board computer on the autonomous driving vehicle and the CAN communication module connected to it; the CAN communication module is connected to the above-mentioned environment sensing unit and receives the above-mentioned motion information; the on-board computer decides the automatic driving vehicle in the current state according to the motion information Best driving behavior;
  • the trajectory planning unit is connected to the above-mentioned real-time decision-making unit, which performs double planning on the path and speed of the vehicle according to the optimal driving behavior of the autonomous vehicle decided by the real-time decision-making unit to obtain the optimal trajectory;
  • the control unit includes the electronic control unit and the active steering motor, brake and accelerator pedal displacement motor electrically connected thereto; the electronic control unit generates the active steering motor, brake and throttle according to the optimal trajectory planned by the above trajectory planning unit The control command of the pedal displacement motor, which in turn controls the driving state of the vehicle.
  • the environment sensing unit includes a GPS, a laser radar, a camera, an ultrasonic radar, and a yaw rate sensor provided on the autonomous driving vehicle.
  • the motion information includes position information, speed information, and yaw angle information.
  • trajectory planning unit considers the safety, efficiency, and stability constraints of the vehicle when planning.
  • the risk constraint That is to ensure that the hazard degree R corresponding to the optimal trajectory is less than the hazard level acceptable to passengers
  • the three-dimensional decision field can be constrained into a two-dimensional plane; the specific solution is as follows:
  • is the weighting factor
  • s t represents the state parameters corresponding to the given trajectory at time t, specifically including the horizontal and vertical coordinates, speed, and yaw angle information
  • p t represents the state parameters corresponding to the surrounding vehicles at time t, specifically including the horizontal and vertical Coordinate, speed, yaw angle information
  • F is the risk assessment function
  • Efficiency constraints ensures that the vehicle travels along the optimal speed provided by the traffic road. According to this efficiency constraint, the above two-dimensional plane can be constrained into a curve; where v final is the corresponding speed after a given trajectory for several cycles, The best speed of the corresponding vehicle under current traffic conditions;
  • Stability constraint ensures that the vehicle changes lanes less, thereby improving riding comfort; according to this constraint, the above curve can be constrained to an optimal decision point, and the autonomous driving vehicle can be determined by the decision point The optimal driving trajectory at the next moment; where ⁇ y is the longitudinal deviation of the determined trajectory from the current road.
  • the invention also discloses a trajectory planning method of the automatic driving vehicle decision system under complex working conditions. Based on the above system, it includes the following steps:
  • each target position point represents a trajectory, that is, the traversal trajectory
  • step 1) specifically includes the following steps:
  • the path corresponding to the future traversal trajectory of the vehicle is fitted using a polynomial, using 4 constraints, and the third-order polynomial is used for fitting, as follows:
  • a 0 , a 1 , a 2 , a 3 are the fitting parameters corresponding to the polynomial path, Is the yaw angle corresponding to the vehicle at the current moment, (x k , y k ) is the position coordinate corresponding to the current state of the automatic driving vehicle, (x p , y p ) is the target point reached by the automatic driving vehicle after a given number of cycles ;
  • T is the time corresponding to each cycle
  • v k is the current speed of the vehicle
  • a is the acceleration corresponding to the process
  • v k is the speed corresponding to the vehicle at the current moment
  • v t is the speed corresponding to the vehicle after t cycles
  • the optimal trajectory search method in step 2) specifically includes the following steps:
  • S r is the road safety factor
  • D sf is the standard safety distance
  • b is the limiting factor
  • Des is the actual distance between the vehicle and the surrounding vehicle
  • t is the time when the surrounding vehicle reaches the front of the vehicle
  • t b is the braking time
  • the constraint of the three-dimensional hazard field in step 22) specifically includes:
  • is the weighting factor
  • s t represents the state parameters corresponding to the given trajectory at time t, specifically including the horizontal and vertical coordinates, speed, and yaw angle information
  • p t represents the state parameters corresponding to the surrounding vehicles at time t, specifically including the horizontal and vertical Coordinate, speed, yaw angle information
  • F is the risk assessment function
  • Efficiency constraints Constrain the above two-dimensional plane to a curve; where v final is the corresponding velocity after a certain period of a given trajectory, The best speed of the corresponding vehicle under current traffic conditions;
  • the invention can realize the automatic driving of vehicles under the complicated traffic conditions of multiple lanes, and can plan a safe and collision-free trajectory in real time.
  • the decision planning system designed by the present invention can fully consider the kinematic and dynamic constraints of the vehicle's actuator, the generated trajectory has continuity, and can be applied to the real-time changing road environment.
  • the trajectory planning of the vehicle by the present invention is a dual planning of path and speed, that is, planning the path of the vehicle while planning the speed of the vehicle at each point on the path.
  • Figure 1 is an overall block diagram of the decision-making system of the present invention.
  • FIG. 2 is a schematic diagram corresponding to an optimization search method for a given target position point.
  • a decision-making system for an autonomous driving vehicle under complex working conditions of the present invention is characterized by comprising: an environment awareness unit, a real-time decision unit, a trajectory planning unit and a control unit; wherein,
  • Environment awareness unit including GPS, lidar, camera, ultrasonic radar, and yaw rate sensor installed on the self-driving vehicle; used to sense the motion information of the self-driving vehicle and surrounding vehicles in real time; the motion information includes: location information, Speed information and yaw angle information.
  • Real-time decision-making unit including the on-board computer on the autonomous driving vehicle and the CAN communication module connected to it; the CAN communication module is connected to the above-mentioned environment sensing unit and receives the above-mentioned motion information; the on-board computer decides the automatic driving vehicle in the current state according to the motion information Best driving behavior;
  • the trajectory planning unit is connected to the above-mentioned real-time decision-making unit, which double-plans the path and speed of the vehicle according to the best driving behavior of the self-driving vehicle decided by the real-time decision-making unit, and obtains the optimal trajectory; the safety of the vehicle is considered when planning Sexual constraints, high efficiency constraints, stability constraints.
  • is the weighting factor
  • s t represents the state parameters corresponding to the given trajectory at time t, specifically including the horizontal and vertical coordinates, speed, and yaw angle information
  • p t represents the state parameters corresponding to the surrounding vehicles at time t, specifically including the horizontal and vertical Coordinate, speed, yaw angle information
  • F is the risk assessment function
  • the control unit includes an electronic control unit (ECU) and an active steering motor, a brake and an accelerator pedal displacement motor electrically connected thereto; the electronic control unit generates an active steering motor, a control system according to the optimal trajectory planned by the above trajectory planning unit Control commands for moving and accelerator pedal displacement motors, which in turn control the vehicle's driving status.
  • ECU electronice control unit
  • active steering motor a brake and an accelerator pedal displacement motor electrically connected thereto
  • the electronic control unit generates an active steering motor, a control system according to the optimal trajectory planned by the above trajectory planning unit Control commands for moving and accelerator pedal displacement motors, which in turn control the vehicle's driving status.
  • the environment awareness unit When working, the environment awareness unit first uses various sensors to obtain the motion information around the autonomous vehicle, and transmits this information to the on-board computer for real-time decision-making.
  • the trajectory contains not only the path information of the self-driving vehicle, but also the speed information, so it is well adapted to the decision planning under the actual complex working conditions.
  • the vehicle's control unit After obtaining the trajectory information, the vehicle's control unit will use the corresponding control algorithm to convert the trajectory information into the control quantity information directly input by the controller, and input these control quantities to the active steering motor, brake, and accelerator pedal displacement motor, thereby Realize the automatic driving of vehicles.
  • the invention also discloses a trajectory planning method of the automatic driving vehicle decision system under complex working conditions. Based on the above system, it includes the following steps:
  • each target position point represents a trajectory, that is, the traversal trajectory
  • the method for obtaining the traversing trajectory in step 1) specifically includes the following steps:
  • the path corresponding to the future traversal trajectory of the vehicle is fitted using a polynomial, and the third-order polynomial is used to fit using 4 constraints, as follows:
  • a 0 , a 1 , a 2 , a 3 are the fitting parameters corresponding to the polynomial path, Is the yaw angle corresponding to the vehicle at the current moment, (x k ,y k ) is the position coordinate corresponding to the current state of the automatic driving vehicle, (x p ,y p ) is the target point reached by the automatic driving vehicle after a given number of cycles ;
  • T is the time corresponding to each cycle
  • v k is the current speed of the vehicle
  • a is the acceleration corresponding to the process
  • v k is the speed corresponding to the vehicle at the current moment
  • v t is the speed corresponding to the vehicle after t cycles
  • the optimal trajectory search method in step 2) specifically includes the following steps:
  • S r is the road safety factor
  • D sf is the standard safety distance
  • b is the limiting factor
  • Des is the actual distance between the vehicle and the surrounding vehicle
  • t is the time when the surrounding vehicle reaches the front of the vehicle
  • t b is the braking time
  • step 22) restricting the three-dimensional hazard field specifically includes:
  • is the weighting factor
  • s t represents the state parameter corresponding to the given trajectory at time t, specifically including the horizontal and vertical coordinates, speed, and yaw angle information
  • p t represents the state parameter corresponding to the surrounding vehicles at time t, specifically including the horizontal and vertical coordinates , Speed, yaw angle information
  • F is the risk assessment function
  • Efficiency constraints ensures that the vehicle can travel along the optimal speed provided by the traffic road. According to the efficiency constraint, the above two-dimensional plane is constrained into a curve; where, v final is the corresponding speed after a certain period of the given trajectory, The best speed of the corresponding vehicle under current traffic conditions;
  • the tracking method based on vehicle dynamics used in step 3 includes the following steps:
  • Is the yaw rate, Y and X are the vertical and horizontal coordinates in the earth coordinate system;
  • T is the sampling time of each discretization period
  • I is the identity matrix
  • f is the dynamic equation
  • (u 0 , ⁇ 0 ) is the reference state point.
  • is the predicted actual state quantity
  • ⁇ ref is the reference state quantity obtained from the planned trajectory
  • ⁇ U is the control quantity increment
  • the first term of the objective function is the deviation of the predicted trajectory from the reference trajectory, reflecting the trajectory followability
  • the second term of the objective function is the deviation of the control quantity, which reflects stability.
  • Control amount constraint where the control amount is defined as follows:
  • the optimal tracking control quantity sequence of the autonomous vehicle can be solved at every moment, so as to ensure that the vehicle travels along the safest trajectory.

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Physics & Mathematics (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Electromagnetism (AREA)
  • Acoustics & Sound (AREA)
  • Business, Economics & Management (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Game Theory and Decision Science (AREA)
  • Medical Informatics (AREA)
  • Control Of Driving Devices And Active Controlling Of Vehicle (AREA)
  • Traffic Control Systems (AREA)

Abstract

一种复杂工况下自动驾驶车辆决策***及其轨迹规划方法,***主要包括环境感知单元、实时决策单元、轨迹规划单元及控制单元。车辆在行驶过程中,环境感知单元将感知到的周围车辆运动信息传输到实时决策单元,用于决策出当前状态下自动驾驶车辆最佳的驾驶行为。在决策出驾驶行为后,轨迹规划单元再通过设定好的轨迹规划方法对车辆的路径和速度进行双规划,得到当前状态下的最优轨迹,并将相应的控制信号通过ECU输入到执行机构保证车辆安全可靠的行驶。在多车道、多障碍车辆的复杂工况下给车辆实时规划出一条安全无碰撞的轨迹,实现车辆的自动驾驶。

Description

一种复杂工况下自动驾驶车辆决策***及其轨迹规划方法 技术领域
本发明属于车辆自动驾驶技术领域,尤其涉及一种复杂工况下自动驾驶车辆决策***及其轨迹规划方法。
背景技术
随着传感器精度的提高,芯片技术的发展、5G通信的到来,对于车辆自动驾驶的研究越来越多,其研究的目的主要在于减少交通事故,缓解交通堵塞,并减轻驾驶员的驾驶负担。目前很多国际化公司,如谷歌,通用,特斯拉都投入了大量精力来研究高水平的自动驾驶车辆。
作为车辆自动驾驶技术的核心部分,决策***实时决策出一条能避开周围障碍物的安全可靠的轨迹显得尤其重要。该***是自动驾驶的大脑,主要根据车辆传感器感知到的交通信息,如周围车辆速度,横摆角速度,车道线及道路边界信息,来给自动驾驶车辆规划出一条安全的轨迹。得到相应的轨迹后,通过车辆的动力学模型得到相应的控制量并传给下层执行***,从而控制方向盘、刹车和油门来实现车辆的自动驾驶。目前对于车辆决策***的研究,主要是对静止工况下即周围车辆运动确定的交通工况下的决策规划,而对于实际复杂交通中,周围车辆的未来运动可能发生变化需要进行预测,面对的也不是规则的跟车、换道工况。在该情境下,决策***需要保证车辆实时避开周围的障碍物,提高车辆的行车安全性。
此外,轨迹规划技术作为自动驾驶车辆决策***的关键技术,目前的研究主要停留在对速度或者对路径的单一规划,如跟车时只对速度进行规划,换道时则保持速度不变只规划路径。而要实现真正意义上的自动驾驶,尤其面对超车工况时,在轨迹规划时需要对速度及路径进行同时规划。因此,如何在实际复杂交通工况下,通过控制车辆的速度及运动方向规划出一条无碰撞的轨迹显得很重要。
发明内容
针对于上述现有技术的不足,本发明的目的在于提供一种复杂工况下自动驾驶车辆决策***及其轨迹规划方法,以解决现有技术中在多车道、高速场景等复杂工况下的智能决策及轨迹规划问题,通过本发明的技术方案能实时规划出一条安全稳定的轨迹,实现车辆的自动驾驶。
为达到上述目的,本发明采用的技术方案如下:
本发明公开了一种复杂工况下自动驾驶车辆决策***,包括:环境感知单元、实时决策单元、轨迹规划单元及控制单元;其中,
环境感知单元,用于实时感知自动驾驶车辆及周围车辆的运动信息;
实时决策单元,包含自动驾驶车辆上的车载计算机及与之相连的CAN通信模块;CAN通信模块与上述环境感知单元连接,并接收上述运动信息;车载计算机根据运动信息决策出当 前状态下自动驾驶车辆最佳的驾驶行为;
轨迹规划单元,与上述实时决策单元相连接,其根据实时决策单元决策出的自动驾驶车辆最佳的驾驶行为对车辆的路径及速度进行双规划,得到最优轨迹;
控制单元,包含电子控制单元及与之电气连接的主动转向电机、制动及油门踏板位移电机;电子控制单元根据上述轨迹规划单元规划出的最优轨迹,产生对主动转向电机、制动及油门踏板位移电机的控制指令,进而控制车辆行驶状态。
进一步地,所述环境感知单元包含自动驾驶车辆上设有的GPS、激光雷达、摄像头、超声波雷达及横摆角速度传感器。
进一步地,所述运动信息包含:位置信息、速度信息及横摆角信息。
进一步地,所述轨迹规划单元在规划时考虑车辆的安全性约束、高效性约束、平稳性约束。
进一步地,所述的危险性约束:
Figure PCTCN2019115628-appb-000001
即保证最优轨迹对应的危险度R小于乘客可接受的危险等级
Figure PCTCN2019115628-appb-000002
根据该危险性约束,即可将三维决策场约束成一个二维平面;具体求解如下:
Figure PCTCN2019115628-appb-000003
式中,γ为权重因子;s t表示t时刻给定轨迹对应的状态参数,具体包括横纵坐标、速度、横摆角信息;p t表示t时刻周围车辆对应的状态参数,具体包括横纵坐标、速度、横摆角信息;F为危险度评估函数;
高效性约束:
Figure PCTCN2019115628-appb-000004
该约束保证车辆沿着交通道路提供的最佳车速行驶,根据该高效性约束,即可将上述二维平面约束成一条曲线;其中,v final为给定轨迹若干个周期后对应的速度,
Figure PCTCN2019115628-appb-000005
为当前交通状况下对应的车辆最佳车速;
平稳性约束:min Δy,该平稳性约束保证车辆少换道,从而提高乘坐舒适性;根据该约束,即可将上述的曲线约束成最优的决策点,通过该决策点来确定自动驾驶车辆下一时刻的最优行驶轨迹;其中,Δy为决策出的轨迹与当前道路的纵向偏差。
本发明还公开了一种复杂工况下自动驾驶车辆决策***的轨迹规划方法,基于上述***,包括以下步骤:
1)获得遍历轨迹:先给定若干周期后车辆可能到达的目标位置点,每个目标位置点代表一条轨迹,即遍历轨迹;
2)根据上述遍历轨迹进行最优化搜索,通过车辆面对的危险性、高效性和稳定性约束来优化出当前时刻车辆最优的一条轨迹。
进一步地,所述步骤1)中获得遍历轨迹的方法,具体包括如下步骤:
11)根据车辆当前位置及给定的目标位置点,利用多项式拟合出车辆未来的遍历轨迹对 应的路径,利用4个约束,使用3次多项式进行拟合,具体如下:
y=a 0+a 1x+a 2x 2+a 3x 3
Figure PCTCN2019115628-appb-000006
其中,a 0,a 1,a 2,a 3分别为多项式路径对应的拟合参数,
Figure PCTCN2019115628-appb-000007
为当前时刻车辆对应的横摆角,(x k,y k)为当前状态自动驾驶车辆对应的位置坐标,(x p,y p)为给定的若干个周期之后自动驾驶车辆到达的目标点;
12)利用积分求出上述路径对应的长度,将该过程视为匀加速运动过程,得到遍历轨迹对应的速度;
每条决策路径的长度S:
Figure PCTCN2019115628-appb-000008
其中,y’为多项式路径的斜率;
将该过程视为匀加速运动;从而得到每个决策点对应的加速度:
Figure PCTCN2019115628-appb-000009
其中,T为每个周期对应的时间,v k为车辆当前速度,a为该过程对应的加速度;
该过程对应的速度表示如下:
v t=v k+a*(t-k)
其中,v k为当前时刻车辆对应的速度,v t表示t个周期后车辆对应的速度;
13)根据步骤11)拟合出的路径和步骤12)得到的在该路径上的速度,即得到车辆遍历轨迹的参数信息(路径+速度)。
进一步的,所述步骤2)中的最优轨迹搜索方法,具体包括如下步骤:
21)获取每条轨迹对应的危险度,该危险度即可形成一个车辆在未来时刻的三维危险场,其中危险度评估函数F建立如下:
Figure PCTCN2019115628-appb-000010
其中,S r为道路安全系数;D sf为标准安全距离;b为限幅系数;D es为自车与周围车辆实际距 离,t为周围车辆到达自车前方的时间,t b为刹车时间;
22)对上述三维危险场进行约束,得到一条最优轨迹。
进一步地,所述步骤22)中对三维危险场进行约束具体包括:
221)危险性约束:
Figure PCTCN2019115628-appb-000011
即保证最优轨迹对应的危险度R小于乘客可接受的危险等级
Figure PCTCN2019115628-appb-000012
根据该约束,即可将三维决策场约束成一个二维平面;具体求解如下:
Figure PCTCN2019115628-appb-000013
式中,γ为权重因子;s t表示t时刻给定轨迹对应的状态参数,具体包括横纵坐标、速度、横摆角信息;p t表示t时刻周围车辆对应的状态参数,具体包括横纵坐标、速度、横摆角信息;F为危险度评估函数;
222)高效性约束:
Figure PCTCN2019115628-appb-000014
将上述二维平面约束成一条曲线;其中,v final为给定轨迹若干个周期后对应的速度,
Figure PCTCN2019115628-appb-000015
为当前交通状况下对应的车辆最佳车速;
223)平稳性约束:min Δy,将上述的曲线约束成最优的决策点,通过该决策点来确定自动驾驶车辆t+1时刻的最优行驶轨迹;其中,Δy为决策出的轨迹与当前道路的纵向偏差。
本发明的有益效果:
1、本发明能实现车辆在多车道复杂交通工况下的自动驾驶,能实时规划出一条安全,无碰撞的轨迹。
2、本发明所设计的决策规划***能充分考虑车辆执行机构的运动学和动力学约束,所生成的轨迹具有连续性,能适用实时变化的道路环境。
3、本发明对车辆进行的轨迹规划,是路径与速度的双规划,即在给车辆规划路径的同时规划处车辆在该路径上各点的速度。
附图说明
图1为本发明决策***总体框图。
图2为给定目标位置点进行最优化搜索方法对应的示意图。
具体实施方式
为了便于本领域技术人员的理解,下面结合实施例与附图对本发明作进一步的说明,实施方式提及的内容并非对本发明的限定。
参照图1所示,本发明的一种复杂工况下自动驾驶车辆决策***,其特征在于,包括:环境感知单元、实时决策单元、轨迹规划单元及控制单元;其中,
环境感知单元,包含自动驾驶车辆上设有的GPS、激光雷达、摄像头、超声波雷达及横摆角速度传感器;用于实时感知自动驾驶车辆及周围车辆的运动信息;所述运动信息包含:位置信息、速度信息及横摆角信息。
实时决策单元,包含自动驾驶车辆上的车载计算机及与之相连的CAN通信模块;CAN通信模块与上述环境感知单元连接,并接收上述运动信息;车载计算机根据运动信息决策出当前状态下自动驾驶车辆最佳的驾驶行为;
轨迹规划单元,与上述实时决策单元相连接,其根据实时决策单元决策出的自动驾驶车辆最佳的驾驶行为对车辆的路径及速度进行双规划,得到最优轨迹;在规划时考虑车辆的安全性约束、高效性约束、平稳性约束。
危险性约束:
Figure PCTCN2019115628-appb-000016
即保证最优轨迹对应的危险度R小于乘客可接受的危险等级
Figure PCTCN2019115628-appb-000017
根据该危险性约束,即可将三维决策场约束成一个二维平面;具体求解如下:
Figure PCTCN2019115628-appb-000018
式中,γ为权重因子;s t表示t时刻给定轨迹对应的状态参数,具体包括横纵坐标、速度、横摆角信息;p t表示t时刻周围车辆对应的状态参数,具体包括横纵坐标、速度、横摆角信息;F为危险度评估函数;
高效性约束:
Figure PCTCN2019115628-appb-000019
将上述二维平面约束成一条曲线;其中,v final为给定轨迹若干个周期后对应的速度,
Figure PCTCN2019115628-appb-000020
为当前交通状况下对应的车辆最佳车速;
平稳性约束:min Δy,将上述的曲线约束成最优的决策点,通过该决策点来确定自动驾驶车辆下一时刻的最优行驶轨迹;其中,Δy为决策出的轨迹与当前道路的纵向偏差。
控制单元,包含电子控制单元(ECU)及与之电气连接的主动转向电机、制动及油门踏板位移电机;电子控制单元根据上述轨迹规划单元规划出的最优轨迹,产生对主动转向电机、制动及油门踏板位移电机的控制指令,进而控制车辆行驶状态。
在工作时,环境感知单元先利用各种传感器来获取自动驾驶车辆周围的运动信息,并将这些信息传输到车载计算机内进行实时决策,具体决策时,利用“遍历轨迹+最优搜索法”来进行轨迹规划,该轨迹既包含了自动驾驶车辆的路径信息,又包含了速度信息,从而很好的适应实际复杂工况下的决策规划。得到轨迹信息后,车辆的控制单元会利用相应的控制算法将轨迹信息转化为控制器直接输入的控制量信息,并将这些控制量输入到主动转向电机、制动及油门踏板位移电机上,从而实现车辆的自动驾驶。
本发明还公开了一种复杂工况下自动驾驶车辆决策***的轨迹规划方法,基于上述***,包括以下步骤:
1)获得遍历轨迹:先给定若干周期后车辆可能到达的目标位置点,每个目标位置点代表一条轨迹,即遍历轨迹;
2)根据上述遍历轨迹进行最优化搜索,通过车辆面对的危险性、高效性和稳定性约束来优化出当前时刻车辆最优的一条轨迹。
所述步骤1)中获得遍历轨迹的方法,具体包括如下步骤:
11)根据车辆当前位置及给定的目标位置点,利用多项式拟合出车辆未来的遍历轨迹对应的路径,利用4个约束,使用3次多项式进行拟合,具体如下:
y=a 0+a 1x+a 2x 2+a 3x 3
Figure PCTCN2019115628-appb-000021
其中,a 0,a 1,a 2,a 3分别为多项式路径对应的拟合参数,
Figure PCTCN2019115628-appb-000022
为当前时刻车辆对应的横摆角,(x k,y k)为当前状态自动驾驶车辆对应的位置坐标,(x p,y p)为给定的若干个周期之后自动驾驶车辆到达的目标点;
12)利用积分求出上述路径对应的长度,将该过程视为匀加速运动过程,得到遍历轨迹对应的速度;
每条决策路径的长度S:
Figure PCTCN2019115628-appb-000023
其中,y’为多项式路径的斜率;
将该过程视为匀加速运动;从而得到每个决策点对应的加速度:
Figure PCTCN2019115628-appb-000024
其中,T为每个周期对应的时间,v k为车辆当前速度,a为该过程对应的加速度;
该过程对应的速度表示如下:
v t=v k+a*(t-k)
其中,v k为当前时刻车辆对应的速度,v t表示t个周期后车辆对应的速度;
13)根据步骤11)拟合出的路径和步骤12)得到的在该路径上的速度,即得到了车辆遍历轨迹的参数信息(路径+速度)。
参照图2所示,所述步骤2)中的最优轨迹搜索方法,具体包括如下步骤:
21)获取每条轨迹对应的危险度,该危险度即可形成一个车辆在未来时刻的三维危险场,其中危险度评估函数F建立如下:
Figure PCTCN2019115628-appb-000025
其中,S r为道路安全系数;D sf为标准安全距离;b为限幅系数;D es为自车与周围车辆实际距离,t为周围车辆到达自车前方的时间,t b为刹车时间;
22)对上述三维危险场进行约束,得到一条最优轨迹。
其中,所述步骤22)中对三维危险场进行约束具体包括:
221)危险性约束:
Figure PCTCN2019115628-appb-000026
即保证最优轨迹对应的危险度R小于乘客可接受的危险等级
Figure PCTCN2019115628-appb-000027
根据该约束,即可将三维决策场约束成一个二维平面;具体求解如下:
Figure PCTCN2019115628-appb-000028
其中,γ为权重因子;s t表示t时刻给定轨迹对应的状态参数,具体包括横纵坐标、速度、横摆角信息;p t表示t时刻周围车辆对应的状态参数,具体包括横纵坐标、速度、横摆角信息;F为危险度评估函数;
222)高效性约束:
Figure PCTCN2019115628-appb-000029
该约束保证车辆能沿着交通道路提供的最佳车速行驶,根据该高效性约束,将上述二维平面约束成一条曲线;其中,v final为给定轨迹若干个周期后对应的速度,
Figure PCTCN2019115628-appb-000030
为当前交通状况下对应的车辆最佳车速;
223)平稳性约束:min Δ y,该平稳性约束能保证车辆少换道,从而提高乘坐舒适性;根据该约束,将上述的曲线约束成最优的决策点,通过该决策点来确定自动驾驶车辆t+1时刻的最优行驶轨迹;其中,Δ y为决策出的轨迹与当前道路的纵向偏差。
3)以上得到了车辆在当前状态下的最优轨迹,下面需要控制车辆跟踪该轨迹,基于轮胎小角度转角下的线性化假设建立如下车辆动力学模型进行跟踪。
其中,步骤3)中使用的基于车辆动力学的跟踪方法,包括如下步骤:
31)基于动力学的车辆状态方程获得:
当车辆控制量u dyn为前轮转角σ f及车速v,即u dyn=[σ f,v]时,状态量为车辆运动轨迹序列
Figure PCTCN2019115628-appb-000031
时,
其中,
Figure PCTCN2019115628-appb-000032
为车辆纵向速度,
Figure PCTCN2019115628-appb-000033
为车辆侧向速度时,
Figure PCTCN2019115628-appb-000034
为车辆横摆角,
Figure PCTCN2019115628-appb-000035
为横摆角速度,Y,X为地球坐标系下的纵横坐标;
可建立如下离散化的状态方程:
Figure PCTCN2019115628-appb-000036
其中,k为当前时刻,T是每个离散化周期的采样时间,I为单位矩阵,f为动力学方程,(u 00)为参考状态点。
32)跟踪目标函数的建立:
有以上基于动力学的预测方程,则可进一步根据预测的轨迹与期望轨迹的偏差来跟踪,从而控制车辆安全行驶,具体采用如下目标函数J(k):
Figure PCTCN2019115628-appb-000037
其中,η为预测的实际状态量,η ref为规划轨迹得到的参考状态量,ΔU为控制量增量;该目标函数的第一项是预测轨迹与参考轨迹的偏差,体现轨迹跟随性;该目标函数的第二项是控制量的偏差,体现稳定性。
33)对于以上目标函数,通过以下两个约束求得最终轨迹跟踪序列;
321)控制量约束:其中控制量定义如下:
u min(t+k)≤u(t+k)≤u max(t+k),k=0,1,L,20
322)控制增量约束:
Δu min(t+k)≤Δu(t+k)≤Δu max(t+k),k=0,1,L,20
有了以上约束,即可在每个时刻求解出自动驾驶车辆最佳的跟踪控制量序列,从而保证车辆沿着最安全的轨迹行驶。
本发明具体应用途径很多,以上所述仅是本发明的优选实施方式,应当指出,对于本技术领域的普通技术人员来说,在不脱离本发明原理的前提下,还可以作出若干改进,这些改进也应视为本发明的保护范围。

Claims (9)

  1. 一种复杂工况下自动驾驶车辆决策***,其特征在于,包括:环境感知单元、实时决策单元、轨迹规划单元及控制单元;其中,
    环境感知单元,用于实时感知自动驾驶车辆及周围车辆的运动信息;
    实时决策单元,包含自动驾驶车辆上的车载计算机及与之相连的CAN通信模块;CAN通信模块与上述环境感知单元连接,并接收上述运动信息;车载计算机根据运动信息决策出当前状态下自动驾驶车辆最佳的驾驶行为;
    轨迹规划单元,与上述实时决策单元相连接,其根据实时决策单元决策出的自动驾驶车辆最佳的驾驶行为对车辆的路径及速度进行双规划,得到最优轨迹;
    控制单元,包含电子控制单元及与之电气连接的主动转向电机、制动及油门踏板位移电机;电子控制单元根据上述轨迹规划单元规划出的最优轨迹,产生对主动转向电机、制动及油门踏板位移电机的控制指令,进而控制车辆行驶状态。
  2. 根据权利要求1所述的复杂工况下自动驾驶车辆决策***,其特征在于,所述环境感知单元包含自动驾驶车辆上设有的GPS、激光雷达、摄像头、超声波雷达及横摆角速度传感器。
  3. 根据权利要求1所述的复杂工况下自动驾驶车辆决策***,其特征在于,所述运动信息包含:位置信息、速度信息及横摆角信息。
  4. 根据权利要求1所述的复杂工况下自动驾驶车辆决策***,其特征在于,所述轨迹规划单元在规划时考虑车辆的安全性约束、高效性约束、平稳性约束。
  5. 根据权利要求4所述的复杂工况下自动驾驶车辆决策***,其特征在于,所述的危险性约束:
    Figure PCTCN2019115628-appb-100001
    即保证最优轨迹对应的危险度R小于乘客可接受的危险等级
    Figure PCTCN2019115628-appb-100002
    根据该危险性约束,即将三维决策场约束成一个二维平面;具体求解如下:
    Figure PCTCN2019115628-appb-100003
    其中,γ为权重因子;s t表示t时刻给定轨迹对应的状态参数,具体包括横纵坐标、速度、横摆角信息;p t表示t时刻周围车辆对应的状态参数,具体包括横纵坐标、速度、横摆角信息;F为危险度评估函数;
    高效性约束:
    Figure PCTCN2019115628-appb-100004
    将上述二维平面约束成一条曲线;其中,v final为给定轨迹若干个周期后对应的速度,
    Figure PCTCN2019115628-appb-100005
    为当前交通状况下对应的车辆最佳车速;
    平稳性约束:minΔy,将上述的曲线约束成最优的决策点,通过该决策点来确定自动驾驶车辆下一时刻的最优行驶轨迹;其中,Δy为决策出的轨迹与当前道路的纵向偏差。
  6. 一种复杂工况下自动驾驶车辆决策***的轨迹规划方法,基于上述权利要求1至5中任意一项所述的***,包括以下步骤:
    1)获得遍历轨迹:先给定若干周期后车辆可能到达的目标位置点,每个目标位置点代表一条轨迹,即遍历轨迹;
    2)根据上述遍历轨迹进行最优化搜索,通过车辆面对的危险性、高效性和稳定性约束来优化出当前时刻车辆最优的一条轨迹。
  7. 根据权利要求6所述的复杂工况下自动驾驶车辆决策***的轨迹规划方法,其特征在于,所述步骤1)中获得遍历轨迹的方法,具体包括如下步骤:
    11)根据车辆当前位置及给定的目标位置点,利用多项式拟合出车辆未来的遍历轨迹对应的路径,利用4个约束,使用3次多项式进行拟合,具体如下:
    y=a 0+a 1x+a 2x 2+a 3x 3
    Figure PCTCN2019115628-appb-100006
    其中,a 0,a 1,a 2,a 3分别为多项式路径对应的拟合参数,
    Figure PCTCN2019115628-appb-100007
    为当前时刻车辆对应的横摆角,(x k,y k)为当前状态自动驾驶车辆对应的位置坐标,(x p,y p)为给定的若干个周期之后自动驾驶车辆到达的目标点;
    12)利用积分求出上述路径对应的长度,将该过程视为匀加速运动过程,得到遍历轨迹对应的速度;
    每条决策路径的长度S:
    Figure PCTCN2019115628-appb-100008
    其中,y’为多项式路径的斜率;
    将该过程视为匀加速运动;从而得到每个决策点对应的加速度:
    Figure PCTCN2019115628-appb-100009
    其中,T为每个周期对应的时间,v k为车辆当前速度,a为该过程对应的加速度;
    该过程对应的速度表示如下:
    v t=v k+a*(t-k)
    其中,v k为当前时刻车辆对应的速度,v t表示t个周期后车辆对应的速度;
    13)根据步骤11)拟合出的路径和步骤12)得到的在该路径上的速度,即得到车辆遍历 轨迹的参数信息。
  8. 根据权利要求6所述的复杂工况下自动驾驶车辆决策***的轨迹规划方法,其特征在于,所述步骤2)中的最优轨迹搜索方法,具体包括如下步骤:
    21)获取每条轨迹对应的危险度,该危险度即可形成一个车辆在未来时刻的三维危险场,其中危险度评估函数F建立如下:
    Figure PCTCN2019115628-appb-100010
    其中,S r为道路安全系数;D sf为标准安全距离;b为限幅系数;D es为自车与周围车辆实际距离,t为周围车辆到达自车前方的时间,t b为刹车时间;
    22)对上述三维危险场进行约束,得到一条最优轨迹。
  9. 根据权利要求8所述的复杂工况下自动驾驶车辆决策***的轨迹规划方法,其特征在于,所述步骤22)中对三维危险场进行约束具体包括:
    221)危险性约束:
    Figure PCTCN2019115628-appb-100011
    即保证最优轨迹对应的危险度R小于乘客可接受的危险等级
    Figure PCTCN2019115628-appb-100012
    根据该约束,即可将三维决策场约束成一个二维平面;具体求解如下:
    Figure PCTCN2019115628-appb-100013
    其中,γ为权重因子;s t表示t时刻给定轨迹对应的状态参数,具体包括横纵坐标、速度、横摆角信息;p t表示t时刻周围车辆对应的状态参数,具体包括横纵坐标、速度、横摆角信息;F为危险度评估函数;
    222)高效性约束:
    Figure PCTCN2019115628-appb-100014
    将上述二维平面约束成一条曲线;其中,v final为给定轨迹若干个周期后对应的速度,
    Figure PCTCN2019115628-appb-100015
    为当前交通状况下对应的车辆最佳车速;
    223)平稳性约束:minΔy,将上述的曲线约束成最优的决策点,通过该决策点来确定自动驾驶车辆t+1时刻的最优行驶轨迹;其中,Δy为决策出的轨迹与当前道路的纵向偏差。
PCT/CN2019/115628 2019-01-08 2019-11-05 一种复杂工况下自动驾驶车辆决策***及其轨迹规划方法 WO2020143288A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910015057.1A CN109669461B (zh) 2019-01-08 2019-01-08 一种复杂工况下自动驾驶车辆决策***及其轨迹规划方法
CN201910015057.1 2019-01-08

Publications (1)

Publication Number Publication Date
WO2020143288A1 true WO2020143288A1 (zh) 2020-07-16

Family

ID=66149351

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/115628 WO2020143288A1 (zh) 2019-01-08 2019-11-05 一种复杂工况下自动驾驶车辆决策***及其轨迹规划方法

Country Status (2)

Country Link
CN (1) CN109669461B (zh)
WO (1) WO2020143288A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111968397A (zh) * 2020-07-31 2020-11-20 东风商用车有限公司 自动驾驶行为判定方法
CN112201070A (zh) * 2020-09-29 2021-01-08 上海交通大学 基于深度学习的自动驾车高速公路瓶颈路段行为决策方法

Families Citing this family (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109669461B (zh) * 2019-01-08 2020-07-28 南京航空航天大学 一种复杂工况下自动驾驶车辆决策***及其轨迹规划方法
CN110244729A (zh) * 2019-06-18 2019-09-17 无锡新图云创科技发展有限公司 一种间断性巡磁的agv导航方法
CN110568841A (zh) * 2019-08-05 2019-12-13 西藏宁算科技集团有限公司 一种自动驾驶决策方法及***
CN110703754B (zh) * 2019-10-17 2021-07-09 南京航空航天大学 一种自动驾驶车辆路径与速度高度耦合的轨迹规划方法
CN110989577B (zh) * 2019-11-15 2023-06-23 深圳先进技术研究院 自动驾驶决策方法及车辆的自动驾驶装置
CN110968095A (zh) * 2019-11-23 2020-04-07 奇瑞汽车股份有限公司 一种自动行车控制方法
CN110884488B (zh) * 2019-11-28 2022-05-31 东风商用车有限公司 自动驾驶工程车辅助定位***及其使用方法
CN112257486B (zh) * 2019-12-23 2023-12-29 北京国家新能源汽车技术创新中心有限公司 算力分配方法、装置、计算机设备和存储介质
CN111258312B (zh) * 2020-01-20 2024-04-02 深圳市商汤科技有限公司 可移动模型及其控制方法、装置、***、设备和存储介质
CN111267846B (zh) * 2020-02-11 2021-05-11 南京航空航天大学 一种基于博弈论的周围车辆交互行为预测方法
CN113002540B (zh) * 2020-04-17 2022-11-25 青岛慧拓智能机器有限公司 矿用自卸车控制方法及装置
CN113552869B (zh) * 2020-04-23 2023-07-07 华为技术有限公司 优化决策规控的方法、控制车辆行驶的方法和相关装置
CN111797780B (zh) * 2020-07-08 2022-12-02 中国第一汽车股份有限公司 一种跟车轨迹规划方法、装置、服务器及存储介质
CN111899625A (zh) * 2020-07-16 2020-11-06 北京理工大学 智能辅助驾驶开发装置
CN112896185A (zh) * 2021-01-25 2021-06-04 北京理工大学 一种面向车路协同的智能驾驶行为决策规划方法及***
CN113085853B (zh) * 2021-04-26 2022-05-17 中汽研(天津)汽车工程研究院有限公司 一种车道内主动躲闪大型车辆的辅助驾驶***
CN113377112B (zh) * 2021-06-30 2023-05-05 东风商用车有限公司 一种自动驾驶速度规划及状态协调方法和装置
CN113386795B (zh) * 2021-07-05 2022-07-01 西安电子科技大学芜湖研究院 一种自动驾驶车辆智能决策及局部轨迹规划方法及其决策***
CN113799799A (zh) * 2021-09-30 2021-12-17 中国第一汽车股份有限公司 一种安全补偿方法、装置、存储介质及电子设备
CN114019978A (zh) * 2021-11-08 2022-02-08 陕西欧卡电子智能科技有限公司 一种无人驾驶游船及无人驾驶方法
CN114407902B (zh) * 2022-01-19 2023-11-28 浙江大学 一种基于道路水层深度估计的驾驶决策的***
CN115743178A (zh) * 2022-11-25 2023-03-07 中国科学院深圳先进技术研究院 一种基于场景自适应识别的自动驾驶方法及***
CN116499486B (zh) * 2023-06-26 2023-09-22 北京理工大学 一种复杂越野环境路径规划方法、***及电子设备

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106094812A (zh) * 2015-04-27 2016-11-09 通用汽车环球科技运作有限责任公司 用于自主驾驶的反应式路径规划
CN106926844A (zh) * 2017-03-27 2017-07-07 西南交通大学 一种基于实时环境信息的动态自动驾驶换道轨迹规划方法
US20180218606A1 (en) * 2017-02-01 2018-08-02 Toyota Research Institute, Inc. Systems and methods for operating a vehicle on a roadway
CN108387242A (zh) * 2018-02-07 2018-08-10 西南交通大学 自动驾驶换道准备和执行一体化轨迹规划方法
CN108460980A (zh) * 2018-05-11 2018-08-28 西南交通大学 自动驾驶车辆下匝道最优意图生成点计算方法
CN108762271A (zh) * 2018-06-04 2018-11-06 北京智行者科技有限公司 车辆控制单元
CN109669461A (zh) * 2019-01-08 2019-04-23 南京航空航天大学 一种复杂工况下自动驾驶车辆决策***及其轨迹规划方法

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101620911B1 (ko) * 2014-09-25 2016-05-13 국방과학연구소 주행정보지도 기반 무인차량 및 이를 이용한 지역경로계획 관리 방법
CN206691107U (zh) * 2017-03-08 2017-12-01 深圳市速腾聚创科技有限公司 无人驾驶汽车***及汽车
CN107195020A (zh) * 2017-05-25 2017-09-22 清华大学 一种面向列车自动驾驶模式学习的列车运行记录数据处理方法
CN107315411B (zh) * 2017-07-04 2020-04-21 合肥工业大学 一种基于车车协同下无人驾驶车辆的换道轨迹规划方法
CN107797534B (zh) * 2017-09-30 2019-05-28 安徽江淮汽车集团股份有限公司 一种纯电动自动驾驶***
CN107797559A (zh) * 2017-11-24 2018-03-13 南京视莱尔汽车电子有限公司 一种智能驾驶测试数据远程监测***
CN108225364B (zh) * 2018-01-04 2021-07-06 吉林大学 一种无人驾驶汽车驾驶任务决策***及方法
CN108919795B (zh) * 2018-06-01 2023-07-25 中国北方车辆研究所 一种自动驾驶汽车换道决策方法及装置

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106094812A (zh) * 2015-04-27 2016-11-09 通用汽车环球科技运作有限责任公司 用于自主驾驶的反应式路径规划
US20180218606A1 (en) * 2017-02-01 2018-08-02 Toyota Research Institute, Inc. Systems and methods for operating a vehicle on a roadway
CN106926844A (zh) * 2017-03-27 2017-07-07 西南交通大学 一种基于实时环境信息的动态自动驾驶换道轨迹规划方法
CN108387242A (zh) * 2018-02-07 2018-08-10 西南交通大学 自动驾驶换道准备和执行一体化轨迹规划方法
CN108460980A (zh) * 2018-05-11 2018-08-28 西南交通大学 自动驾驶车辆下匝道最优意图生成点计算方法
CN108762271A (zh) * 2018-06-04 2018-11-06 北京智行者科技有限公司 车辆控制单元
CN109669461A (zh) * 2019-01-08 2019-04-23 南京航空航天大学 一种复杂工况下自动驾驶车辆决策***及其轨迹规划方法

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111968397A (zh) * 2020-07-31 2020-11-20 东风商用车有限公司 自动驾驶行为判定方法
CN112201070A (zh) * 2020-09-29 2021-01-08 上海交通大学 基于深度学习的自动驾车高速公路瓶颈路段行为决策方法
CN112201070B (zh) * 2020-09-29 2022-03-01 上海交通大学 基于深度学习的自动驾车高速公路瓶颈路段行为决策方法

Also Published As

Publication number Publication date
CN109669461A (zh) 2019-04-23
CN109669461B (zh) 2020-07-28

Similar Documents

Publication Publication Date Title
WO2020143288A1 (zh) 一种复杂工况下自动驾驶车辆决策***及其轨迹规划方法
CN113386795B (zh) 一种自动驾驶车辆智能决策及局部轨迹规划方法及其决策***
CN110597245B (zh) 基于二次型规划和神经网络的自动驾驶换道轨迹规划方法
CN110703754B (zh) 一种自动驾驶车辆路径与速度高度耦合的轨迹规划方法
CN111338340B (zh) 基于模型预测的无人驾驶汽车局部路径规划方法
CN110187639B (zh) 一种基于参数决策框架的轨迹规划控制方法
CN109855639B (zh) 基于障碍物预测与mpc算法的无人驾驶轨迹规划方法
CN109032131B (zh) 一种应用于无人驾驶汽车的动态超车避障方法
CN109727469B (zh) 一种多车道下自动驾驶车辆综合危险度评估方法
US10240933B2 (en) Method for guiding a vehicle and a driver assistance system
WO2020029462A1 (zh) 一种电动汽车的无人驾驶***
WO2021102957A1 (zh) 一种车道保持方法、车载设备和存储介质
CN111806467A (zh) 一种基于车辆行驶规律的变速动态换道轨迹规划方法
CN111409641A (zh) 一种基于纵横向协调的轨迹跟踪控制方法及***
CN106681327A (zh) 一种大惯性电动客车的智能驾驶横纵向解耦控制方法及***
WO2019167518A1 (ja) 車両走行制御装置
CN111681452A (zh) 一种基于Frenet坐标系下的无人驾驶汽车动态换道轨迹规划方法
CN109164814A (zh) 面向高速公路场景的自动驾驶控制***
CN113467474A (zh) 自动驾驶层次化控制***
CN212828326U (zh) 一种基于纵横向协调的轨迹跟踪控制***
Lin et al. Time-to-collision-aware lane-change strategy based on potential field and cubic polynomial for autonomous vehicles
CN115140048A (zh) 一种自动驾驶行为决策与轨迹规划模型与方法
CN117279816A (zh) 用于控制自主车辆的方法
CN114212107A (zh) 一种基于价值图的智能网联驾驶规划方法
CN116300970B (zh) 车辆自主编队方法及装置

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: 19909496

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19909496

Country of ref document: EP

Kind code of ref document: A1