CN111599179B - No-signal intersection vehicle motion planning method based on risk dynamic balance - Google Patents

No-signal intersection vehicle motion planning method based on risk dynamic balance Download PDF

Info

Publication number
CN111599179B
CN111599179B CN202010438570.4A CN202010438570A CN111599179B CN 111599179 B CN111599179 B CN 111599179B CN 202010438570 A CN202010438570 A CN 202010438570A CN 111599179 B CN111599179 B CN 111599179B
Authority
CN
China
Prior art keywords
vehicle
intersection
risk
track
automatic driving
Prior art date
Legal status (The legal status 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 status listed.)
Active
Application number
CN202010438570.4A
Other languages
Chinese (zh)
Other versions
CN111599179A (en
Inventor
鲁光泉
谭海天
陈发城
丁川
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beihang University
Original Assignee
Beihang University
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 Beihang University filed Critical Beihang University
Priority to CN202010438570.4A priority Critical patent/CN111599179B/en
Publication of CN111599179A publication Critical patent/CN111599179A/en
Application granted granted Critical
Publication of CN111599179B publication Critical patent/CN111599179B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/01—Detecting movement of traffic to be counted or controlled
    • G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
    • G08G1/0125—Traffic data processing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/26—Government or public services
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/01—Detecting movement of traffic to be counted or controlled
    • G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
    • G08G1/0137—Measuring and analyzing of parameters relative to traffic conditions for specific applications
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/09—Arrangements for giving variable traffic instructions
    • G08G1/0962—Arrangements for giving variable traffic instructions having an indicator mounted inside the vehicle, e.g. giving voice messages
    • G08G1/0967—Systems involving transmission of highway information, e.g. weather, speed limits
    • G08G1/096708—Systems involving transmission of highway information, e.g. weather, speed limits where the received information might be used to generate an automatic action on the vehicle control
    • G08G1/096725—Systems involving transmission of highway information, e.g. weather, speed limits where the received information might be used to generate an automatic action on the vehicle control where the received information generates an automatic action on the vehicle control

Landscapes

  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Tourism & Hospitality (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Chemical & Material Sciences (AREA)
  • Marketing (AREA)
  • Theoretical Computer Science (AREA)
  • Analytical Chemistry (AREA)
  • General Business, Economics & Management (AREA)
  • Development Economics (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Health & Medical Sciences (AREA)
  • Educational Administration (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Atmospheric Sciences (AREA)
  • Traffic Control Systems (AREA)
  • Control Of Driving Devices And Active Controlling Of Vehicle (AREA)

Abstract

The invention develops a no-signal intersection vehicle motion planning method based on risk dynamic balance. The method comprises the following steps: (1) completing the prediction of the running tracks of other moving vehicles in the no-signal intersection on the basis of a no-signal intersection vehicle track prediction model; (2) constructing a dynamic risk field which changes at any time in the non-signalized intersection through a non-signalized intersection risk field model; (3) obtaining the expected vehicle track distribution based on the expected vehicle track distribution model at the signalless intersection; (4) combining the dynamic risk field with the expected track distribution of the vehicle to determine risk values corresponding to different expected tracks; (5) obtaining an acceptable risk level of the autonomous vehicle according to the model of the acceptable risk level of the autonomous vehicle; (6) comparing the risk values corresponding to different expected tracks with acceptable risk levels to obtain acceptable risk track distribution of the automatic driving vehicle; (7) and selecting the track with the highest comprehensive benefit from the acceptable risk track distribution through a comprehensive benefit function.

Description

No-signal intersection vehicle motion planning method based on risk dynamic balance
Technical Field
The invention relates to the field of traffic safety, in particular to a no-signal intersection vehicle motion planning method based on risk dynamic balance.
Background
With the rapid development of artificial intelligence, wireless communication, informatization and other technologies, the carrying tool enters an intelligent era. Among a plurality of core technical modules for automatic driving, a motion planning module is a key factor for reflecting the intelligent level of an automatic driving vehicle, directly determines the specific behavior of the automatic driving vehicle, and modeling the driving safety risk is the basis of the automatic decision and motion planning of the automatic driving vehicle, and directly influences the safety and efficiency of a motion planning path.
A great deal of research is carried out on traffic safety risk modeling at home and abroad. The method mainly uses the time distance and the space distance to describe the relative motion relation between vehicles and evaluate the transverse and longitudinal driving risks. The typical indicators of the time distance include TTC (time to precision), TH (time headway) and derivatives thereof. Charly et al uses an improved driving risk indicator based on TTC to identify possible conflicts between different types of vehicles. Ali et al quantify the safety in the lane change process using minimum gap time in an internet environment. Under the condition of vehicle-vehicle information interaction, Lu Guangquan and other people design a derived index safety margin, and the risk degree subjectively sensed by a driver in the following process is quantized. Ryan et al propose a driving risk assessment method for an autonomous vehicle based on driving behavior and critical safety events, and identify behavior hotspots by using natural driving data, and identify similar risk groups by a self-organizing map. Li ya Yong et al consider two vehicles in front and back as two potential field centers, consider the position and speed of the front and back vehicles, evaluate the longitudinal safety of the vehicles under crowded conditions. Li and the like are adapted to different road line shapes, an elliptical traffic safety field is introduced based on a potential field theory, and the traffic risks under different scenes are evaluated by considering the influence difference of the tangential direction and the normal direction of the lane line on the traffic risks.
However, the above-mentioned research is mostly directed to a single target such as vehicle safety control and road safety assessment, and fails to comprehensively analyze the influence of different elements in the human-vehicle-road system on the vehicle motion planning path risk from the perspective of the human-vehicle-road system. To solve this problem, field theory is increasingly applied to the description of risks in the human-vehicle-road system at home and abroad. Wolf et al established a potential energy field between the own vehicle and another vehicle by using the Tochu potential and electric field thought, and studied the anti-collision path planning in typical scenes such as following and overtaking, but the research result can only be applied to a partially simplified specific scene. The Wangjiaqiang et al put forward a driving safety field theory considering human-vehicle-road factors, describe the driving risk of vehicles by using physical quantities such as field intensity, field force and potential energy, and establish a safety field combined model based on safety potential energy and time change rate thereof, but the safety field combined model can only provide assistance for driving safety and cannot provide accurate basis for safety assessment in complex environment, and the model excessively pursues absolute safety without considering the acceptable risk level of automatically driving vehicles, resulting in imbalance of driving safety and efficiency.
At present, existing research focuses on traffic element description and driving safety risk evaluation in a simple scene, a signalless intersection is taken as a typical complex traffic scene, and existing research results are difficult to adapt to the environment of the signalless intersection. In addition, the existing signalless intersection automatic driving motion planning considers less problems of bearable capability of automatic driving vehicles to risks and balance of risks and efficiency in a motion planning path. Therefore, how to implement hierarchical modeling of a no-signal intersection scene by using dynamic motion data, how to comprehensively evaluate the influence of a dynamic traffic environment in the no-signal intersection on driving safety under a person-vehicle-road coordination condition, and how to consider the balance between the safety and efficiency of a motion planning path are all problems to be solved urgently.
Disclosure of Invention
Aiming at the problems, the invention aims to provide a no-signal intersection vehicle motion planning method based on dynamic risk balance and capable of providing no-signal intersection environment downlink vehicle risk identification and track distribution selection for an automatic driving system
In order to achieve the above object, the present invention provides a no-signalized intersection vehicle motion planning method based on risk dynamic balance, which comprises the following steps:
step 1, acquiring static environment element information and other moving vehicle state data in a no-signal intersection through an automatic driving vehicle sensor and an internet of vehicles, and finishing the prediction of the driving tracks of other vehicles in the no-signal intersection by combining a no-signal intersection vehicle track prediction model;
step 2, constructing a dynamic risk field changing at any time in the signalless intersection by combining the static environment element information in the signalless intersection and the running track data of other moving vehicles obtained in the step 1 based on the signalless intersection risk field model;
step 3, based on the no-signal intersection vehicle expected track distribution model, combining the vehicle expectation and the static environment element information in the no-signal intersection obtained in the step 1 to obtain the expected track distribution of the automatic driving vehicle in the no-signal intersection;
step 4, substituting the expected track distribution of the automatic driving vehicle in the no-signal intersection obtained in the step 3 into the dynamic risk field obtained in the step 2, and determining risk values corresponding to different expected tracks in the expected track distribution;
step 5, considering the delay of the information acquired by the automatic driving vehicle, the error of the automatic driving vehicle on the sensing of the surrounding environment and the influence on the prediction error of the other moving vehicle tracks based on the vehicle acceptable risk level model, and determining the acceptable risk level of the automatic driving vehicle;
step 6, comparing the risk values corresponding to the different expected tracks obtained in the step 4 with the acceptable risk level of the automatic driving vehicle obtained in the step 5, and screening to obtain acceptable risk track distribution of the automatic driving vehicle;
and 7, determining the comprehensive benefits of the trajectories with different distributions of acceptable risk trajectories of the automatically-driven vehicles obtained in the step 6 through a comprehensive benefit function, and selecting one trajectory with the highest comprehensive benefit as the motion trajectory planning of the automatic driving of the signalless intersection based on the risk dynamic balance theory.
Further, the method for obtaining the "travel track prediction" in step 1 includes the following steps:
step 11, acquiring required basic data including the shape and size of the signalless intersection and moving objects in the intersection at the initial t through the sensor of the automatic driving vehicle and the Internet of vehicles0Transverse and longitudinal position of time (x (t)0),y(t0) Transverse and longitudinal velocity (v)x(t0),vy(t0) Transverse and longitudinal acceleration (a)x(t0),ay(t0) And yaw angle ψ (t)0)。
Step 12, establishing a vehicle track prediction model at the no-signal intersection as follows:
SP(xp,yp,vp,x,vp,y,ap,x,ap,y,ψp,t)＝fP(typev,x(t0),y(t0),vx(t0),vy(t0),ax(t0),ay(t0),ψ(t0))
in the formula, typevRefers to predicting the type of vehicle, (x (t)0),y(t0) ) means an initial t0(v) predicting the lateral and longitudinal position of the vehicle at that time, (v)x(t0),vy(t0) ) means an initial t0Predicting the transverse and longitudinal speed of the vehicle at the moment (a)x(t0),ay(t0) ) means an initial t0Predicting the transverse and longitudinal acceleration of the vehicle at a time, psi (t)0) Means the initial t0The moment predicts the yaw angle of the vehicle.
Step 13, combining the obtained basic data with the prediction model of the vehicle track of the signalless intersection to obtain the running tracks S of all other moving vehicles in the signalless intersectionP。
Further, the method for obtaining the "dynamic risk field" in step 2 includes the following steps:
step 21, considering the constraint form, the constraint strength and the constraint range of the static traffic environment elements in the signalless intersection to the automatic driving vehicle, so as to determine a risk field model around the static traffic environment elements in the signalless intersection:
Re(xe,ye,t)＝fe(typee,lengthe,widthe,heighte)
in the formula, typeeMeans the type, length, of the static environment elementeIs the length, width, of the static environment elementeRefers to the width, height, of the static environment elementeRefers to the height of the static environmental element;
step 22, considering the constraint form, the constraint strength and the constraint range of other vehicles in the signalless intersection to the automatic driving vehicle, so as to determine a risk field model around the moving object in the signalless intersection:
Rv(xv,yv,t)＝fv(lengthv,widthv,x(t),y(t),vx(t),vy(t),ax(t),ay(t),ψ(t))
in the formula, lengthvIs the length, width, of the vehiclevRefers to the width of the vehicle, (x (t), y (t)) refers to the spatial position of the vehicle over time, (v)x(t),vy(t)) means the speed of movement of the vehicle over time, (a)x(t),ay(t)) means the acceleration of the vehicle over time, ψ (t) means the yaw angle of the vehicle over time;
step 23, transferring the risk field model obtained in the above steps to the same coordinate system according to the following coordinate system change formula:
Figure GDA0003159519800000031
in the formula (x)i,yi) The coordinate position of a coordinate point in the ith individual element risk field model coordinate system is defined, (x, y) is the coordinate position of the coordinate point in the final unified xoy coordinate system, and (delta x (i), (delta y (i)), alpha (i)) is the position and the rotation angle of the ith individual element risk field model coordinate system relative to the final unified xoy coordinate system;
and 24, superposing all the individual element risk field models to obtain a dynamic risk field in the signal-free intersection:
R0(x,y,t)＝max(Re,1(x,y,t),…,Re,n1(x,y,t),Rv,1(x,y,t),…,Rv,n2(x,y,t))
in the formula, Re,1(x, y, t) is the 1 st static environment element applying a risk value to the coordinate point (x, y) at time t, Rv,1(x, y, t) is the 1 st vehicle applying a risk value to the coordinate point (x, y) at time t, n1Is the number of static environment elements, n, in the signalless intersection2The number of vehicles traveling in the no-signal intersection;
further, the method for obtaining the "desired trajectory distribution" in step 3 includes the following steps:
step 31, determining the driving direction of the automatic driving vehicle at the no-signal intersection, wherein the driving direction is straight, left-turning or right-turning;
step 32, establishing an expected track distribution model of the automatically-driven vehicles at the signalless intersection as follows:
the expected track distribution model of the automatic driving vehicle straight running in the signalless intersection is as follows:
TS(x,y,vx,vy,ax,ay,ψ,t)＝fS(Si,Ae,Al,x(t0),y(t0),vx(t0),vy(t0),ax(t0),ay(t0),ψ(t0))
the distribution model of the expected track of the left turn of the automatic driving vehicle in the signalless intersection comprises the following steps:
TL(x,y,vx,vy,ax,ay,ψ,t)＝fL(Si,Ae,Al,x(t0),y(t0),vx(t0),vy(t0),ax(t0),ay(t0),ψ(t0))
the distribution model of the expected track of the automatic driving vehicle turning right at the signalless intersection comprises the following steps:
TR(x,y,vx,vy,ax,ay,ψ,t)＝fR(Si,Ae,Al,x(t0),y(t0),vx(t0),vy(t0),ax(t0),ay(t0),ψ(t0))
in the formula, SiIn the shape of the intersection, AeMeans the coordinate range of the entering lane of the vehicle entering the intersection, AlThe coordinate range of the possible departure lane of the vehicle from the intersection is defined, (x (t)0),y(t0) ) means an initial t0The transverse and longitudinal positions of the vehicle at the moment of time, (v)x(t0),vy(t0) ) means an initial t0The transverse and longitudinal speed of the vehicle (a)x(t0),ay(t0) ) means an initial t0Moment transverse and longitudinal acceleration of the vehicle, psi (t)0) Means the initial t0The yaw angle of the vehicle at that moment.
Step 33, the obtained current state data of the vehicle and the no-signal exchange are carried outCombining the static environment element information of the fork with the expected track distribution model to obtain the expected track distribution T of the automatic driving vehicle at the signalless intersection0(x,y,vx,vy,ax,ay,ψ,t)。
Further, the method for obtaining the "risk values corresponding to different expected trajectories" in step 4 includes the following steps:
desired trajectory profile T at the signalless intersection for the autonomous vehicle0(x,y,vx,vy,ax,ayψ, t) of a track Si(x (t), y (t)), calculating the risk R of the trajectory over timei(t)＝R0(Si(x (t), y (t)) and t), and further obtaining the risk size corresponding to all different expected track distributions in the whole expected track distribution.
Further, the method of determining the "acceptable risk level of autonomous vehicle" in step 5 includes the steps of:
RA＝fA(uC,uP,uT)
in the formula uCIs the delay of obtaining information from an autonomous vehicle, uPRefers to the possible error, u, of the perception of the surrounding environment by the autonomous vehicleTRefers to the possible errors in predicting the trajectories of other moving vehicles.
Further, the method for obtaining the acceptable risk trajectory distribution in step 6 includes the following steps:
when the track S isi(x (t), y (t)) corresponding risk R over timei(t) is less than the acceptable risk level R for the autonomous vehicle calculated from step 5AI.e. Ri(t)≤RAThen the trace distribution belongs to the acceptable risk trace distribution Taccept(x,y,vx,vy,ax,ayψ, T) to obtain an acceptable risk trajectory distribution Taccept(x,y,vx,vy,ax,ay,ψ,t)。
Further, the method for obtaining the "motion trail plan" in step 7 includes the following steps:
step 71, calculating the comprehensive benefits of all the distribution tracks belonging to the acceptable risk tracks:
Figure GDA0003159519800000051
i is 0,1, …, n, wherein Si,accept(x(t),y(t))∈Taccept(x,y,vx,vy,ax,ay,ψ,t)
Step 72, comparing all the distributions T belonging to the acceptable risk trajectoryaccept(x,y,vx,vy,ax,ayψ, t) of the track Si(x (t), y (t)) corresponding general profit GiSelecting the track in which the maximum comprehensive profit can be obtained as the motion track plan of the automatic driving vehicle in the signal-free intersection:
Figure GDA0003159519800000052
its comprehensive benefits
Figure GDA0003159519800000053
Then pair
Figure GDA0003159519800000054
Its comprehensive benefits
Figure GDA0003159519800000055
Has Gj≥GiIn the formula, g (S)i,accept(x (t), y (t))) and g (S)j,accept(x (t), y (t))) is a Lagrangian function of the desired trajectory, GiAnd GjRespectively acceptable risk trajectory distribution Taccept(x,y,vx,vy,ax,ayThe comprehensive income corresponding to the ith and jth tracks in psi, t) is obtained as the jth track Sj,accept(x (t), y (t)) for planning the motion track of the autonomous vehicle in the signal-free intersection.
Drawings
FIG. 1 is an intersection environment according to an embodiment of the present disclosure;
FIG. 2 is a travel path of another moving vehicle according to an embodiment of the present disclosure;
FIG. 3a is a schematic view of a risk field around a static environmental element according to an embodiment of the present disclosure;
FIG. 3b is a schematic view of a risk field around a moving vehicle according to an embodiment of the present disclosure;
FIG. 3c is a schematic view of a signalless intersection risk scenario according to an embodiment of the present disclosure;
FIG. 4 is a desired trajectory profile for a straight-ahead driving of an autonomous vehicle according to an embodiment of the present disclosure;
FIG. 5 is a graph illustrating risk values over time for different expected trajectory profiles, according to an embodiment of the present disclosure;
FIG. 6 is a graph of an acceptable risk trajectory profile for an autonomous vehicle according to an embodiment of the present disclosure;
fig. 7 is a path plan for risk-based dynamic balancing for autonomous vehicles according to embodiments of the present disclosure.
Fig. 8 is an overall logic block diagram according to an embodiment of the present disclosure.
Detailed Description
The invention is described in detail below with reference to the figures and examples.
The no-signalized intersection vehicle motion planning method based on risk dynamic balance provided by the embodiment comprises the following steps:
step 1, acquiring static environment element information and other moving vehicle state data in a no-signal intersection through an automatic driving vehicle sensor and an internet of vehicles, and combining a no-signal intersection vehicle track prediction model to complete the prediction of the running tracks of other vehicles at the no-signal intersection. In this embodiment, as shown in fig. 1, static environment element information such as the no-signal intersection marking line and motion information of two other vehicles at the initial time are acquired, and then the prediction of the traveling tracks of the two vehicles is completed, and the traveling tracks of the two vehicles are predicted as shown in fig. 2.
And 2, constructing a dynamic risk field which changes along with time and space in the signalless intersection based on the signalless intersection risk field model by combining the static environment element information in the signalless intersection and the other moving vehicle driving track data obtained in the step 1. In the invention, the risk fields around static environment elements such as a solid line are shown in fig. 3a, the risk fields around moving vehicles are shown in fig. 3b, and the risk fields around all the elements in the intersection are superposed to obtain a dynamic risk field which changes along with space and time in the intersection without signals, as shown in fig. 3 c.
And 3, obtaining expected track distribution of the automatically driven vehicle in the non-signal intersection based on the non-signal intersection vehicle expected track distribution model by combining the vehicle expectation and the static environment element information in the non-signal intersection obtained in the step 1. In the present embodiment, the traveling direction of the autonomous vehicle is straight. Therefore, the straight expected trajectory distribution of the autonomous vehicle at the no-signal intersection is as shown in fig. 4.
And 4, substituting the expected track distribution of the automatic driving vehicle in the no-signal intersection obtained in the step 3 into the dynamic risk field obtained in the step 2, and determining risk values corresponding to different expected tracks in the expected track distribution. In the present embodiment, the risk values with time change corresponding to all the expected tracks shown in fig. 4 are referred to, as shown in fig. 5.
And 5, determining the acceptable risk level of the automatic driving vehicle based on the acceptable risk level model of the vehicle, considering the delay of the information acquired by the automatic driving vehicle, the error of the automatic driving vehicle on the perception of the surrounding environment and the influence on the prediction error of the track of other moving vehicles. In this embodiment, the delay of the automatic driving vehicle in the no-signal intersection through the self sensor and the internet of vehicles to acquire the information in the no-signal intersection, the error possibly existing in the acquired data, and the error possibly existing in the running tracks of the other two vehicles in the intersection have a great influence on the acceptable risk level of the vehicle, and when the delay and the error are great, the acceptable risk level of the vehicle is low.
And 6, comparing the risk values corresponding to the different expected tracks obtained in the step 4 with the acceptable risk level of the automatic driving vehicle obtained in the step 5, and screening to obtain the acceptable risk track distribution of the automatic driving vehicle. In the present embodiment, the distribution of acceptable risk trajectories for the autonomous driving straight ahead obtained by the screening is shown in fig. 6.
And 7, determining the comprehensive benefits of the trajectories with different distributions of acceptable risk trajectories of the automatically-driven vehicles obtained in the step 6 through a comprehensive benefit function, and selecting one trajectory with the highest comprehensive benefit as the motion trajectory planning of the automatic driving of the signalless intersection based on the risk dynamic balance theory. In this embodiment, the motion profile as shown in fig. 7 has a risk that is less than an acceptable level of risk for the autonomous vehicle, while the combined benefit is higher than for other profiles that fall within the acceptable risk profile.
In one embodiment, the method for obtaining the "travel track prediction" in step 1 includes the following steps:
step 11, obtaining basic data required for track prediction in the intersection shown in fig. 1 through the sensor of the automatic driving vehicle and the internet of vehicles, wherein the basic data comprise the shape and the size of the signalless intersection and two vehicles in the intersection at the initial t0Transverse and longitudinal position of time (x (t)0),y(t0) Transverse and longitudinal velocity (v)x(t0),vy(t0) Transverse and longitudinal acceleration (a)x(t0),ay(t0) And yaw angle ψ (t)0)。
Step 12, establishing a vehicle track prediction model at the no-signal intersection as follows:
SP(xp,yp,vp,x,vp,y,ap,x,ap,y,ψp,t)＝fP(typev,x(t0),y(t0),vx(t0),vy(t0),ax(t0),ay(t0),ψ(t0))
in the formula, typevIt is meant to predict the type of vehicle,(x(t0),y(t0) ) means an initial t0(v) predicting the lateral and longitudinal position of the vehicle at that time, (v)x(t0),vy(t0) ) means an initial t0Predicting the transverse and longitudinal speed of the vehicle at the moment (a)x(t0),ay(t0) ) means an initial t0Predicting the transverse and longitudinal acceleration of the vehicle at a time, psi (t)0) Means the initial t0The moment predicts the yaw angle of the vehicle.
Step 13, combining the obtained basic data with the prediction model of the vehicle track of the signalless intersection to obtain the running tracks S of the two vehicles in the signalless intersectionP,1And SP,2As shown in fig. 2.
In one embodiment, the method for obtaining the "dynamic risk field" in step 2 comprises the following steps:
step 21, considering the constraint form, the constraint strength and the constraint range of the static traffic environment element in the signalless intersection to the autonomous vehicle, so as to determine a risk field model around the static traffic environment element in the signalless intersection, as shown in fig. 3 a:
Re(xe,ye,t)＝fe(typee,lengthe,widthe,heighte)
in the formula, typeeMeans the type, length, of the static environment elementeIs the length, width, of the static environment elementeRefers to the width, height, of the static environment elementeRefers to the height of the static environmental element;
step 22, considering the constraint form, the constraint strength and the constraint range of other vehicles in the signalless intersection to the automatic driving vehicle, so as to determine a risk field model around the moving object in the signalless intersection, as shown in fig. 3 c:
Rv(xv,yv,t)＝fv(lengthv,widthv,x(t),y(t),vx(t),vy(t),ax(t),ay(t),ψ(t))
in the formula, lengthvRefers to the length of the vehicle，widthvRefers to the width of the vehicle, (x (t), y (t)) refers to the spatial position of the vehicle over time, (v)x(t),vy(t)) means the speed of movement of the vehicle over time, (a)x(t),ay(t)) means the acceleration of the vehicle over time, ψ (t) means the yaw angle of the vehicle over time;
step 23, transferring the risk field model obtained in the above steps to the same coordinate system according to the following coordinate system change formula:
Figure GDA0003159519800000071
in the formula (x)i,yi) The coordinate position of a coordinate point in the ith individual element risk field model coordinate system is defined, (x, y) is the coordinate position of the coordinate point in the final unified xoy coordinate system, and (delta x (i), (delta y (i)), alpha (i)) is the position and the rotation angle of the ith individual element risk field model coordinate system relative to the final unified xoy coordinate system;
step 24, superposing all the individual element risk field models to obtain a dynamic risk field in the signal-free intersection, as shown in fig. 3 c:
R0(x,y,t)＝max(Re,1(x,y,t),…,Re,n1(x,y,t),Rv,1(x,y,t),Rv,2(x,y,t))
in the formula, Re,1(x, y, t) is the 1 st static environment element applying a risk value to the coordinate point (x, y) at time t, Rv,1(x, y, t) is the 1 st vehicle applying a risk value to coordinate point (x, y) at time t, Rv,2(x, y, t) is the 2 nd vehicle applying a risk value to the coordinate point (x, y) at time t, n1The number of static environment elements in the signalless intersection;
in one embodiment, the method for obtaining the "desired trajectory distribution" in step 3 includes the following steps:
and step 31, determining the driving direction of the automatic driving vehicle at the no-signal intersection, wherein the driving direction is straight, left-turning or right-turning. In this embodiment, the direction of travel of the autonomous vehicle at the no-signal intersection is straight.
Step 32, establishing an expected track distribution model of the automatically-driven vehicles at the signalless intersection as follows:
the expected track distribution model of the automatic driving vehicle straight running in the signalless intersection is as follows:
TS(x,y,vx,vy,ax,ay,ψ,t)＝fS(Si,Ae,Al,x(t0),y(t0),vx(t0),vy(t0),ax(t0),ay(t0),ψ(t0))
in the formula, SiIn the shape of the intersection, AeMeans the coordinate range of the entering lane of the vehicle entering the intersection, AlThe coordinate range of the possible departure lane of the vehicle from the intersection is defined, (x (t)0),y(t0) ) means an initial t0The transverse and longitudinal positions of the vehicle at the moment of time, (v)x(t0),vy(t0) ) means an initial t0The transverse and longitudinal speed of the vehicle (a)x(t0),ay(t0) ) means an initial t0Moment transverse and longitudinal acceleration of the vehicle, psi (t)0) Means the initial t0The yaw angle of the vehicle at that moment.
Step 33, combining the obtained current state data of the self-vehicle and the static environment element information of the no-signal intersection with the straight expected track distribution model of the no-signal intersection to obtain the expected track distribution T of the automatic driven vehicle at the no-signal intersection0(x,y,vx,vy,ax,ayψ, t) as shown in fig. 4.
In one embodiment, the method for obtaining the "risk values corresponding to different expected trajectories" in step 4 includes the following steps:
desired trajectory profile T at the signalless intersection for the autonomous vehicle0(x,y,vx,vy,ax,ayψ, t) of a track Si(x (t), y (t)) calculating the trajectory at any timeRisk of inter-variation Ri(t)＝R0(Si(x (t), y (t)), and t), and then obtaining the risk size corresponding to all different expected trajectory distributions in the whole expected trajectory distribution, as shown in fig. 5.
In one embodiment, the method of determining the "acceptable risk level for autonomous vehicle" in step 5 comprises the steps of:
RA＝fA(uC,uP,uT)
in the formula uCIs the delay of obtaining information from an autonomous vehicle, uPRefers to the possible error, u, of the perception of the surrounding environment by the autonomous vehicleTRefers to the possible errors in predicting the trajectories of other moving vehicles.
In one embodiment, the method for obtaining the acceptable risk trajectory distribution in step 6 includes the following steps:
when the track S isi(x (t), y (t)) corresponding risk R over timei(t) is less than the acceptable risk level R for the autonomous vehicle calculated from step 5AI.e. Ri(t)≤RAThen the trace distribution belongs to the acceptable risk trace distribution Taccept(x,y,vx,vy,ax,ayψ, T) to obtain an acceptable risk trajectory distribution Taccept(x,y,vx,vy,ax,ayψ, t) as shown in fig. 6.
Further, the method for obtaining the "motion trail plan" in step 7 includes the following steps:
step 71, calculating the comprehensive benefits of all the distribution tracks belonging to the acceptable risk tracks:
Figure GDA0003159519800000081
i is 0,1, …, n, wherein Si,accept(x(t),y(t))∈Taccept(x,y,vx,vy,ax,ay,ψ,t)
Step 72, comparing all the trajectories belonging to the acceptable riskDistribution Taccept(x,y,vx,vy,ax,ayψ, t) of the track Si(x (t), y (t)) corresponding general profit GiSelecting the track in which the maximum comprehensive profit can be obtained as the motion track plan of the automatic driving vehicle in the signal-free intersection:
Figure GDA0003159519800000091
its comprehensive benefits
Figure GDA0003159519800000092
Then pair
Figure GDA0003159519800000093
Its comprehensive benefits
Figure GDA0003159519800000094
Has Gj≥GiIn the formula, g (S)i,accept(x (t), y (t))) and g (S)j,accept(x (t), y (t))) is a Lagrangian function of the desired trajectory, GiAnd GjRespectively acceptable risk trajectory distribution Taccept(x,y,vx,vy,ax,ayThe comprehensive income corresponding to the ith and jth tracks in psi, t) is obtained as the jth track Sj,accept(x (t), y (t)) planning the motion trail of the automatic driving vehicle in the signalless intersection, and finally selecting the motion trail of the automatic driving signalless intersection based on the dynamic risk balance as shown in fig. 7.
The above embodiments are only for illustrating the present invention, and the steps of the method and the like can be changed, and all equivalent changes and modifications based on the technical scheme of the present invention should not be excluded from the protection scope of the present invention.

Claims (8)

1. A no-signal intersection vehicle motion planning method based on risk dynamic balance is characterized by comprising the following steps:
step 1, acquiring static environment element information and other moving vehicle state data in a no-signal intersection through an automatic driving vehicle sensor and an internet of vehicles, and finishing the prediction of the driving tracks of other vehicles in the no-signal intersection by combining a no-signal intersection vehicle track prediction model;
step 2, on the basis of a signal-free intersection risk field model, combining the static environment element information in the signal-free intersection obtained in the step 1 and the running track data of other moving vehicles to construct a dynamic risk field which changes along with time and space in the signal-free intersection;
step 3, based on the no-signal intersection vehicle expected track distribution model, combining the vehicle expectation and the static environment element information in the no-signal intersection obtained in the step 1 to obtain the expected track distribution of the automatic driving vehicle in the no-signal intersection;
step 4, substituting the expected track distribution of the automatic driving vehicle in the no-signal intersection obtained in the step 3 into the dynamic risk field obtained in the step 2, and determining risk values corresponding to different expected tracks in the expected track distribution;
step 5, considering the delay of the information acquired by the automatic driving vehicle, the error of the automatic driving vehicle on the sensing of the surrounding environment and the influence on the prediction error of the other moving vehicle tracks based on the vehicle acceptable risk level model, and determining the acceptable risk level of the automatic driving vehicle;
step 6, comparing the risk values corresponding to the different expected tracks obtained in the step 4 with the acceptable risk level of the automatic driving vehicle obtained in the step 5, and screening to obtain acceptable risk track distribution of the automatic driving vehicle;
and 7, determining the comprehensive benefits of the trajectories with different distributions of acceptable risk trajectories of the automatically-driven vehicles obtained in the step 6 through a comprehensive benefit function, and selecting one trajectory with the highest comprehensive benefit as the motion trajectory planning of the automatic driving of the signalless intersection based on the risk dynamic balance theory.
2. The method according to claim 1, wherein the method for obtaining the "vehicle travel track prediction" in step 1 comprises the following steps:
step 11, acquiring required basic data including the shape and the size of the signalless intersection, the transverse and longitudinal position, the transverse and longitudinal speed, the transverse and longitudinal acceleration and the yaw angle of a moving object in the intersection at an initial moment through a sensor of an automatic driving vehicle and an internet of vehicles;
step 12, establishing a vehicle track prediction model at the no-signal intersection as follows:
SP(xp,yp,vp,x,vp,y,ap,x,ap,y,ψp,t)＝fP(typev,x(t0),y(t0),vx(t0),vy(t0),ax(t0),ay(t0),ψ(t0))
in the formula, typevRefers to predicting the type of vehicle, (x (t)0),y(t0) ) means an initial t0(v) predicting the lateral and longitudinal position of the vehicle at that time, (v)x(t0),vy(t0) ) means an initial t0Predicting the transverse and longitudinal speed of the vehicle at the moment (a)x(t0),ay(t0) ) means an initial t0Predicting the transverse and longitudinal acceleration of the vehicle at a time, psi (t)0) Means the initial t0Predicting the yaw angle of the vehicle at any moment;
step 13, combining the obtained basic data with the prediction model of the vehicle track of the signalless intersection to obtain the running tracks S of all other moving vehicles in the signalless intersectionP。
3. The method according to claim 1, wherein the method for obtaining the "dynamic risk field" in step 2 comprises the following steps:
step 21, considering the constraint form, the constraint strength and the constraint range of the static traffic environment elements in the signalless intersection to the automatic driving vehicle, and establishing the following risk field model around the static traffic environment elements in the signalless intersection:
Re(xe,ye,t)＝fe(typee,lengthe,widthe,heighte)
in the formula, typeeMeans the type, length, of the static environment elementeIs the length, width, of the static environment elementeRefers to the width, height, of the static environment elementeRefers to the height of the static environmental element;
step 22, considering the constraint form, the constraint strength and the constraint range of other vehicles in the signalless intersection to the automatic driving vehicle, establishing the following risk field model around the moving object in the signalless intersection:
Rv(xv,yv,t)＝fv(lengthv,widthv,x(t),y(t),vx(t),vy(t),ax(t),ay(t),ψ(t))
in the formula, lengthvIs the length, width, of the vehiclevRefers to the width of the vehicle, (x (t), y (t)) refers to the spatial position of the vehicle over time, (v)x(t),vy(t)) means the speed of movement of the vehicle over time, (a)x(t),ay(t)) means the acceleration of the vehicle over time, ψ (t) means the yaw angle of the vehicle over time;
step 23, transferring the risk field model obtained in the above steps to the same coordinate system according to the following coordinate system change formula:
Figure FDA0003159519790000021
in the formula (x)i,yi) The coordinate position of a coordinate point in the ith individual element risk field model coordinate system is defined, (x, y) is the coordinate position of the coordinate point in the final unified xoy coordinate system, and (delta x (i), (delta y (i)), alpha (i)) is the position and the rotation angle of the ith individual element risk field model coordinate system relative to the final unified xoy coordinate system;
and 24, superposing all the individual element risk field models to obtain a dynamic risk field in the signal-free intersection:
Figure FDA0003159519790000022
in the formula, Re,1(x, y, t) is the 1 st static environment element applying a risk value to the coordinate point (x, y) at time t, Rv,1(x, y, t) is the 1 st vehicle applying a risk value to the coordinate point (x, y) at time t, n1Is the number of static environment elements, n, in the signalless intersection2The number of vehicles traveling in the no-signal intersection.
4. The method according to claim 3, wherein the obtaining method of "desired trajectory distribution" in step 3 comprises the steps of:
step 31, determining the driving direction of the automatic driving vehicle at the no-signal intersection, wherein the driving direction is straight, left-turning or right-turning;
step 32, establishing an expected track distribution model of the automatically-driven vehicles at the signalless intersection as follows:
the expected track distribution model of the automatic driving vehicle straight running in the signalless intersection is as follows:
TS(x,y,vx,vy,ax,ay,ψ,t)＝fS(Si,Ae,Al,x(t0),y(t0),vx(t0),vy(t0),ax(t0),ay(t0),ψ(t0))
the distribution model of the expected track of the left turn of the automatic driving vehicle in the signalless intersection comprises the following steps:
TL(x,y,vx,vy,ax,ay,ψ,t)＝fL(Si,Ae,Al,x(t0),y(t0),vx(t0),vy(t0),ax(t0),ay(t0),ψ(t0))
the distribution model of the expected track of the automatic driving vehicle turning right at the signalless intersection comprises the following steps:
TR(x,y,vx,vy,ax,ay,ψ,t)＝fR(Si,Ae,Al,x(t0),y(t0),vx(t0),vy(t0),ax(t0),ay(t0),ψ(t0))
in the formula, SiIn the shape of the intersection, AeMeans the coordinate range of the entering lane of the vehicle entering the intersection, AlThe coordinate range of the possible departure lane of the vehicle from the intersection is defined, (x (t)0),y(t0) ) means an initial t0The transverse and longitudinal positions of the vehicle at the moment of time, (v)x(t0),vy(t0) ) means an initial t0The transverse and longitudinal speed of the vehicle (a)x(t0),ay(t0) ) means an initial t0Moment transverse and longitudinal acceleration of the vehicle, psi (t)0) Means the initial t0The yaw angle of the vehicle at the moment;
step 33, combining the obtained current state data of the self-vehicle and the static environment element information of the no-signal intersection with an expected track distribution model to obtain an expected track distribution T of the automatic driving vehicle at the no-signal intersection0(x,y,vx,vy,ax,ay,ψ,t)。
5. The method according to claim 4, wherein the method for obtaining the risk values corresponding to different expected trajectories in step 4 comprises the following steps:
desired trajectory profile T at the signalless intersection for the autonomous vehicle0(x,y,vx,vy,ax,ayψ, t) of a track Si(x (t), y (t)), calculating the risk R of the trajectory over timei(t)＝R0(Si(x (t), y (t)) and t), and further obtaining the risk size corresponding to all different expected track distributions in the whole expected track distribution.
6. The method of claim 1, the method of determining an "acceptable risk level for autonomous vehicle" in step 5 comprising the steps of:
based on the autopilot system parameters, the expression for the acceptable risk level for an autopilot vehicle is:
RA＝fA(uC,uP,uT)
in the formula uCIs the delay of obtaining information from an autonomous vehicle, uPRefers to the possible error, u, of the perception of the surrounding environment by the autonomous vehicleTRefers to the possible errors in predicting the trajectories of other moving vehicles.
7. The method according to claim 1, wherein the step 6 of obtaining the acceptable risk trajectory distribution comprises the following steps:
when the track S isi(x (t), y (t)) corresponding risk R over timei(t) is less than the acceptable risk level R for the autonomous vehicle calculated from step 5AI.e. Ri(t)≤RAThen the trace distribution belongs to the acceptable risk trace distribution Taccept(x,y,vx,vy,ax,ayψ, T) to obtain an acceptable risk trajectory distribution Taccept(x,y,vx,vy,ax,ay,ψ,t)。
8. The method according to claim 1, wherein the step 7 of obtaining the "motion trail plan" comprises the following steps:
step 71, calculating the comprehensive benefits of all the distribution tracks belonging to the acceptable risk tracks:
Figure FDA0003159519790000031
wherein Si,accept(x(t),y(t))∈Taccept(x,y,vx,vy,ax,ay,ψ,t)
Step 72, comparing all the distributions T belonging to the acceptable risk trajectoryaccept(x,y,vx,vy,ax,ayψ, t) of the track Si(x (t), y (t)) corresponding general profit GiSelecting the track in which the maximum comprehensive profit can be obtained as the motion track plan of the automatic driving vehicle in the signal-free intersection:
Figure FDA0003159519790000032
its comprehensive benefits
Figure FDA0003159519790000033
Then pair
Figure FDA0003159519790000034
Its comprehensive benefits
Figure FDA0003159519790000035
Has Gj≥GiIn the formula, g (S)i,accept(x (t), y (t))) and g (S)j,accept(x (t), y (t))) is a Lagrangian function of the desired trajectory, GiAnd GjRespectively acceptable risk trajectory distribution Taccept(x,y,vx,vy,ax,ayThe comprehensive income corresponding to the ith and jth tracks in psi, t) is obtained as the jth track Sj,accept(x (t), y (t)) are used for planning the motion track of the automatic driving vehicle in the signal-free intersection.
CN202010438570.4A 2020-05-21 2020-05-21 No-signal intersection vehicle motion planning method based on risk dynamic balance Active CN111599179B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010438570.4A CN111599179B (en) 2020-05-21 2020-05-21 No-signal intersection vehicle motion planning method based on risk dynamic balance

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010438570.4A CN111599179B (en) 2020-05-21 2020-05-21 No-signal intersection vehicle motion planning method based on risk dynamic balance

Publications (2)

Publication Number Publication Date
CN111599179A CN111599179A (en) 2020-08-28
CN111599179B true CN111599179B (en) 2021-09-03

Family

ID=72192471

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010438570.4A Active CN111599179B (en) 2020-05-21 2020-05-21 No-signal intersection vehicle motion planning method based on risk dynamic balance

Country Status (1)

Country Link
CN (1) CN111599179B (en)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113055474B (en) * 2021-03-12 2022-11-11 成都格林希尔德交通科技有限公司 Micro road right transaction system
CN113345221B (en) * 2021-05-13 2023-03-28 北京航空航天大学 Method for matching and organizing vehicles at entrance lane of signalless intersection based on parallel lanes
CN113296541B (en) 2021-07-27 2021-11-30 北京三快在线科技有限公司 Future collision risk based unmanned equipment control method and device
CN113635897B (en) * 2021-09-24 2023-03-31 北京航空航天大学 Safe driving early warning method based on risk field
CN113942526B (en) * 2021-11-23 2023-11-03 同济大学 Automatic driving overtaking track planning method based on acceptable risk
CN114543827A (en) * 2022-02-11 2022-05-27 齐鲁工业大学 Path planning method and device
CN114724376B (en) * 2022-05-05 2023-04-28 北京航空航天大学 Intersection safety evaluation method based on risk field theory
CN116168550B (en) * 2022-12-30 2024-07-26 福州大学 Traffic coordination method for intelligent network-connected vehicles at signalless intersections

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107168305B (en) * 2017-04-01 2020-03-17 西安交通大学 Bezier and VFH-based unmanned vehicle track planning method under intersection scene
JP2019105684A (en) * 2017-12-11 2019-06-27 株式会社デンソー Running orbit data generation device in intersection, running orbit data generation program in intersection, and storage medium
CN109389838A (en) * 2018-11-26 2019-02-26 爱驰汽车有限公司 Unmanned crossing paths planning method, system, equipment and storage medium
CN109760675B (en) * 2019-03-12 2021-05-25 百度在线网络技术（北京）有限公司 Method, device, storage medium and terminal equipment for predicting vehicle track
CN110298122B (en) * 2019-07-03 2021-04-27 北京理工大学 Unmanned vehicle urban intersection left-turn decision-making method based on conflict resolution

Also Published As

Publication number Publication date
CN111599179A (en) 2020-08-28

Similar Documents

Publication Publication Date Title
CN111599179B (en) No-signal intersection vehicle motion planning method based on risk dynamic balance
CN113165652B (en) Verifying predicted trajectories using a mesh-based approach
CN110539752B (en) Intelligent automobile multi-prediction-range model prediction trajectory tracking control method and system
EP3814909B1 (en) Using divergence to conduct log-based simulations
CN110562258B (en) Method for vehicle automatic lane change decision, vehicle-mounted equipment and storage medium
CN112700470B (en) Target detection and track extraction method based on traffic video stream
CN110843789B (en) Vehicle lane change intention prediction method based on time sequence convolution network
CN110553660B (en) Unmanned vehicle trajectory planning method based on A-star algorithm and artificial potential field
Yao et al. On-road vehicle trajectory collection and scene-based lane change analysis: Part II
CN107886750B (en) Unmanned automobile control method and system based on beyond-visual-range cooperative cognition
CN105009175A (en) Modifying behavior of autonomous vehicles based on sensor blind spots and limitations
CN110304074A (en) A kind of hybrid type driving method based on stratification state machine
CN105809130A (en) Binocular depth perception-based vehicle travelable area calculation method
CN110673602A (en) Reinforced learning model, vehicle automatic driving decision method and vehicle-mounted equipment
CN112577506B (en) Automatic driving local path planning method and system
CN109643118A (en) The function of vehicle is influenced based on the information relevant to function of the environment about vehicle
CN112249008B (en) Unmanned automobile early warning method aiming at complex dynamic environment
CN109878530B (en) Method and system for identifying lateral driving condition of vehicle
Lefèvre et al. Context-based estimation of driver intent at road intersections
Xu et al. Group vehicle trajectory prediction with global spatio-temporal graph
Guo et al. Toward human-like behavior generation in urban environment based on Markov decision process with hybrid potential maps
US11429843B2 (en) Vehicle operation labeling
Ren et al. Self-learned intelligence for integrated decision and control of automated vehicles at signalized intersections
CN111103882A (en) Autonomous following control method for unmanned electric vehicle
Shan et al. Vehicle collision risk estimation based on RGB-D camera for urban road

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant