CN110488802A - A kind of automatic driving vehicle dynamic behaviour decision-making technique netted under connection environment - Google Patents
A kind of automatic driving vehicle dynamic behaviour decision-making technique netted under connection environment Download PDFInfo
- Publication number
- CN110488802A CN110488802A CN201910774762.XA CN201910774762A CN110488802A CN 110488802 A CN110488802 A CN 110488802A CN 201910774762 A CN201910774762 A CN 201910774762A CN 110488802 A CN110488802 A CN 110488802A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- risk
- decision
- road user
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 41
- 230000033001 locomotion Effects 0.000 claims abstract description 20
- 230000009471 action Effects 0.000 claims abstract description 19
- 238000004891 communication Methods 0.000 claims description 18
- 230000001133 acceleration Effects 0.000 claims description 17
- 238000004364 calculation method Methods 0.000 claims description 10
- 230000008859 change Effects 0.000 claims description 7
- 238000012512 characterization method Methods 0.000 claims description 3
- 230000006399 behavior Effects 0.000 description 62
- 230000006870 function Effects 0.000 description 55
- 238000005259 measurement Methods 0.000 description 12
- 238000005516 engineering process Methods 0.000 description 9
- 238000003860 storage Methods 0.000 description 8
- 230000008447 perception Effects 0.000 description 7
- 230000009467 reduction Effects 0.000 description 6
- 238000011156 evaluation Methods 0.000 description 5
- 230000004888 barrier function Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000002093 peripheral effect Effects 0.000 description 3
- 206010039203 Road traffic accident Diseases 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 239000003550 marker Substances 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 239000003595 mist Substances 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000012502 risk assessment Methods 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 238000013517 stratification Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 230000009885 systemic effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/0055—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots with safety arrangements
- G05D1/0061—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots with safety arrangements for transition from automatic pilot to manual pilot and vice versa
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses the automatic driving vehicle dynamic behaviour decision-making techniques under a kind of net connection environment.The described method includes: step S1, from vehicle in the case where V2X nets connection environment, surrounding road user obtains surrounding enviroment information；And centered on from vehicle mass center, region division is carried out with different radiuses, estimates risk zones；Step S2 based on surrounding road user's surrounding enviroment information and estimates risk zones, carries out first stage behaviour decision making, is determined as the possible action set for guaranteeing to take from vehicle traffic safety；Step S3 carries out second stage behaviour decision making: considering non-safety constraint condition, from the possible action set, the movement that optimum choice finally executes carries out driving behavior decision.
Description
Technical field
The present invention relates to a kind of automatic Pilot fields, join automatic driving vehicle dynamic row under environment especially with regard to a kind of net
For decision-making technique.
Background technique
Intelligent transportation development brings convenience, but also results in potential traffic accident simultaneously.Therefore vehicle safe driving has become
For the key factor of modern intelligent transportation system.Although current active safety system and passive safety system gradually mature and give
Using to avoid vehicle collision, and the influence of reduction accident to the maximum extent, but reduction traffic accident incidence, raising vehicle
Level of security still has strong demand.In recent years, autonomous driving vehicle due to its potential application in avoiding collision and
Cause the strong interest of auto industry.However, automatic Pilot is still one multiple with the target for realizing " on the way zero accident "
Miscellaneous task.
Many basic functions may be implemented because there is various driving assistance systems in automatic driving vehicle at present,
Can such as carry out lane-change auxiliary, adaptive cruise drive, give warning in advance function.However, existing automatic driving vehicle behavior
Decision making algorithm can not carry out accurate risk assessment to potential risks on road, i.e. these automatic Pilot algorithms cannot be identified tight
Which movement is safer in anxious situation, can not make the behaviour decision making of effective and safe.In order to keep automatic driving vehicle more intelligent
Change, existing decision system needs to identify and understand driver-vehicle-road traffic environment integrated information, automatic Pilot vehicle
Decision making algorithm be considered as it is comprehensive drive risk, and graded by severity of the driving risk index to risk of driving a vehicle.
Particularly with complicated road environment, autonomous driving vehicle needs to adapt to mixed traffic form, while really realizing and improving traffic effect
Rate and reduction accident.
Existing programmed decision-making technology can mainly be subdivided into the tool such as mission planning, route planning, conduct programming and motion planning
Body technique link.Typical Vehicle Decision Method plans the different type object being related on road.Current some research work will be certainly
It is dynamic to drive the different operation of vehicle and be considered as individual drive mode, and only consider in one of these modes, A.Furda and
L.Vlacic also proposed the hierarchical decision making method for determining behaviour decision making switching.However, these methods often rely on it is predefined
Strategy, these strategy designs get up very laborious and are easy to unreliable under emergency case.Meanwhile CVX and Ipopt are also realized
Fast and reliable coordinates measurement.This method can be realized from starting and be configured to the complete trajectory that target configures, and complete decision
Planning process, but this kind of methods often do not consider the dynamic and randomness of environment, and cannot accomplish accurately to estimate
Risk, to realize safe avoidance decision.
In face of complex changeable running environment, automatic driving vehicle is required to accurate judgement current environment value-at-risk
And then carry out effective behaviour decision making.The limitation of existing decision-making technique causes intelligent vehicle level of intelligence to be restricted, and therefore, has
Necessity develops a kind of automatic driving vehicle dynamic behaviour decision-making technique and device netted under connection environment.
Summary of the invention
The purpose of the present invention is to provide the automatic driving vehicle dynamic behaviour decision-making techniques and dress under a kind of net connection environment
It sets, the dangerous feelings of automatic driving vehicle effective testing and evaluation current traffic condition under DYNAMIC COMPLEX environment can be enable
Condition, and make vaild act decision automatic driving vehicle can be avoided and collided with other road users.
To achieve the above object, the present invention provides a kind of automatic driving vehicle dynamic behaviour decision-making party netted under connection environment
Method is automatic driving vehicle from vehicle, the described method comprises the following steps S1-S3.
Step S1, from vehicle in the case where V2X nets connection environment, surrounding road user obtains surrounding enviroment information；And with from vehicle
Centered on mass center, region division is carried out with different radiuses, estimates risk zones.
Step S2 based on surrounding road user's surrounding enviroment information and estimates risk zones, carries out first stage behavior
Decision is determined as the possible action set for guaranteeing to take from vehicle traffic safety.
First stage behaviour decision making is that safe and feasible decision is done under special scenes.It is grasped about feasible, safe driving
Vertical decision can be influenced by automatic driving vehicle surrounding enviroment and route planning instruction.If riding manipulation can be in specific friendship
Road traffic regulation is safely executed and meets in logical situation, then the riding manipulation is defined as feasible.In order to ensure peace
Quan Xing, it is assumed that automatic driving vehicle will observe traffic rules and regulations always, while meet safe hard constraint function.Under first layer decision,
It can there are a variety of feasible riding manipulations (for example, overtake other vehicles parking or it is waited to continue to drive) under any traffic conditions.
Step S3 carries out second stage behaviour decision making: non-safety constraint condition is considered, from the possible action set
In, the movement that optimum choice finally executes carries out driving behavior decision.Step S3 be equivalent to from the first stage it is a variety of it is feasible certainly
The decision of the most suitable riding manipulation of selection in plan (possible action set).Second stage behaviour decision making selects and starts to execute list
A riding manipulation (driving behavior), the driving behavior are selected as being most suitable for special traffic conditions.Non-safety constraint condition
Demand or constraint condition e.g. in terms of efficiency, comfort or traffic flow.
V2V (vehicle to vehicle) net connection environment can be a part of existing V2X net connection environment, by the communication technology,
Realize the authentic communication exchange between automatic driving vehicle.V2X net connection environment further realizes automatic Pilot vehicle on this basis
Authentic communication exchange between road infrastructure (V2I) etc..The present invention can effectively realize vehicle in the case where net joins environment
Between cooperation, improve the efficiency and safety of automatic driving vehicle.
From vehicle by using perception algorithm, information is obtained from onboard sensor system, effectively identifies relevant traffic characteristic,
Such as traffic sign, pavement marker, barrier, pedestrian, vehicle etc..It, can for the sensor technology of onboard sensor system
In any light, the accurate and reliable information in relation to vehicle environmental is provided under weather and road conditions.Meanwhile obtaining car's location
The information such as (warp, latitude information), height, speed, acceleration are sent to surrounding road user, while receiving surrounding road use
The same information that person sends, can be by being calculated the opposite distance and position from vehicle of surrounding road user.
Carry out information storage as needed after obtaining information from vehicle, at the same it is necessary in the case where, carry out information more
Newly.Storage environment prior information, such as road, intersection and traffic sign；The information that storage sensor provides, such as obstacle
The traffic sign of object, traffic lane and perception；Storage is by communicating the information obtained with other vehicles or traffic control center；No
The disconnected information for updating prior information and continuously being obtained from sensor and communication component.
Preferably, in step sl, risk zones are estimated in the following manner:
It defines centered on from vehicle mass center, with safe stopping distance LriskBorder circular areas for radius is risk zones,
In formula, viFor from vehicle present speed, amaxFor from vehicle acceleration maximum value, L is from vehicle commander's degree；If road around certain
Road user is located in risk zones, and surrounding road user is defined as risk road user；
Definition is centered on from vehicle mass center, safe early warning distance Lp-riskFor radius border circular areas remove risk zones it
Annular region afterwards is potential risk region,
Wherein, adecFor from the maximum value of vehicle deceleration, if certain surrounding road user is located in potential risk region,
Surrounding road user is defined as potential risk road user；
Definition is centered on from vehicle mass center, safe early warning distance Lp-riskIt is peace for the region except the border circular areas of radius
Entire area, if certain surrounding road user is located at except potential risk region, or when in from except the communication range of vehicle,
Surrounding road user is defined as green route user.
Preferably, step S2 includes the following steps:
Step S21, for the surrounding road user in risk zones and potential risk region, calculation risk degree C,
In, it is first calculated for the surrounding road user in risk zones, then make for the surrounding road in potential risk region
User calculates,
It is clashed between current state and surrounding road user's state locating for risk C characterization automatic driving vehicle
Probability, i.e., the probability for clashing or colliding in the case where being kept the current status unchanged from vehicle and surrounding road user.It should
Probability is estimated value, and is calculated as follows:
Wherein, t is the estimated collision time of estimation, if having two or more in risk zones and potential risk region
A surrounding road user, t are the estimated of two or more estimations for occurring to expect to collide with each surrounding road user
The minimum value of collision time, if not having surrounding road user in risk zones and potential risk region, t is greater than tcSet
Definite value,
tcIt can be rung according to the braking ability of vehicle, the state of emergency for the crash time avoided collision for the constant of setting
Speed, road performance etc. is answered to predefine.For example, defining t in one embodiment of the inventionc=4s.
Work as C=0, then determines that there is no traffic conflicts in this state, determine risk magnitude friskIt is zero, and goes to step
S23；
Work as C=1, then determine there is potential traffic conflict in this state, go to step S22,
Step S22, calculation risk metric frisk,
Step S23, according to risk magnitude friskMovement selection is carried out, determines possible action set.
Preferably, the estimated collision time t of estimation is calculated as follows,
T=min { TTC, PET, TTB }
Wherein,
XiBe from truck position,
XjIt is the position for the surrounding road user being followed by,
viFor from vehicle present speed,
vjFor his vehicle present speed,
LiFor from vehicle commander's degree,
PET is the time t for entering conflict point from vehicleiThe time t of this conflict point is reached to another surrounding road userj's
Difference between time,
PET=t=| ti-tj|
TTB is suitable for from Che Hou, his Che Qian scene for assessing forward region,
XiBe from truck position,
XjIt is his truck position being followed by,
viFor from vehicle present speed,
LiFor from vehicle commander's degree.
In one embodiment, the estimated collision time t of estimation is calculated in the following manner.
On the one hand, when scene can be distinguished, the estimated collision time t of the estimation for the scene is only calculated, wherein right
In straight way follow the bus scene, t=TTC；For intersection scene, t=PET；For colliding scene, t=from Che Hou forward direction
TTB；That is, need to only calculate the risk function for the scene, specifically, TTC is main when scene can be distinguished
For straight way follow the bus scene, PET is suitable for intersection scene, and TTB is suitable for colliding scene from Che Hou forward direction.Calculating wind
When the measurement of danger, preferred TTC index is assessed, and choosing other again for not being suitable for TTC scene is applicable in index.When appearance three
When difference occurs in a index evaluation result, the index for choosing greatest risk carries out decision output.And then according to multiple target output
Risk is further analyzed and Decision-Making Intervention.
PET mainly can capture influence of other the certain intersection features to safety, because including other crosspoints spy
Limit influence can be generated to predictive ability by levying (such as sighting distance, grade and other parameters) only, therefore mainly applicable scene is to intersect
Mouthful.TTC can be applied to different types of conflict, such as rear end, front and right angle collision, but it is more quasi- in straight way scene
Really.TTB is mainly for assessment of forward region, i.e., measurement is not used in Background Region, is suitable for from vehicle in his rear Che Qian scene.
On the other hand, three indexs are calculated when scene complicated difficult is to distinguish scene to be minimized,
T=min { TTC, PET, TTB },
Wherein,
XiBe from truck position,
XjIt is the position for the surrounding road user being followed by,
viFor from vehicle present speed,
vjFor his vehicle present speed,
LiFor from vehicle commander's degree,
PET is the time t for entering conflict point from vehicleiThe time t of this conflict point is reached to another surrounding road userj's
Difference between time,
PET=t=| ti-tj|
TTB is suitable for from Che Hou, his Che Qian scene for assessing forward region,
XiBe from truck position,
XjIt is his truck position being followed by,
viFor from vehicle present speed,
LiFor from vehicle commander's degree.
In different traffic scenes, if there are other multiple road users under Same Scene, individually assessment is each
The estimated collision time of the estimation of road user, then determined by using multi-object Threat assessment algorithm " effective " or
The estimated collision time risk t of " equivalent " estimation.Multi-object Threat assessment algorithm for example, by using being minimized algorithm, or
Weighting is minimized algorithm.
Weighting is minimized algorithm for example are as follows: to the estimated collision time t1 of the smallest estimation imparting weighting coefficient a, a >
0.5；Weighting coefficient a* (1-a) is assigned to the estimated collision time t2 of the second small estimation；The estimated of the estimation small to third touches
It hits time t2 and assigns weighting coefficient a* (1-a-a* (1-a)) etc..A is, for example, 0.6 or 2/3.
Decision-making structures proposed by the invention be suitable for multiple target scene, and then according to multiple target output risk come into
The analysis of one step and Decision-Making Intervention.
Preferably, in step s 2, risk magnitude f is calculated as followsrisk,
Or
It preferably, does not include the decision attribute of any influence safety in the second stage behaviour decision making of step S3, but
Consider efficient soft-constraint function fe, comfortable soft-constraint function fcWith traffic flow soft-constraint function ftCarry out optimizing decision.
Second stage behaviour decision making i.e. the second decision phase.Optimal decision method is found in second decision phase.From
The decision of most suitable riding manipulation is selected in one layer of a variety of feasible decisions.The stage selects and starts to execute single driving behaviour
It is vertical, it is selected as being most suitable for special traffic conditions.Due to this stage only consider those be selected as it is feasible (therefore be safety
) riding manipulation, therefore the stage do not include safety it is most important because it does not include the decision category of any influence safety
Property.Second decision phase depended primarily on efficient soft-constraint function fe, comfortable soft-constraint function fc, traffic flow soft-constraint function ft。
Preferably, efficient soft-constraint function feIs defined as:
Wherein, t0Initially to set out the moment from vehicle, tfTo arrive at the destination the moment from vehicle, S (t) be from vehicle from starting point to mesh
The walked path in ground, be (t) from vehicle speed.
Preferably, comfortable soft-constraint function fcIs defined as:
Wherein, a is from vehicle acceleration, alatFor transverse acceleration, alonFor longitudinal acceleration.
Comfort is reflected by vehicle performance index, mainly for because of automatic driving vehicle decision in the present invention
Change on passenger mentality physiological comfort caused by vehicle own mechanical structure caused by mobility and assembling manufacturing horizontal vibration
Change, wherein the manipulative behaviors such as anxious acceleration, anxious deceleration can significantly be impacted to passenger.
Preferably, traffic flow soft-constraint function ftIs defined as:
minft=α (vave-vder)2+β(dave-dder)2
Wherein,
vaveFor the velocity levels of the preceding periphery traffic flow of decision,
vderFor the desired velocity levels of decision rear perimeter edge traffic flow,
daveFor the average following distance of the preceding periphery traffic flow of decision,
dderFor the desired average following distance of decision rear perimeter edge traffic flow,
α, β are weight coefficient, both greater than 0 and less than 1.
So that the disturbance to surrounding vehicles is minimum.The disturbance of the dynamic characteristic of surrounding traffic stream is imitated from car state variation
It answers, is embodied in the variation of their desired average speeds, distance level of other vehicle distances.
Because of the decision behavior bring potential impact of automatic driving vehicle, i.e., surrounding traffic stream is moved from car state variation
The disturbance effect of step response is embodied in the influence to the traffic capacity and traffic flow stability and patency of peripheral path.
Therefore, traffic flow soft-constraint function f is definedtAutomatic driving vehicle is differentiated when making behaviour decision making, needs to measure the behavior
Disadvantage influence, the interruption for being embodied in the reduction of speed or giving it the gun are generated favorably or had on traffic flow.
In the case where net joins environment, periphery traffic flow range is defined as from vehicle in vehicle communication range.
Preferably, in the second stage behaviour decision making of step S3, it is as follows to define cost function J:
w1、w2、w3For weight coefficient, both greater than 0 and less than 1, and w1+w2+w3=1
Wherein, fe0、fc0、ft0It respectively indicates safe and efficient, easypro after assuming to continue to execute from vehicle according to state before decision
Suitable, traffic stream function.
Judgement makes whether the decision is reasonable, that is, the cost function J after wanting comparison decision.The minimum value of cost function J is
Corresponding optimal solution.
Method of the invention estimates risk area according to the traffic risk function by the information exchange between vehicle
Domain carries out behaviour decision making stage by stage, improves speed of decision, efficiency and safety, can help to realize and be driven automatically by control
Sail the movement of vehicle and it is safe and efficient, cosily arrive at the destination.
Detailed description of the invention
Fig. 1 is the schematic flow chart of automatic driving vehicle dynamic behaviour decision-making technique provided in an embodiment of the present invention.
Fig. 2 is shown using the control system of automatic driving vehicle dynamic behaviour decision-making technique provided in an embodiment of the present invention
Meaning property frame figure；
Fig. 3 is risk class region division schematic diagram provided in an embodiment of the present invention.
Fig. 4 is the schematic diagram for illustrating first stage behaviour decision making in the present invention.
Fig. 5 is the schematic diagram for illustrating first stage behaviour decision making and second stage behaviour decision making in the present invention.
Specific embodiment
The embodiment of the present invention is described in detail with reference to the accompanying drawing.
Referring to Fig. 1, a kind of net of the embodiment of the present invention joins the automatic driving vehicle dynamic behaviour decision-making technique under environment, institute
The method of stating includes the following steps S1-S3.
Step S1, from vehicle in the case where V2X nets connection environment, surrounding road user obtains surrounding enviroment information；And with from vehicle
Centered on mass center, region division is carried out with different radiuses, estimates risk zones.It is automatic driving vehicle from vehicle.Surrounding enviroment
Position, velocity information etc. of the information for example including road users such as surrounding vehicles, bicyclist, pedestrian and barriers.Estimate wind
Danger zone domain is equivalent to the pretreatment of the anticipation link of the risk in Fig. 4, determines risk zones.
Further, in decision phase 1 (i.e. first stage behaviour decision making), for example, as shown in Fig. 2, being prejudged in risk
Link determines risk road user, potential risk road occupation by determining whether road user is in risk zones
Person and green route user.In one embodiment, Ta Che and from distance definition between vehicle be his vehicle headstock to from vehicle mass center
Distance on line direction.
In addition, in step sl, can be precalculated or be determined for the first stage behaviour decision making in step S2, example
Such as, potential conflict, estimation conflict time, probability etc. are determined by traffic risk stratification evaluation module (referring to fig. 2).Above-mentioned sentences
Fixed, estimation etc., can also be calculated in step s 2.For example, collision detection link shown in Fig. 4 carries out above-mentioned estimation
With judgement.
V2X (Vehicle to Everything) function vehicle is carried to typically refer to refer to the advanced vehicle-mounted sensing of carrying
The devices such as device, controller, actuator (i.e. car networking network aware module in Fig. 2), and modern communications and network technology are merged,
It realizes that vehicle and exchanging for X (people, vehicle, road, backstage etc.) intelligent information are shared, has complex environment perception, intelligent decision, collaboration
Control and the functions such as execute, it can be achieved that safety, comfortable, energy conservation, efficiently traveling, and final alternative people is come the vehicle that operates.From
It is dynamic to drive the automatic driving vehicle that vehicle includes each automatic Pilot kind grade.
Compared with camera common in automatic Pilot technology or laser radar, V2X possesses wider array of use scope, it has
Have and break through visual dead angle and the information obtaining ability across shelter, while can be with the shared driving in real time of other vehicles and facility
Status information can also generate predictive information by studying and judging algorithm.In addition, V2X be uniquely do not influenced by weather conditions it is automobile-used
Sensing technology, no matter rain, mist or strong illumination all will not influence its normal work.
In addition, V2X further highlights " intelligence except the shared function with environment sensing of traditional intelligence automobile information exchange
Decision ", " Collaborative Control and execution " function, based on powerful back-end data analysis, decision, dispatching service system.And
It realizes that automatic Pilot, vehicle must have sensory perceptual system, the environment of surrounding can be observed as people, so in addition to sensing
Device, V2X technology also belong to a perception means of automatic Pilot.
V2V (vehicle to vehicle) net connection environment can be a part of existing V2X net connection environment, by the communication technology,
Realize the authentic communication exchange between automatic driving vehicle.V2X net connection environment further realizes automatic Pilot vehicle on this basis
Authentic communication exchange between road infrastructure (V2I).The present invention can net join environment under effectively realize vehicle it
Between cooperation, improve the efficiency and safety of automatic driving vehicle.
From vehicle by using perception algorithm, information is obtained from onboard sensor system, effectively identifies relevant traffic characteristic,
Such as traffic sign, pavement marker, barrier, pedestrian, vehicle etc..It, can for the sensor technology of onboard sensor system
In any light, the accurate and reliable information in relation to vehicle environmental is provided under weather and road conditions.Meanwhile obtaining car's location
The information such as (warp, latitude information), height, speed, acceleration are sent to surrounding road user, while receiving surrounding road use
The same information that person sends, can be by being calculated the opposite distance and position from vehicle of surrounding road user.
Carry out information storage as needed after obtaining information from vehicle, at the same it is necessary in the case where, carry out information more
Newly.Storage environment prior information, such as road, intersection and traffic sign；The information that storage sensor provides, such as obstacle
The traffic sign of object, traffic lane and perception；Storage is by communicating the information obtained with other vehicles or traffic control center；No
The disconnected information for updating prior information and continuously being obtained from sensor and communication component.
Step S2 and S3 is executed in multilayer behaviour decision making module.
Step S2 based on surrounding road user's surrounding enviroment information and estimates risk zones, carries out first stage behavior
Decision (decision phase 1 in Fig. 2), is determined as the possible action set for guaranteeing to take from vehicle traffic safety.First stage
Behaviour decision making meets the requirement of safe hard constraint function.The result is that obtaining possible action set (i.e. all feasible decisions in Fig. 2
Collection).
As shown in figure 4, first stage behaviour decision making generally includes following several links: risk prejudges link, conflict inspection
Grommet section, risk measurement link and movement selection link.Risk anticipation link mainly judges risk road user.Conflict inspection
It surveys link and is mainly used for whether preliminary judgement can clash (collision).Risk measurement is mainly calculation risk metric.It can be with
Carry out calculation risk metric in any suitable manner.Other than the risk magnitude introduced in detail below,
First stage behaviour decision making is that safe and feasible decision is done under special scenes.It is grasped about feasible, safe driving
Vertical decision can be influenced by automatic driving vehicle surrounding enviroment and route planning instruction.If riding manipulation can be in specific friendship
Road traffic regulation is safely executed and meets in logical situation, then the riding manipulation is defined as feasible.In order to ensure peace
Quan Xing, it is assumed that automatic driving vehicle will observe traffic rules and regulations always, while meet safe hard constraint function.Under first layer decision,
It can there are a variety of feasible riding manipulations (for example, overtake other vehicles parking or it is waited to continue to drive) under any traffic conditions.
Step S3 is carried out second stage behaviour decision making (i.e. decision phase 2 in Fig. 2): carry out second stage behaviour decision making:
Consider non-safety constraint condition, from the possible action set, the movement that optimum choice finally executes carries out driving behavior
Decision.Step S3 is equivalent to the most suitable riding manipulation of selection from a variety of feasible decisions (possible action set) of first stage
Decision.Second stage behaviour decision making selects and starts to execute single riding manipulation (driving behavior), and the driving behavior is selected
It is selected as being most suitable for special traffic conditions.Non-safety constraint condition is, for example, the need in terms of efficiency, comfort or traffic flow
It asks or constraint condition.It is carried out that is, second stage behaviour decision making is based primarily upon efficient, comfortable, traffic flow constraint function.Knot
Fruit is to obtain or make optimum behavior decision.Referring to Fig. 5, second stage behaviour decision making is exported based on first stage behaviour decision making
Optimum behavior decision is done on the basis of optional set of actions.
According to some embodiments of the present invention, in step sl, risk zones are estimated in the following manner:
It defines centered on from vehicle mass center, with safe stopping distance LriskBorder circular areas for radius is risk zones,
In formula, viFor from vehicle present speed, amaxFor from vehicle acceleration maximum value, L is from vehicle commander's degree；If road around certain
Road user is located in risk zones, and surrounding road user is defined as risk road user；
Definition is centered on from vehicle mass center, safe early warning distance Lp-riskFor radius border circular areas remove risk zones it
Annular region afterwards is potential risk region,
Wherein, adecFor from the maximum value of vehicle deceleration, if certain surrounding road user is located in potential risk region,
Surrounding road user is defined as potential risk road user；
Definition is centered on from vehicle mass center, safe early warning distance Lp-riskIt is peace for the region except the border circular areas of radius
Entire area, if certain surrounding road user is located at except potential risk region, or when in from except the communication range of vehicle,
Surrounding road user is defined as green route user.It is L that i.e. green route user region, which is in radius,safe
=min { Lcom,Lp-riskExcept region.Wherein with communication radius LcomFor round range refer to net join environment under, from parking stall in
The range of peripheral path user's information can be observed and obtained in its specified region.
Specifically, as shown in figure 3, the present invention is to centered on vehicle centroid, radius is communication radius length LcomCircle into
Row region division, being specifically divided into risk road user, potential risk road user and green route according to risk class makes
User.For the vehicle observed, three distance metrics relevant to above-mentioned three kinds of risk class are defined, if locating range
More than threshold value, then the risk measurement for the vehicle observed can then change correspondingly.Wherein, as shown in figure 3, defining risk road occupation
Person, potential risk road user line of demarcation be with safe stopping distance LriskIt is wind within border circular areas for the circle of radius
Danger zone domain.The line of demarcation for defining potential risk road user and green route user is with safe early warning distance Lp-risk's
Circle is (L when being in radiusrisk,Lp-risk) between annular region within be potential risk region.When vehicle and conflict point
When distance is in greater than security warning distance or his vehicle not in the communication range from vehicle, then define the vehicle makes for green route
User.It is L that i.e. green route user region, which is in radius,safe=min { Lcom,Lp-riskExcept region.Wherein
With communication radius LcomRefer to for round range in the case where net joins environment, from parking stall in can observe and obtain week in its specified region
The range of side road user information.
In one embodiment, step S2 includes the following steps:
Step S21, for the surrounding road user in risk zones and potential risk region, calculation risk degree C,
In, it is first calculated for the surrounding road user in risk zones, then make for the surrounding road in potential risk region
User calculates,
It is clashed between current state and surrounding road user's state locating for risk C characterization automatic driving vehicle
Probability, i.e., the probability for clashing or colliding in the case where being kept the current status unchanged from vehicle and surrounding road user.It should
Probability is estimated value, and is calculated as follows:
Wherein, t is the estimated collision time of estimation, if having two or more in risk zones and potential risk region
A surrounding road user, t are the estimated of two or more estimations for occurring to expect to collide with each surrounding road user
The minimum value of collision time, if not having surrounding road user in risk zones and potential risk region, t is greater than tcSet
Definite value,
tcIt can be rung according to the braking ability of vehicle, the state of emergency for the crash time avoided collision for the constant of setting
Speed, road performance etc. is answered to predefine.For example, defining t in one embodiment of the inventionc=4s.
Work as C=0, then determines that there is no traffic conflicts in this state, determine risk magnitude friskIt is zero, and goes to step
S23。
Work as C=1, then determines there is potential traffic conflict in this state, go to step S22.
Step S22, calculation risk metric frisk, risk magnitude friskIt can carry out in any suitable manner
It calculates.For convenient for evaluation, risk magnitude friskThe number being set as between 0-1.
Step S23, according to risk magnitude friskMovement selection is carried out, determines possible action set.
It should be pointed out that the total collection for the movement that vehicle is able to carry out is different under different concrete scenes.Example
Such as, referring to the movement total collection under each scene of table 1.
Automatic driving vehicle acts total collection under the different traffic scenes of table 1
| Action sequence traffic scene | Situation 1 | Situation 2 | Situation 3 | Situation 4 | Situation p |
| Accelerate to pass through | √ | √ | √ | — | |
| Directly pass through | √ | √ | — | ||
| Slow down | √ | √ | √ | — | |
| Parking waiting | √ | √ | — | ||
| Successively follow | √ | √ | — | ||
| Turning is passed through | √ | √ | — | ||
| Act q | — | — | — | — | — |
For some classical scenes, by decision phase 1 optional, the results are shown in Table 1.Such as to the further explanation of table 1
Under:
Situation 1: it is traveled freely scene.Automatic driving vehicle is along road driving, apart from upcoming intersection phase
To farther out.In this case, risk can be exported under based on first layer assessment judgement, then carries out second layer decision, provided
Feasible three kinds of speed executes selection (that is, accelerating, former speed is slowed down).
Situation 2: crossroad scene.For automatic driving vehicle close to a crossroad, there are other dynamic roads at crossing
Road user (e.g. a, pedestrian to go across the road, cyclist, other vehicles etc.).In this case, it is assessed based on first layer
Judgement is lower can to export risk, then carries out second layer decision, provide feasible four kinds of operations execute selection (that is, wait for parking, according to
Secondary to follow, directly pass through, turning is passed through) and three kinds of speed executes selection (that is, accelerating, original is fast, deceleration).
Situation 3: follow the bus scene.Autonomous driving vehicle is chasing after another automobile.In this case, it is being based on first layer
Assessment judgement is lower can to export risk, then carries out second layer decision, determine the feasible two kinds of operations of driver behavior select (that is, according to
Secondary to follow, turning is passed through) and speed to execute selection (that is, acceleration, deceleration) be feasible.
Situation 4: barrier obstruction scene.Static-obstacle thing (tree) is located at the front of automatic driving vehicle.Unique feasible
Driver behavior is emergent stopping.
Situation p: refer to other Driving Scenes except above-mentioned 4 kinds of situations.
Movement q: refer to other movements except above-mentioned 7 kinds of movements.
"-": it indicates just to be capable of determining whether according to specific scene and specific movement feasible.
First stage behaviour decision making is that under the premise that security is guaranteed, from the movement total collection of table 1, selecting can take action
Work gathers (i.e. all feasible decision sets in Fig. 2).More specifically, in step S23, according to risk magnitude friskIt is acted
Selection, determines possible action set.
Table 2 illustrates different risk magnitude f by taking the scene of crossroad as an exampleriskCorresponding possible action set.
The optional movement Choose for user table of automatic driving vehicle under the different risk magnitudes of table 2
| Movement selection risk magnitude | 0 | (0,0.4] | (0.4,0.7] | (0.7,1] |
| Accelerate to pass through | √ | |||
| Directly pass through | √ | |||
| Slow down | √ | √ | ||
| Parking waiting | √ | √ | √ | |
| Successively follow | √ | √ | ||
| Turning is passed through | √ | √ | √ | √ |
In the specific embodiment of the invention, specific decision process, which can be divided into, judges carry out state under external condition driving
The security attribute of selection is determined by safe hard constraint function；It takes decision to act then according to concrete scene, such as adds and subtracts
Speed traveling, turning lane-change, speed holding and parking waiting etc..
For example, being calculated by optimization, in the case where considering efficiency (efficient), the comfortable and concrete condition of traffic flow for difference
Specific driving situation, final decision use different movements.
Table 3 is based on different efficiency, comfortable and traffic flow demand, determining movement
In some embodiments of the invention, in the decision phase 1, safe hard constraint function (risk magnitude letter is introduced
Number), safe and feasible decision is done under special scenes.
In an alternative embodiment of the invention, the estimated collision time t of estimation is calculated as follows,
T=min { TTC, PET, TTB }
Wherein,
XiBe from truck position,
XjIt is the position for the surrounding road user being followed by,
viFor from vehicle present speed,
vjFor his vehicle present speed,
LiFor from vehicle commander's degree,
PET is the time t for entering conflict point from vehicleiThe time t of this conflict point is reached to another surrounding road userj's
Difference between time,
PET=t=| ti-tj|
TTB is suitable for from Che Hou, his Che Qian scene for assessing forward region,
XiBe from truck position,
XjIt is his truck position being followed by,
viFor from vehicle present speed,
LiFor from vehicle commander's degree.
In one embodiment, the estimated collision time t of estimation is calculated in the following manner.
On the one hand, when scene can be distinguished, the estimated collision time t of the estimation for the scene is only calculated, wherein right
In straight way follow the bus scene, t=TTC；For intersection scene, t=PET；For colliding scene, t=from Che Hou forward direction
TTB；That is, need to only calculate the risk function for the scene, specifically, TTC is main when scene can be distinguished
For straight way follow the bus scene, PET is suitable for intersection scene, and TTB is suitable for colliding scene from Che Hou forward direction.Calculating wind
When the measurement of danger, preferred TTC index is assessed, and choosing other again for not being suitable for TTC scene is applicable in index.When appearance three
When difference occurs in a index evaluation result, the index for choosing greatest risk carries out decision output.And then according to multiple target output
Risk is further analyzed and Decision-Making Intervention.
PET mainly can capture influence of other the certain intersection features to safety, because including other crosspoints spy
Limit influence can be generated to predictive ability by levying (such as sighting distance, grade and other parameters) only, therefore mainly applicable scene is to intersect
Mouthful.TTC can be applied to different types of conflict, such as rear end, front and right angle collision, but it is more quasi- in straight way scene
Really.TTB is mainly for assessment of forward region, i.e., measurement is not used in Background Region, is suitable for from vehicle in his rear Che Qian scene.
On the other hand, three indexs are calculated when scene complicated difficult is to distinguish scene to be minimized,
T=min { TTC, PET, TTB },
Wherein,
XiBe from truck position,
XjIt is the position for the surrounding road user being followed by,
viFor from vehicle present speed,
vjFor his vehicle present speed,
LiFor from vehicle commander's degree,
PET is the time t for entering conflict point from vehicleiThe time t of this conflict point is reached to another surrounding road userj's
Difference between time,
PET=t=| ti-tj|
TTB is suitable for from Che Hou, his Che Qian scene for assessing forward region,
XiBe from truck position,
XjIt is his truck position being followed by,
viFor from vehicle present speed,
LiFor from vehicle commander's degree.
In different traffic scenes, if there are other multiple road users under Same Scene, individually assessment is each
The estimated collision time of the estimation of road user, then determined by using multi-object Threat assessment algorithm " effective " or
The estimated collision time risk t of " equivalent " estimation.Multi-object Threat assessment algorithm for example, by using being minimized algorithm, or
Weighting is minimized algorithm.
Weighting is minimized algorithm for example are as follows: to the estimated collision time t1 of the smallest estimation imparting weighting coefficient a, a >
0.5；Weighting coefficient a* (1-a) is assigned to the estimated collision time t2 of the second small estimation；The estimated of the estimation small to third touches
It hits time t2 and assigns weighting coefficient a* (1-a-a* (1-a)) etc..A is, for example, 0.6 or 2/3.
Decision-making structures proposed by the invention be suitable for multiple target scene, and then according to multiple target output risk come into
The analysis of one step and Decision-Making Intervention.
Risk class locating for vehicle is more accurately assessed by calculation risk metric.Specifically, the vehicle observed
Risk measurement depend on the relative velocities of the distance between the vehicle observed and automatic driving vehicle and two vehicles.For
The potential collision threat between vehicle and automatic driving vehicle that assessment is observed, TTC are used as common threat measure, because
It is time-based for it and includes spatial proximity and speed difference.But in certain highways, only TTC is
Inadequate.For example, it is contemplated that such situation: the vehicle observed in the lane adjacent with automatic driving vehicle with drive automatically
It sails vehicle closely and is travelled with similar speed.This will result in big TTC value, to the situation is evaluated as low
Risk (if TTC is used only).However, in fact, sample situation is actually breakneck；Therefore, in this case,
Lane changing manipulation (to avoid collision) should not be executed.In addition, TTC does not consider exception, such as the vehicle that forward observation arrives
The case where stopping suddenly for some reason.Therefore, in order to supplement TTC, other two kinds of assessment sides also are used to situation assessment
Method: (1) PET, for describing the time difference that two vehicles enter conflict point, (2) TTB, be defined as should using emergency braking to prevent
The only remaining time of unexpected shutdown or motor-driven collision.According to the definition of TTB, the measurement of TTB is not used in Background Region.Therefore, fixed
Risk class locating for the vehicle of different zones is assessed with different risk magnitudes stage by stage in adopted subregion.
The judgement that conflict whether there is usually is measured by estimating the criticality of traffic condition.To being at present
Only, different safety indexs, such as collision time (TTC) have been developed, time (PET) after intrusion, uneasy theoretical density (UD),
The rate of deceleration (DRAC) avoided collision, the ratio (PSD) of stop distance, off time (GT), generalized time measure (CTM), chase after
Tail collision probability (RECP) etc., wherein most widely used index first is that collision time TTC.Its definition is and two cars phase
Remaining time before hitting, this two cars are travelled on same route with initial velocity.However, when TTC judges dangerous situation, one
As two vehicles it is closer apart from the point of impingement, therefore cannot reflect whether that very well there are potential conflicts.The present invention uses PET
(postencrotime) this index is measured, and PET is to be measured with two cars by the time of same position.Permit
Perhaps a kind of measure of observable of being consistent property is the time (PET) after intrusion between observer and position.PET is first car
Terminate conflict area and second car into the difference between the time of conflict area.PET only needs two timestamps to carry out
It calculates, it has specific boundary to distinguish the advantage of collapse and non-crash events.PET value indicates to collide for 0, and non-zero
PET value indicates that collision is close.Although it does not describe the initial stage of conflict, also without the row taken of description correlation driver
It is dynamic, but it shows the result of final stage and offer is relatively close to the measurement of collision.Current research has evaluated PET as replacing
For the validity of property safety measure, to prevent from opposing by vehicle collision.
In an alternative embodiment of the invention, in step s 2, risk magnitude f is calculated as followsrisk,
In this alternative embodiment, it is assumed that t=3s, frisk=(4-3)/4=0.25.Assuming that t=1s, frisk=(4-1)/4
=0.75.Assuming that t=0.4s, frisk=(4-0.4)/4=0.9.
In another alternative embodiment of the invention, in step s 2, risk magnitude f is calculated as followsrisk,
In another alternative embodiment, with the reduction for estimating collision time, risk magnitude is quicklyd increase.From
And further protrude risk.Specifically, in this alternative embodiment, it is assumed that t=3s, frisk=(4*4-3*3)/(4*4)=
0.44.Assuming that t=1s, frisk=(4*4-1*1)/(4*4)=0.94 assumes t=0.4s, frisk=(4*4-0.4*0.4)/(4*
4)=0.99.
It specifically, does not include the decision attribute of any influence safety in the second stage behaviour decision making of step S3, but
Consider efficient soft-constraint function fe, comfortable soft-constraint function fcWith traffic flow soft-constraint function ftCarry out optimizing decision.It can also be only
One or two of only consider other constraint functions, or only consider above-mentioned soft-constraint function.
Second stage behaviour decision making i.e. the second decision phase.Optimal decision method is found in second decision phase.From
The decision of most suitable riding manipulation is selected in one layer of a variety of feasible decisions.The stage selects and starts to execute single driving behaviour
It is vertical, it is selected as being most suitable for special traffic conditions.Due to this stage only consider those be selected as it is feasible (therefore be safety
) riding manipulation, therefore the stage do not include safety it is most important because it does not include the decision category of any influence safety
Property.Second decision phase depended primarily on efficient soft-constraint function fe, comfortable soft-constraint function fc, traffic flow soft-constraint function ft。
Preferably, efficient soft-constraint function feIs defined as:
Wherein, t0Initially to set out the moment from vehicle, tfTo arrive at the destination the moment from vehicle, S (t) be from vehicle from starting point to mesh
The walked path in ground, v (t) is from vehicle speed.
Preferably, comfortable soft-constraint function fcIs defined as:
Wherein, a is from vehicle acceleration, alatFor transverse acceleration, alonFor longitudinal acceleration.
Comfort is reflected by vehicle performance index, mainly for because of automatic driving vehicle decision in the present invention
Change on passenger mentality physiological comfort caused by vehicle own mechanical structure caused by mobility and assembling manufacturing horizontal vibration
Change, wherein the manipulative behaviors such as anxious acceleration, anxious deceleration can significantly be impacted to passenger.
Preferably, traffic flow soft-constraint function ftIs defined as:
minft=α (vave-vder)2+β(dave-dder)2
Wherein,
vaveFor the velocity levels of the preceding periphery traffic flow of decision,
vderFor the desired velocity levels of decision rear perimeter edge traffic flow,
daveFor the average following distance of the preceding periphery traffic flow of decision,
dderFor the desired average following distance of decision rear perimeter edge traffic flow,
α, β are weight coefficient, both greater than 0 and less than 1.
So that the disturbance to surrounding vehicles is minimum.The disturbance of the dynamic characteristic of surrounding traffic stream is imitated from car state variation
It answers, is embodied in the variation of their desired average speeds, distance level of other vehicle distances.
Because of the decision behavior bring potential impact of automatic driving vehicle, i.e., surrounding traffic stream is moved from car state variation
The disturbance effect of step response is embodied in the influence to the traffic capacity and traffic flow stability and patency of peripheral path.
Therefore, traffic flow soft-constraint function f is definedtAutomatic driving vehicle is differentiated when making behaviour decision making, needs to measure the behavior
Disadvantage influence, the interruption for being embodied in the reduction of speed or giving it the gun are generated favorably or had on traffic flow.
In the case where net joins environment, periphery traffic flow range is defined as from vehicle in vehicle communication range.
Preferably, in the second stage behaviour decision making of step S3, it is as follows to define cost function J:
w1、w2、w3For weight coefficient, both greater than 0 and less than 1, and w1+w2+w3=1.w1、w2、w3It can be systemic presupposition
, it can also be adjusted according to the demand of motroist.For example, motroist's current demand is that most fast speed arrives at the destination, then
w1It is set as larger, for example, 0.5,0.6,0.8, even 1.System can also force setting w3More than or equal to minimum value, such as
0.1。
Wherein, fe0、fc0、ft0Respectively indicate efficient, comfortable, friendship after assuming to continue to execute from vehicle according to state before decision
Through-flow function.It should be pointed out that efficient, comfortable, traffic stream function is not limited to the function being given above, it can also use and appoint
The corresponding constraint function of what form known.Moreover, using corresponding efficient, comfortable, traffic flow constraint function the reality of form known
Example, and the embodiment including other non-security constraint functions are applied, all within protection scope of the present invention.
Judgement makes whether the decision is reasonable, that is, the cost function J after wanting comparison decision.The minimum value of cost function J is
Corresponding optimal solution.
The method of the embodiment of the present invention is by the information exchange between vehicle, according to the traffic risk function, in advance
Estimate risk zones, carry out behaviour decision making stage by stage, improve speed of decision, efficiency and safety, can help to realize and pass through control
The movement of automatic driving vehicle processed and it is safe and efficient, cosily arrive at the destination.
The particular embodiment of the present invention can be realized following advantages.
1, by using different risk measurement value functions to different scenes, it can guarantee the accuracy of assessment.
2, the automatic driving vehicle dynamic behaviour decision-making technique under the net connection environment proposed in the embodiment of the present invention combines
Branch scape hierarchical decision making thought will incorporate safety guarantee function, can guarantee the safety of each step decision, together in the decision phase
When it is further contemplated that automatic driving vehicle class human nature, that is, introduce efficient soft-constraint function, comfortable soft-constraint function, traffic flow soft-constraint letter
Several soft-constraints can further improve intelligent vehicle intelligent level.
4, the present invention is based on the information that perception subsystem provides, Real-time Decision and riding manipulation control subsystem make driving
Decision.Designed each algorithm (for example, road tracks, can overtake other vehicles, intersection etc.) under special traffic conditions and grasp
Vertical vehicle.Algorithm can be used for assisting driver (the especially driver of the automatic driving vehicle of lower grade) in many ways,
For example, warning driver will collide or automatically apply braking under crucial traffic conditions or turn to, or directly apply
Decision rule is carried out to vehicle behavior in automatic driving vehicle.
Finally it is noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations.This
The those of ordinary skill in field is it is understood that be possible to modify the technical solutions described in the foregoing embodiments or right
Part of technical characteristic is equivalently replaced；These are modified or replaceed, and it does not separate the essence of the corresponding technical solution originally
Invent the spirit and scope of each embodiment technical solution.
Claims (11)
1. the automatic driving vehicle dynamic behaviour decision-making technique under a kind of net connection environment, is automatic driving vehicle, feature from vehicle
It is, the described method comprises the following steps:
Step S1, from vehicle in the case where V2X nets connection environment, surrounding road user obtains surrounding enviroment information；And with from vehicle mass center
Centered on, region division is carried out with different radiuses, estimates risk zones；
Step S2 based on surrounding road user's surrounding enviroment information and estimates risk zones, carries out first stage behaviour decision making,
It is determined as the possible action set for guaranteeing to take from vehicle traffic safety；
Step S3 carries out second stage behaviour decision making: consider non-safety constraint condition, it is excellent from the possible action set
Change the movement that selection finally executes, carries out driving behavior decision.
2. automatic driving vehicle dynamic behaviour decision-making technique as described in claim 1, which is characterized in that in step sl, with
Following manner estimates risk zones:
It defines centered on from vehicle mass center, with safe stopping distance LriskBorder circular areas for radius is risk zones,
In formula, viFor from vehicle present speed, amaxFor from vehicle acceleration maximum value, L is from vehicle commander's degree；If certain surrounding road uses
Person is located in risk zones, and surrounding road user is defined as risk road user；
Definition is centered on from vehicle mass center, safe early warning distance Lp-riskAfter the border circular areas removal risk zones of radius
Annular region is potential risk region,
Wherein, adecFor from the maximum value of vehicle deceleration, if certain surrounding road user is located in potential risk region, this week
It encloses road user and is defined as potential risk road user；
Definition is centered on from vehicle mass center, safe early warning distance Lp-riskIt is safety zone for the region except the border circular areas of radius
Domain, if certain surrounding road user is located at except potential risk region, or when in from except the communication range of vehicle, this week
It encloses road user and is defined as green route user.
3. automatic driving vehicle dynamic behaviour decision-making technique as claimed in claim 2, which is characterized in that step S2 includes following
Step:
Step S21, for the surrounding road user in risk zones and potential risk region, calculation risk degree C, wherein first
It is calculated for the surrounding road user in risk zones, then for the surrounding road user meter in potential risk region
It calculates,
It is clashed between current state and surrounding road user's state locating for risk C characterization automatic driving vehicle general
Rate,
T is the estimated collision time of estimation, if had around two or more in risk zones and potential risk region
Road user, t are the estimated collision time that estimated two or more estimations collided occur with each surrounding road user
Minimum value, if there is no surrounding road user in risk zones and potential risk region, t be greater than tcSetting value,
tcFor the crash time avoided collision, for the constant of setting,
Work as C=0, then determines that there is no traffic conflicts in this state, determine risk magnitude friskIt is zero, and goes to step S23；
Work as C=1, then determine there is potential traffic conflict in this state, go to step S22,
Step S22, calculation risk metric frisk,
Step S23, according to risk magnitude friskMovement selection is carried out, determines possible action set.
4. automatic driving vehicle dynamic behaviour decision-making technique as claimed in claim 3, which is characterized in that the estimated collision of estimation
Time, t was calculated as follows,
T=min { TTC, PET, TTB }
Wherein,
XiBe from truck position,
XjIt is the position for the surrounding road user being followed by,
viFor from vehicle present speed,
vjFor his vehicle present speed,
LiFor from vehicle commander's degree,
PET is the time t for entering conflict point from vehicleiThe time t of this conflict point is reached to another surrounding road userjTime
Between difference,
PET=t=| ti-tj|
TTB is suitable for from Che Hou, his Che Qian scene for assessing forward region,
XiBe from truck position,
XjIt is his truck position being followed by,
viFor from vehicle present speed,
LiFor from vehicle commander's degree.
5. automatic driving vehicle dynamic behaviour decision-making technique as claimed in claim 3, which is characterized in that the estimated collision of estimation
Time, t was calculated in the following manner,
When scene can be distinguished, the estimated collision time t of the estimation for the scene is only calculated, wherein for straight way follow the bus
Scene, t=TTC；For intersection scene, t=PET；For colliding scene, t=TTB from Che Hou forward direction；
When scene complexity, t=min { TTC, PET, TTB },
Wherein,
XiBe from truck position,
XjIt is the position for the surrounding road user being followed by,
viFor from vehicle present speed,
vjFor his vehicle present speed,
LiFor from vehicle commander's degree,
PET is the time t for entering conflict point from vehicleiThe time t of this conflict point is reached to another surrounding road userjTime
Between difference,
PET=t=| ti-tj|
TTB is suitable for from Che Hou, his Che Qian scene for assessing forward region,
XiBe from truck position,
XjIt is his truck position being followed by,
viFor from vehicle present speed,
LiFor from vehicle commander's degree.
6. automatic driving vehicle dynamic behaviour decision-making technique as claimed in claim 3, which is characterized in that in step s 2, with
Following formula calculation risk metric frisk,
Or
7. the automatic driving vehicle dynamic behaviour decision-making technique as described in claim 1-6, which is characterized in that the of step S3
Do not include the decision attribute of any influence safety in two-stage behaviour decision making, but considers efficient soft-constraint function fe, it is comfortable soft
Constraint function fcWith traffic flow soft-constraint function ftCarry out optimizing decision.
8. automatic driving vehicle real-time track planing method as claimed in claim 7, which is characterized in that efficient soft-constraint function
feIs defined as:
Wherein, t0Initially to set out the moment from vehicle, tfTo arrive at the destination the moment from vehicle, v (t) is from vehicle speed.
9. automatic driving vehicle real-time track planing method as claimed in claim 7, which is characterized in that comfortable soft-constraint function
fcIs defined as:
Wherein, a is from vehicle acceleration, alatFor transverse acceleration, alonFor longitudinal acceleration.
10. automatic driving vehicle real-time track planing method as claimed in claim 7, which is characterized in that traffic flow soft-constraint
Function ftIs defined as:
minft=α (vave-vder)2+β(dave-dder)2
Wherein,
vaveFor the velocity levels of the preceding periphery traffic flow of decision,
vderFor the desired velocity levels of decision rear perimeter edge traffic flow,
daveFor the average following distance of the preceding periphery traffic flow of decision,
dderFor the desired average following distance of decision rear perimeter edge traffic flow,
α, β are weight coefficient, and both greater than 0 less than 1.
11. automatic driving vehicle real-time track planing method as claimed in claim 7, which is characterized in that the second of step S3
In stage behaviour decision making, it is as follows to define cost function J:
w1、w2、w3For weight coefficient, both greater than 0 less than 1, and w1+w2+w3=1,
Wherein, fe0、fc0、ft0Respectively indicate safe and efficient, comfortable, friendship after assuming to continue to execute from vehicle according to state before decision
Through-flow function.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910774762.XA CN110488802B (en) | 2019-08-21 | 2019-08-21 | Decision-making method for dynamic behaviors of automatic driving vehicle in internet environment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910774762.XA CN110488802B (en) | 2019-08-21 | 2019-08-21 | Decision-making method for dynamic behaviors of automatic driving vehicle in internet environment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110488802A true CN110488802A (en) | 2019-11-22 |
| CN110488802B CN110488802B (en) | 2020-05-12 |
Family
ID=68552636
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910774762.XA Active CN110488802B (en) | 2019-08-21 | 2019-08-21 | Decision-making method for dynamic behaviors of automatic driving vehicle in internet environment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110488802B (en) |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111615053A (en) * | 2020-05-22 | 2020-09-01 | 英华达（上海）科技有限公司 | Internet of vehicles system and method for dynamically marking risk areas |
| CN111775952A (en) * | 2019-11-26 | 2020-10-16 | 北京京东乾石科技有限公司 | Automatic driving method, device and storage medium |
| CN112017438A (en) * | 2020-10-16 | 2020-12-01 | 宁波均联智行科技有限公司 | Driving decision generation method and system |
| CN112150046A (en) * | 2020-11-06 | 2020-12-29 | 上海交通大学 | Road intersection safety risk index calculation method |
| CN112216146A (en) * | 2020-09-16 | 2021-01-12 | 浙江吉利控股集团有限公司 | Vehicle risk early warning method, device and equipment and vehicle |
| CN112256033A (en) * | 2020-10-30 | 2021-01-22 | 安徽江淮汽车集团股份有限公司 | Driving state recognition method, device, storage medium and device |
| CN112550284A (en) * | 2020-11-16 | 2021-03-26 | 东南大学 | Self-adaptive cruise driving risk assessment method under vehicle-road coupling effect |
| CN112885113A (en) * | 2019-11-29 | 2021-06-01 | 阿尔斯通运输科技公司 | Driving assistance method for public transport vehicle |
| CN112947390A (en) * | 2021-04-02 | 2021-06-11 | 清华大学 | Intelligent networking automobile safety control method and system based on environmental risk assessment |
| CN113022488A (en) * | 2019-12-25 | 2021-06-25 | 北京宝沃汽车股份有限公司 | Vehicle and control method and device thereof |
| CN113129624A (en) * | 2019-12-30 | 2021-07-16 | 百度（美国）有限责任公司 | Fastest lane determining algorithm under traffic jam condition |
| CN113177663A (en) * | 2021-05-20 | 2021-07-27 | 启迪云控(上海)汽车科技有限公司 | Method and system for processing intelligent network connection application scene |
| CN113433936A (en) * | 2021-06-02 | 2021-09-24 | 北京迈格威科技有限公司 | Mobile equipment obstacle-avoiding method and device, mobile equipment and storage medium |
| CN113548049A (en) * | 2021-07-27 | 2021-10-26 | 武汉理工大学 | Intelligent vehicle driving behavior decision method and system based on finite-state machine |
| CN113682299A (en) * | 2020-05-19 | 2021-11-23 | 广州汽车集团股份有限公司 | Vehicle forward collision early warning method and device |
| CN113954824A (en) * | 2020-07-02 | 2022-01-21 | 现代摩比斯株式会社 | Automatic driving system, method and readable medium for preventing collision with cut-in vehicle |
| CN114056332A (en) * | 2022-01-14 | 2022-02-18 | 清华大学 | Intelligent automobile following decision and control method based on cognitive risk balance |
| CN115331436A (en) * | 2022-08-04 | 2022-11-11 | 西安理工大学 | Road section automatic driving automobile movement planning method considering pedestrian safety |
| CN115620536A (en) * | 2022-10-18 | 2023-01-17 | 北京航空航天大学 | Method for improving crossing traffic efficiency based on danger degree in automatic driving environment |
| CN116978257A (en) * | 2023-08-11 | 2023-10-31 | 山东大学 | Highway confluence region conflict early warning method and system based on track prediction |
| US11878685B2 (en) | 2020-12-22 | 2024-01-23 | Beijing Baidu Netcom Science Technology Co., Ltd. | Vehicle control method and apparatus, electronic device and self-driving vehicle |
| CN117877272A (en) * | 2024-03-11 | 2024-04-12 | 中国市政工程华北设计研究总院有限公司 | Intersection safety evaluation method based on unmanned aerial vehicle detection |
| CN116978257B (en) * | 2023-08-11 | 2024-05-24 | 山东大学 | Highway confluence region conflict early warning method and system based on track prediction |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103754224A (en) * | 2014-01-24 | 2014-04-30 | 清华大学 | Vehicle multi-target coordinating lane changing assisting adaptive cruise control method |
| CN105291965A (en) * | 2015-11-04 | 2016-02-03 | 石海水 | Automobile anti-collision method and device |
| CN109375632A (en) * | 2018-12-17 | 2019-02-22 | 清华大学 | Automatic driving vehicle real-time track planing method |
| CN109709956A (en) * | 2018-12-26 | 2019-05-03 | 同济大学 | A kind of automatic driving vehicle speed control multiple-objection optimization with algorithm of speeding |
| CN109910879A (en) * | 2019-04-03 | 2019-06-21 | 大连理工大学 | A kind of vehicle safety anticollision control method of combination safe distance and collision time |
| CN109941211A (en) * | 2019-03-22 | 2019-06-28 | 清华大学 | A kind of Intelligent Vehicle Driving System structure common type framework and construction method |
| CN110020504A (en) * | 2019-04-23 | 2019-07-16 | 吉林大学 | Unmanned vehicle running environment complexity quantization method based on rear intrusion and collision time |
| CN110136254A (en) * | 2019-06-13 | 2019-08-16 | 吉林大学 | Driving assistance information display methods based on dynamic probability driving map |
-
2019
- 2019-08-21 CN CN201910774762.XA patent/CN110488802B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103754224A (en) * | 2014-01-24 | 2014-04-30 | 清华大学 | Vehicle multi-target coordinating lane changing assisting adaptive cruise control method |
| CN105291965A (en) * | 2015-11-04 | 2016-02-03 | 石海水 | Automobile anti-collision method and device |
| CN109375632A (en) * | 2018-12-17 | 2019-02-22 | 清华大学 | Automatic driving vehicle real-time track planing method |
| CN109709956A (en) * | 2018-12-26 | 2019-05-03 | 同济大学 | A kind of automatic driving vehicle speed control multiple-objection optimization with algorithm of speeding |
| CN109941211A (en) * | 2019-03-22 | 2019-06-28 | 清华大学 | A kind of Intelligent Vehicle Driving System structure common type framework and construction method |
| CN109910879A (en) * | 2019-04-03 | 2019-06-21 | 大连理工大学 | A kind of vehicle safety anticollision control method of combination safe distance and collision time |
| CN110020504A (en) * | 2019-04-23 | 2019-07-16 | 吉林大学 | Unmanned vehicle running environment complexity quantization method based on rear intrusion and collision time |
| CN110136254A (en) * | 2019-06-13 | 2019-08-16 | 吉林大学 | Driving assistance information display methods based on dynamic probability driving map |
Non-Patent Citations (2)
| Title |
|---|
| XUNJIA ZHENG 等: "Behavioral decision-making model of the intelligent vehicle based on driving risk assessment", 《IEEE》 * |
| 宋威龙: "城区动态环境下智能车辆行为决策研究", 《中国博士学位论文全文数据库》 * |
Cited By (34)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111775952A (en) * | 2019-11-26 | 2020-10-16 | 北京京东乾石科技有限公司 | Automatic driving method, device and storage medium |
| CN112885113A (en) * | 2019-11-29 | 2021-06-01 | 阿尔斯通运输科技公司 | Driving assistance method for public transport vehicle |
| CN113022488A (en) * | 2019-12-25 | 2021-06-25 | 北京宝沃汽车股份有限公司 | Vehicle and control method and device thereof |
| CN113129624A (en) * | 2019-12-30 | 2021-07-16 | 百度（美国）有限责任公司 | Fastest lane determining algorithm under traffic jam condition |
| US11880201B2 (en) | 2019-12-30 | 2024-01-23 | Baidu Usa Llc | Fastest lane determination algorithm under traffic jam |
| CN113682299A (en) * | 2020-05-19 | 2021-11-23 | 广州汽车集团股份有限公司 | Vehicle forward collision early warning method and device |
| CN111615053A (en) * | 2020-05-22 | 2020-09-01 | 英华达（上海）科技有限公司 | Internet of vehicles system and method for dynamically marking risk areas |
| US11731616B2 (en) | 2020-07-02 | 2023-08-22 | Hyundai Mobis Co., Ltd. | Autonomous driving system for preventing collision of cut-in vehicle and autonomous driving method thereof |
| CN113954824A (en) * | 2020-07-02 | 2022-01-21 | 现代摩比斯株式会社 | Automatic driving system, method and readable medium for preventing collision with cut-in vehicle |
| CN113954824B (en) * | 2020-07-02 | 2024-03-01 | 现代摩比斯株式会社 | Automatic driving system, method and readable medium for preventing collision with cut-in vehicle |
| CN112216146A (en) * | 2020-09-16 | 2021-01-12 | 浙江吉利控股集团有限公司 | Vehicle risk early warning method, device and equipment and vehicle |
| CN112017438B (en) * | 2020-10-16 | 2021-08-27 | 宁波均联智行科技股份有限公司 | Driving decision generation method and system |
| CN112017438A (en) * | 2020-10-16 | 2020-12-01 | 宁波均联智行科技有限公司 | Driving decision generation method and system |
| CN112256033A (en) * | 2020-10-30 | 2021-01-22 | 安徽江淮汽车集团股份有限公司 | Driving state recognition method, device, storage medium and device |
| CN112256033B (en) * | 2020-10-30 | 2022-12-20 | 安徽江淮汽车集团股份有限公司 | Driving state recognition method, device, storage medium and device |
| CN112150046A (en) * | 2020-11-06 | 2020-12-29 | 上海交通大学 | Road intersection safety risk index calculation method |
| CN112550284A (en) * | 2020-11-16 | 2021-03-26 | 东南大学 | Self-adaptive cruise driving risk assessment method under vehicle-road coupling effect |
| CN112550284B (en) * | 2020-11-16 | 2022-03-22 | 东南大学 | Self-adaptive cruise driving risk assessment method under vehicle-road coupling effect |
| US11878685B2 (en) | 2020-12-22 | 2024-01-23 | Beijing Baidu Netcom Science Technology Co., Ltd. | Vehicle control method and apparatus, electronic device and self-driving vehicle |
| CN112947390A (en) * | 2021-04-02 | 2021-06-11 | 清华大学 | Intelligent networking automobile safety control method and system based on environmental risk assessment |
| US11518409B2 (en) | 2021-04-02 | 2022-12-06 | Tsinghua University | Safety control method and system based on environmental risk assessment for intelligent connected vehicle |
| CN113177663B (en) * | 2021-05-20 | 2023-11-24 | 云控智行(上海)汽车科技有限公司 | Processing method and system of intelligent network application scene |
| CN113177663A (en) * | 2021-05-20 | 2021-07-27 | 启迪云控(上海)汽车科技有限公司 | Method and system for processing intelligent network connection application scene |
| CN113433936A (en) * | 2021-06-02 | 2021-09-24 | 北京迈格威科技有限公司 | Mobile equipment obstacle-avoiding method and device, mobile equipment and storage medium |
| CN113548049B (en) * | 2021-07-27 | 2022-05-31 | 武汉理工大学 | Intelligent vehicle driving behavior decision method and system based on finite-state machine |
| CN113548049A (en) * | 2021-07-27 | 2021-10-26 | 武汉理工大学 | Intelligent vehicle driving behavior decision method and system based on finite-state machine |
| CN114056332B (en) * | 2022-01-14 | 2022-04-12 | 清华大学 | Intelligent automobile following decision and control method based on cognitive risk balance |
| CN114056332A (en) * | 2022-01-14 | 2022-02-18 | 清华大学 | Intelligent automobile following decision and control method based on cognitive risk balance |
| CN115331436A (en) * | 2022-08-04 | 2022-11-11 | 西安理工大学 | Road section automatic driving automobile movement planning method considering pedestrian safety |
| CN115331436B (en) * | 2022-08-04 | 2024-04-30 | 西安理工大学 | Road section automatic driving automobile motion planning method considering pedestrian safety |
| CN115620536A (en) * | 2022-10-18 | 2023-01-17 | 北京航空航天大学 | Method for improving crossing traffic efficiency based on danger degree in automatic driving environment |
| CN116978257A (en) * | 2023-08-11 | 2023-10-31 | 山东大学 | Highway confluence region conflict early warning method and system based on track prediction |
| CN116978257B (en) * | 2023-08-11 | 2024-05-24 | 山东大学 | Highway confluence region conflict early warning method and system based on track prediction |
| CN117877272A (en) * | 2024-03-11 | 2024-04-12 | 中国市政工程华北设计研究总院有限公司 | Intersection safety evaluation method based on unmanned aerial vehicle detection |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110488802B (en) | 2020-05-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110488802A (en) | A kind of automatic driving vehicle dynamic behaviour decision-making technique netted under connection environment | |
| Ye et al. | Prediction-based eco-approach and departure at signalized intersections with speed forecasting on preceding vehicles | |
| KR102127741B1 (en) | Systems and methods for navigation of vehicles | |
| EP3842303B1 (en) | Systems and methods for navigating with safe distances | |
| CN108313054B (en) | The autonomous lane-change decision-making technique of automatic Pilot and device and automatic driving vehicle | |
| CN110298122A (en) | Automatic driving vehicle urban intersection left-hand rotation decision-making technique based on conflict resolution | |
| CN110992695B (en) | Vehicle urban intersection traffic decision multi-objective optimization method based on conflict resolution | |
| KR20200049888A (en) | Navigational system with imposed liability constraints | |
| CN107792065A (en) | The method of road vehicle trajectory planning | |
| CN108225364A (en) | A kind of pilotless automobile driving task decision system and method | |
| CN109131326A (en) | A kind of adaptive learning algorithms device and its working method with lane-change miscellaneous function | |
| CN107346612A (en) | A kind of vehicle collision avoidance method and system based on car networking | |
| US20230286536A1 (en) | Systems and methods for evaluating domain-specific navigation system capabilities | |
| WO2022053026A1 (en) | Automatic driving meeting scene processing method and apparatus, vehicle, and storage medium | |
| CN113895459B (en) | Barrier screening method and system | |
| CN114861514A (en) | Planning method and device for vehicle driving scheme and storage medium | |
| Pan et al. | Research on the behavior decision of connected and autonomous vehicle at the unsignalized intersection | |
| US20230322208A1 (en) | Steering limiters for vehicle navigation | |
| Guo et al. | Toward human-like behavior generation in urban environment based on Markov decision process with hybrid potential maps | |
| CN113291318A (en) | Unmanned vehicle blind area turning planning method based on partially observable Markov model | |
| Tomar et al. | Neural network based lane change trajectory predictions for collision prevention | |
| Guo et al. | Self-defensive coordinated maneuvering of an intelligent vehicle platoon in mixed traffic | |
| Lu et al. | Decision-making method of autonomous vehicles in urban environments considering traffic laws | |
| CN114228707A (en) | Anti-collision method and system for unmanned vehicle | |
| Yue et al. | An augmentation function for active pedestrian safety system based on crash risk evaluation |
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 |