CN110298131A - Automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment - Google Patents
Automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment Download PDFInfo
- Publication number
- CN110298131A CN110298131A CN201910603412.7A CN201910603412A CN110298131A CN 110298131 A CN110298131 A CN 110298131A CN 201910603412 A CN201910603412 A CN 201910603412A CN 110298131 A CN110298131 A CN 110298131A
- Authority
- CN
- China
- Prior art keywords
- lane
- change
- vehicle
- strategy
- under
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/29—Graphical models, e.g. Bayesian networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
Abstract
The invention discloses automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment, the multistep dynamic game lane-change model that automatic driving vehicle LV and the mankind under mixing driving environment drive vehicle RV is established, and devises a set of nested game playing algorithm for automatic driving vehicle.The present invention has initially set up multistep dynamic game frame, both sides' vehicle can all determine next action according to other side's policy selection, and define the potential conflict point of lane-change vehicle and target lane rear car, and according to vehicle initial information, the tactful start-stop condition criterion that game is formulated with game step number, then two cars carry out dynamic game using respective strategy and acceleration selection method, until meeting lane-change termination condition, to make automatic driving vehicle in the case where being unsatisfactory for the condition of direct lane-change, also lane-change space can be manufactured by driving vehicle game with the mankind, and realize safe lane-change.
Description
Technical field
The present invention relates to automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment.
Background technique
In recent years, automatic Pilot technology, which receives in world wide, more and more pays close attention to, it is considered to be can overturn biography
It unites an emerging technology of traffic, to solving to play an important role in traffic safety and traffic jam issue.However, from current
Road on all mankind drive vehicle and be transitioned into entirely automatic Pilot, a necessarily long-term process, in this mistake
Journey, traffic environment are the mixed environment of mankind's driving and automatic driving vehicle.Automatic driving vehicle not only needs to complete complete
Safe and efficient traveling under automatic Pilot environment, it is also desirable to adapt to mixing driving environment.
According to the research of existing automatic Pilot lane-change decision, the content of lane-change decision specifically includes that lane-changing intention generates
With the assessment of lane-change environment.Vehicle is influenced by front truck/barrier in the process of moving, or have to enter into/when going out ring road,
Vehicle generates lane-changing intention, determines whether vehicle needs to be implemented lane-change operation.Lane-change environmental assessment is that vehicle is changed in determining needs
Behind road, for the safety and high efficiency for ensuring lane-change, vehicle lane-changing environment is assessed, so that it is determined that can vehicle lane-change.
Wei et al. thinks that vehicle generates lane-changing intention when most seeking driving efficiency, based on prediction and cost function algorithm from vehicle follow gallop,
In choosing lane and scene, the highest control strategy of efficiency is selected.The threshold value in left and right lane does not have in decision, does not haggle over directly
Tactful cost, vehicle preferentially select left-hand rotation lane-change or overtake other vehicles.Use the Boss mould in DARPA Urban Challenge 2007
Quasi- device is tested in mandatory lane-change scene, and test result shows selection and lane-change result table of the decision making algorithm in lane
Now preferably.Equally, Habenicht et al. generates lane-changing intention when pursuing driving efficiency, and the cost letter based on fuzzy logic
Counting method establishes vehicle lane-changing auxiliary system, to behaviour such as not changing Lane, acceleration lane-change, deceleration lane-change and the direct lane-changes of non-speed change
Make carry out decision, and lane-change time, lane-change direction and required plus/minus speed are provided.System is also provided with can not lane-change success
Module propose to abandon the warning of lane-change when system detection to lane-change, which exists, collides.System structure is described in detail in text
And man-machine interface, but lack the assessment of exchange road scene and the background introduction of decision making algorithm.In Kim et al. research, vehicle exists
Lane-changing intention is generated when avoidance, vehicle realizes avoidance process using emergency braking and lane-change, and vehicle cannot achieve avoidance in deceleration
When, lane-changing intention is generated to realize avoidance, and then scene is assessed again, determines that adjacent lane can provide one comfortably
Lane-change environment when, set transverse acceleration as constant, the receptible maximum lateral acceleration of the mankind used to judge as comfort
Standard, then by execute lane-change realize avoidance.In addition, Sivaraman and Trivedi have studied in city vehicle lane-changing and
Lane merges the lane-changing intention generated under environment, drives figure to indicate road environment using probability, calculates vehicle using dynamic probability figure
Thus the cost of mandatory lane-change solves vehicle lane-changing decision problem.Jula et al. has formulated peace when vehicle lane-changing/merging
Full rule, it is assumed that surrounding vehicles constant airspeed, lane-change vehicle are travelled with constant speed or constant acceleration, calculate lane-change vehicle with
Longitudinal minimum safe spacing that surrounding vehicles do not collide.Using minimum safe distance as the whether feasible judgement of vehicle lane-changing
Standard, but the model only considered the case where vehicle does not collide, for the emergency cases such as emergency brake discomfort.
Kanaris et al. solves the problems, such as that Jula et al. is unconformable to emergency, proposes a kind of completely new analysis standard,
Safety regulation be when urgent formulate occurs for a vehicle any in lane-change vehicle and surrounding vehicles, remaining vehicle can brake without
It collides.Lane-change vehicle is adjusted self-acceleration and speed, only when all vehicle spacing all meet minimum peace
Full distance, lane-change environment just execute lane-change operation after meeting safety regulation.Wan et al. is based on identifiable Environmental Studies vehicle certainly
Dynamic lane change decision rule algorithm estimates prediction vehicle lane-changing speed and relative position using algorithm, and rule is in addition to requiring entirely to change
Road process vehicle must keep suitable safe distance with surrounding vehicles, and the lane-change transverse and longitudinal acceleration of estimation is also required to meet vehicle
Performance requirement.Furda and Vlacic carries out Real-time Decision using Petri network and Multi-objective Decision Model, by lane-change decision point
Solution is two continuous stages, and the first stage determines whether lane-change is safe using Petri, if it violates the traffic regulations, second
Stage recycles the comfort and efficiency of multiobjective decision-making raising vehicle, and two step cycles execute, and obtains meeting safety, relax
The Driving Decision-making of adaptive and efficiency.Chen et al. has studied the automatic driving vehicle decision under complicated urban environment, lane-change decision
Same to use, rule determines deletes inexecutable operation first, such as vehicle is deleted to the right when on the lane of the rightmost side
The operation of lane-change；Secondly, deleting the decision that do not observe traffic rules and regulations；Then, consider the efficiency and safety of traveling, select optimal
Driving Decision-making.Wherein, vehicle lane-changing safety is characterized by the spacing of lane-change vehicle and surrounding other vehicles, lane-change efficiency
It is characterized by the time of arrival purpose, so that decision goes out safety and meets the lane-change operation of driver characteristics.Nie et al. is automatic
It drives in vehicle lane-changing tactics research, proposes vehicle lane-changing preparation process, but without analyzing lane-change preparation process and effect,
The free lane-change decision process of automatic driving vehicle is analyzed in text using NGSIM data track of vehicle collection.It is horizontal to introduce vehicle lane-changing
The starting point that vehicle lane-changing is identified to the threshold value of speed establishes autonomous lane-change decision model, root based on support vector machine classifier
It gets on the bus the gap between the vehicle of position and speed assessment front and back according to target lane, determines vehicle lane-changing execution point and execute the time.
The system that Mccall et al. establishes driver intention deduction based on sparse Bayesian, by vehicle-state, environmental variance and driver
State is inputted as system, the calculatings vehicle to be changed one's intention using sparse Bayesian to lane whether the probability of lane-change, thus reality
Existing vehicle lane-changing decision operation.
The present invention is concerned with lane-change problem of the automatic driving vehicle in the case where mixing driving environment.Lane-change is as automatic Pilot
The basic driving behavior of vehicle has vital meaning to driving safety.How reasonable vehicle lane-changing algorithm is designed,
Guarantee automatic driving vehicle in actual traffic, the safety and efficiency of lane-change, are one ten especially under mixing driving environment
Divide challenging problem.
The lane-change scene of automatic driving vehicle is as shown in Figure 1, during lane-change, and the vehicle being related to has four, including certainly
It is dynamic to drive lane-change vehicle LV (Lane-changing vehicle), current lane front truck PV (Preceding vehicle), mesh
It marks lane front truck FV (Front vehicle), target lane rear car RV (Rear vehicle).It is often met in actual lane-change
To such case: since safe lane-change distance is insufficient, in order to realize lane-change, needing target lane rear car RV to slow down is that LV creates peace
Full lane-change space.In this case, LV can it is expected RV evacuation to obtain lane-change income, however RV deceleration can then damage itself
Interests, so RV is reluctant to selection evacuation.At this point, the interests of LV and the interests of RV have conflict, the two needs to carry out a system
The game of column could finally determine behavioral strategy.In the case where mixing driving environment, although LV is automatic driving vehicle, when RV is
When the mankind drive vehicle, there is still a need for the policy selections real-time dynamicly for RV to carry out the Developing Tactics of itself by LV, to mention
High lane-change safety and success rate.In this case, one exclusively for automatic driving vehicle design lane-change control algolithm just
It is very necessary.
Although the research can not drive vehicle according to the mankind currently, having there is relevant research for this problem
Practical strategies selection, pointedly adjust itself strategy, and simply by behavior of the aggressiveness to RV of assessment RV vehicle
Make the assumption that, estimate to aggressiveness there are in the case where error, may result in the case where RV is not avoided, LV into
Row lane-change, to cause security risk.
Summary of the invention
In order to overcome the disadvantages mentioned above of the prior art, the invention proposes automatic Pilot lane-changes under a kind of mixing driving environment
Decision model method for building up considers the policy selection between vehicle and moves using game theory-kinematics coupling model method
State gambling process establishes the multistep dynamic game lane-change that automatic driving vehicle LV and the mankind under mixing driving environment drive vehicle RV
Model, and a set of nested game playing algorithm is devised for automatic driving vehicle.The present invention has initially set up multistep dynamic game frame,
Both sides' vehicle can all determine next action according to other side's policy selection, and define lane-change vehicle and target lane rear car
Potential conflict point, and the start-stop condition criterion of game is formulated according to vehicle initial information, strategy and game step number, secondly consider
The income of involved vehicle during automatic Pilot lane-change, including three speed income, security gain and comfortable income aspects, so
Afterwards, LV deduces out the corresponding financial value of vehicle in automatic Pilot nesting game lane-change frame using kinematic method, solves LV
Corresponding longitudinal acceleration under the strategy used and the strategy；RV is then compared the income under current Different Strategies, choosing
Select corresponding longitudinal acceleration under optimal policy and the strategy.Finally, two cars use respective strategy and acceleration selecting party
Method carries out dynamic game, until meeting lane-change termination condition, so that automatic driving vehicle be made to be unsatisfactory for direct lane-change
In the case where condition, also lane-change space can be manufactured by driving vehicle game with the mankind, and realize safe lane-change.
The technical solution adopted by the present invention to solve the technical problems is: automatic Pilot lane-change under a kind of mixing driving environment
Decision model method for building up, includes the following steps:
Step 1: LV generates lane-changing intention；
Step 2: judging whether that meeting lane-change game starts condition: if it is, entering step three；If it is not, then judging LV
Whether lane-change condition is met: if so, nine are entered step, if it is not, then waiting generate lane-changing intention next time, subsequently into step
One；
Step 3: LV building Dynamic Game Model and two vehicle incomes of calculating；
Step 4: finding out the optimal policy that LV should be selected；
Step 5: judging whether to meet lane-change game termination condition: if so, ten are entered step, if it is not, then entering step
Six；
Step 6: LV opens turn signal and exploratory lateral shift；
Step 7: LV selects lane-change acceleration；
Step 8: judging whether RV selects to avoid: if it is not, then return step three；If so, entering step nine；
Step 9: LV starts lane-change, until lane-change terminates；
Step 10: LV continues in former lane with speeding, this lane-changing intention terminates.
Compared with prior art, the positive effect of the present invention is:
1) automatic Pilot lane-change model can make LV actively create lane-change space by game, in general automatic Pilot model
It can not carry out also being able to carry out safe lane-change behavior under the scene of lane-change.
2) automatic driving vehicle makes the policy selection of oneself in practical lane-change according to RV previous step strategy, and existing
Model is compared, and has higher safety.
3) thinking gambling process is removed from the angle of game both sides' vehicle respectively, lane-change game framework has better integrality
And dynamic.
4) acceleration of vehicle RV is driven to the mankind and policy selection is also modeled, and distinguished mankind's vehicle and oneself
The dynamic difference for driving vehicle in lane-change gambling process, can be more reasonable predict under the following mixed running environment, automatically
Drive the dynamic process of lane-change vehicle and the game of mankind's vehicle lane-changing.
Detailed description of the invention
Examples of the present invention will be described by way of reference to the accompanying drawings, in which:
Fig. 1 is lane-change scene；
Fig. 2 is automatic driving vehicle lane-change model framework；
Fig. 3 is that automatic Pilot and the mankind drive vehicle lane-changing gambling process；
Fig. 4 is potential conflict point；
Fig. 5 is automatic Pilot nesting game playing algorithm frame；
Fig. 6 is example solution procedure；
Fig. 7 is lane-change scene figure；
Fig. 8 is the time difference box figure that conflicts at the lane-change moment；
Fig. 9 is existing model lane-change schematic diagram；
Figure 10 is this model lane-change schematic diagram；
Figure 11 is simulation analysis figure.
Specific embodiment
Automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment, including following content:
1, model framework
The invention proposes the vehicle lane-changing models under the automatic mixed running environment of hand-, set for automatic Pilot lane-change vehicle
The mechanism of complete set is counted, to complete the lane-change gambling process with mankind's vehicle.As shown in Fig. 2, firstly, lane-change vehicle LV can be produced
Raw lane-changing intention；Then, lane-change vehicle LV judges two according to the position and speed information of itself and target lane rear car RV
It whether there is Game Relationship between vehicle, in the case where being unsatisfactory for game and starting condition, if LV can satisfy the condition of lane-change,
LV can execute lane-change；Secondly, then LV starts the lane-change game of automatic Pilot nesting for meeting in the case that game starts condition
Algorithm constructs the Dynamic Game Model frame of LV and two vehicle of RV first, is advised by the generality in the driving process of mankind's vehicle
Rule simulates entire gambling process, and predicts the income in LV and RV game each stage, solves the current generation strategy of LV；Most
Afterwards, according to the policy selection of RV, LV will carry out multiple dynamic game with RV, until meeting game termination condition.
2, lane-change game theory analysis
(1) multistep dynamic lane-change game
The present invention passes through the characteristic to game and automatic driving vehicle during vehicle lane-changing under mankind's driving environment
It is analyzed, building hand mixes under driving environment automatically, and automatic driving vehicle and the mankind drive complete lane-change dynamic between vehicle
The model of game is able to reflect the following hand and mixes mankind's driving vehicle and automatic driving vehicle progress lane-change under driving environment automatically
The dynamic process of game.
The analysis of opponent of the present invention-automatic driving vehicle lane-change gambling process is as shown in figure 3, lane-change vehicle LV is certainly in figure
Dynamic to drive vehicle, target lane rear car RV is that the mankind drive vehicle, and before lane-change starts, lane-change vehicle LV can generate lane-change meaning
Figure, and determine target lane and target gap.Next in the case where lane-change vehicle LV judges that lane-change game condition meets,
Game can be carried out with target lane rear car RV, lane-change vehicle LV will do it policy selection, if selection lane-change, turns to using opening
Lamp and lateral displacement are soundd out, if not lane-change, target lane rear car RV continues with the target lane front truck FV that speeds, game knot
Beam.It then, can be to the receipts of the next strategy of itself in the case that target lane rear car RV observes LV selection lane-change strategy
Benefit is compared, and carries out policy selection.Likewise, lane-change vehicle LV is observing target lane rear car RV selection avoidance strategy
In the case where, then lane-change is selected, is reached an agreement between two vehicles, lane-change vehicle LV can carry out cooperation lane-change, and game terminates.If target
Lane rear car RV selects not avoidance strategy, then two vehicle games continue.After lane-change vehicle LV game step number reaches setting quantity, this
When for automatic driving vehicle, then carry out game and can then have biggish risk, thus lane-change vehicle LV actively exit it is rich
It plays chess, i.e., when target lane rear car selects not avoidance strategy again, lane-change vehicle LV directly selects not lane-change strategy.
(2) potential conflict point
The generation of game in lane-change decision is since there are conflicting on space between two vehicles to vehicle LV with RV.Such as Fig. 4
Shown, traveling curve when LV lane-change can cross with the RV travel route for continuing straight trip in the certain point on target lane, by this
Point is defined as the potential conflict point of two vehicles.Assuming that LV carries out lane-change after generating lane-changing intention at once, and RV is with present speed
Move forward with the acceleration of selection, then by two vehicles reach potential conflict point time difference be defined as conflict the time difference (TDTC,
Time Difference to Collision), the conflict time difference is smaller, it is meant that the lane-change safety of two vehicles is lower.When arriving
When reaching certain threshold value up to time difference of potential conflict point, two vehicles just need to carry out game, thus determine final traveling strategy,
For example whether LV will carry out lane-change, whether RV will be avoided.So, it is necessary first to calculate the potential conflict point of LV and RV.
The coordinate for enabling potential conflict point is (xc,yc), wherein ycCalculation formula it is as follows:
yc=ye-wcar (1)
Wherein, yeFor the ordinate of lane-change geometric locus terminal, wcarIndicate vehicle width.
It is travelled assuming that being along lane center under vehicle normal condition, then yeIt can directly acquire.So in ycIt is known
When, xcIt is that can be obtained by the lane-change geometric locus function of automatic Pilot vehicle.The lane-change geometric locus of automatic driving vehicle is most
Common polynomial curve, is shown below,
Y (x)=a0+a1x+a2x2+a3x3 (2)
Wherein, x and y is the horizontal and vertical position of vehicle LV headstock left end, a0, a1, a2, a3It is parameter to be determined.
A in formula (2)0, a1, a2, a3It is the unknown quantity of lane-change trajectory curve equation, is solved.The equation of locus is
It is designed based on lane-change vehicle, so coordinate system used in equation of locus is using LV headstock left end as seat herein
Mark origin.By deriving, after finding out the unknown parameter in above formula, the lane-change equation of locus of this paper is as follows,
Wherein, x and y is the horizontal and vertical position of vehicle LV headstock left end, xeAnd yeLane-change final on trajectory (ending)
Horizontal and vertical coordinate.
By ycValue bring trajectory curve equation (3) above into, then the abscissa x of available potential conflict pointc, finally may be used
To obtain the position of potential conflict point.
(3) lane-change game start-stop condition
1) lane-change game starts condition
During vehicle lane-changing, not every lane-change requires lane-change vehicle-to-target lane rear car and carries out game.
Lane-change game is carried out there are two condition, first is that lane-change vehicle and current lane front truck keep safe distance, and second is lane-change vehicle
With the time difference that conflicts of target lane rear car no more than some threshold value.First condition be in order to guarantee during lane-change with
The safety of PV conflicts if second condition stub directly carries out lane-change LV with RV presence.
The Calculation of Safety Distance of vehicle LV and PV are shown below using classical Gipps safe distance formula,
Wherein, GlFor the safe distance of lane-change vehicle LV and current lane front truck PV, xlIt (t) is LV, t moment lane-change vehicle
It sets, xpIt (t) is the position t moment current lane front truck PV, lpFor the length of wagon of current lane front truck PV, blSubtract for the maximum of LV
Speed, τlFor the reaction time of LV, vl(t) speed for being t moment LV, vp(t) speed for being t moment PV.
The conflict time difference for calculating LV and RV needs to calculate separately the time of two arrival potential conflict points.LV is from lane-change
Starting point reaches the arc length that the distance travelled in potential conflict point process is trajectory curve equation, and calculation formula is as follows,
Wherein, LlThe operating range of potential conflict point is changed lane to from current location for LV,
RV is from the distance that current location drives to potential conflict point,
Lr=xc+d (6)
Wherein, LrIndicate the operating range of RV point from current location to potential conflict, xcIndicate the left front angle point of lane-change vehicle away from
Fore-and-aft distance from potential conflict point, d indicate LV and the space headway of RV in the longitudinal direction.
So the game condition of LV and RV is as follows,
Wherein, TMIt indicates that the critical conflict time difference of game occurs for two vehicles, while being also that RV influences facing for LV lane-change safety
Boundary's point, vr(t) and vl(t) speed of the current t moment of RV and LV is respectively indicated.
2) lane-change game termination condition
It being driven in vehicle lane-changing gambling process in automatic driving vehicle and the mankind, both sides' vehicle can all have security risk,
And game number is more, caused by security risk it is also bigger, while also will affect the running efficiency of both sides' vehicle.So
The setting for needing game termination condition is also can not ignore the problem of.
In normal lane-change gambling process, a side has selected to give way only in both sides' vehicle, as lane-change vehicle LV is selected
Not lane-change is selected, target lane rear car RV was selected under the case where avoiding, then lane-change game at this time terminates, and in addition side's vehicle selects
Stand fast strategy.So one of game termination condition is exactly that game both sides have a side to make a concession, game is actively exited.
In addition, for automatic driving vehicle being controlled by computer, there is no the mankind to drive the so strong spirit of vehicle
Activity, so needing to be arranged stringent condition to guarantee that it is driving the safety in vehicle lane-changing gambling process with the mankind.This
In invention, for the lane-change termination condition of automatic Pilot lane-change Car design are as follows:
A. (number can make tune with different lane-change scenes when automatic Pilot lane-change vehicle LV game step number arrival n times
It is whole), if automatic driving vehicle actively exits game if mankind's driving vehicle RV is not avoided still.
B. during lane-change automatic driving vehicle LV and current lane front truck PV be unsatisfactory for the safety in formula (4) with
When condition of speeding, then automatic driving vehicle LV should actively abandon lane-change, find lane-change chance next time.
(4) lane-change game element and type
1) lane-change game element
In betting model, needing specific three fundamentals is game object, strategy and income respectively.It changes herein
The three elements of road game are as follows,
A. game object: automatic Pilot lane-change vehicle LV and manual drive target lane rear car RV.
B. tactful: there are two kinds of strategies of lane-change and non-lane-change in game, target lane rear car RV exists lane-change vehicle LV
It avoids and does not avoid two kinds of strategies.When LV selects lane-change strategy, it is meant that LV can be moved to target from current lane center line
Lane-change process is completed in lane；When LV selects not lane-change strategy, LV needs temporarily to abandon lane-change, continues to seek in current lane traveling
Look for next lane-change opportunity.RV selects its when avoidance strategy actively can create lane-change condition for LV, and cooperation LV completes lane-change；RV choosing
The lane-change behavior of LV can be then prevented when selecting not avoidance strategy.
C. income: LV and RV uses three proceeds indicatiors, respectively when carrying out lane-change game when evaluating strategy
For speed income, security gain and comfort income.
2) lane-change game types
Lane-change game types are determined according to the information and move sequence of game both sides.Firstly, since automatic Pilot vehicle
There can be more specific understanding to the income and strategy of its own and opponent vehicle.And mankind's vehicle can only recognize portion
Divide information, specific judgement can not be compared to other side's income, so the game is Incompletely information games；Secondly, hand from
Move in gambling process, there is no communications between vehicle, and after lane-change vehicle LV generates lane-changing intention, RV needs the plan to LV
Slightly judge the reaction for making next step again.Since driver and vehicle itself have the reaction time, both sides' decision exists first
Afterwards sequence, as long as and game can just be terminated in the case where wherein side's vehicle actively exits game, so the game is multistep
Dynamic game.
3, income analysis
During lane-change, lane-change vehicle LV carries out lane-change to pursue higher income, and the lane-change row of lane-change vehicle LV
To will affect the normally travel of target lane rear car RV, to keep the income of RV impaired, this is also lane-change vehicle LV and target carriage
The basic reason clashed between road rear car RV.So needing to lane-change vehicle LV during lane-change and target lane rear car RV
Income analyzed, the income of vehicle includes speed income, security gain and comfort income, concrete condition during lane-change
As follows.
(1) speed income
1) the speed income of lane-change vehicle LV
The reason of lane-change of lane-change vehicle LV is to reach target lane by lane-change to obtain higher speed income, so fast
Spending income is essential proceeds indicatior for LV.
A. not lane-change situation
When lane-change vehicle LV selects the not strategy of lane-change, the speed continued with speed current lane front truck PV, LV can be by
To the restriction of PV, so the difference of the speed of the speed and LV of PV is speed income of the LV not under lane-change strategy, it is shown below:
Wherein,Indicate speed income of the LV under not lane-change strategy, vp(t) speed of the PV in t moment, v are indicatedl(t)
Indicate LV in the speed of t moment.
B. lane-change situation
When lane-change vehicle LV selects lane-change strategy, LV expectation is target lane front truck FV with the vehicle speeded, so FV
Speed is the desired speed of LV, and the difference of the speed of the speed and LV of FV is the speed income under LV lane-change strategy, such as following formula institute
Show:
Wherein,Indicate speed income of the LV under lane-change strategy, vf(t) speed of the FV in t moment, v are indicatedl(t) table
Show LV in the speed of t moment.
2) the speed income of target lane rear car RV
Speed income is an important indicator of target lane rear car RV policy selection, and RV can take strategy that LV is prevented to change
Road, main cause are exactly that LV lane-change needs the deceleration of RV to cooperate, and disturb the normally travel of RV, and RV vehicle has evacuation and do not keep away
It allows two kinds of strategies, the speed income of RV under both strategies is analyzed below.
A. situation is avoided
Assuming that lane-change vehicle LV executes lane-change at once, RV has an expected evacuation speed under avoidance strategy, and RV uses this
A speed, which continues traveling, can just guarantee the safety of LV lane-change.The expected evacuation speed of RV should meet following formula,
Wherein,Indicate the expected evacuation speed of RV, TlIndicate that LV executes lane-change from current location and reaches potential conflict point
Running time.
The goal pace under RV avoidance strategy can be found out by formula (10):
Wherein TlDerivation process it is as follows, it is assumed that LV vehicle is in t moment with speed vl(t) and acceleration al(t) it is changed
Road, then LV prediction lane-change execution track on potential conflict point distance LlMeet following formula:
Wherein, vl(t) speed of t moment LV, a are indicatedl(t) acceleration of LV selection is indicated.Tl(t) it indicates that LV is reached to dive
The time required to conflict point.
It can show that LV reaches potential conflict point required time T by above formula derivationl(t) expression formula:
Then the difference of the expected evacuation speed of RV vehicle and RV speed is the speed income under RV avoidance strategy, such as following formula institute
Show:
Wherein,Indicate the speed income under RV avoidance strategy.
B. situation is not avoided
Under the strategy that rear car RV selection in target lane does not avoid, RV be will continue to target lane front truck FV traveling of speeding, then
The speed difference of target lane front truck FV and RV are speed income of the RV not under avoidance strategy, are shown below:
Wherein,Indicate speed income of the RV not under avoidance strategy.
(2) comfort income
For automatic driving vehicle, comfort is also a key factor in need of consideration, in vehicle traveling, vehicle
The amplitude of variation of acceleration, will affect the comfort of driving.It is retouched herein using the changing value of acceleration between adjacent step sizes
The comfort income for stating vehicle, is shown below:
Wherein,Indicate the comfort income of LV, alThe acceleration of (t- λ) expression long LV of previous step.
Identical as LV vehicle, the comfort income of RV vehicle is also indicated with the difference of the acceleration of adjacent step sizes, such as following formula
It is shown:
Wherein,Indicate the comfort income of RV, arThe acceleration of (t- λ) expression long RV of previous step.
(3) security gain
In the case where mixing driving environment, vehicle safety is vital factor, herein next will be respectively to LV and RV
Security gain during lane-change is described in detail.
The security gain of two cars can be calculated by the time difference that conflicts.Assuming that RV is with speed vr(t) and acceleration ar
(t) it moves forward, if RV reaches the running time of potential conflict point from current location as Tr(t), then RV drives to potential conflict
The distance L of pointrAre as follows:
T can be derived by formula (18)rExpression formula it is as follows,
Wherein, Tr(t) indicate that RV reaches the running time of potential conflict point from current location.
The time T of LV arrival potential conflict pointlIt (t) is formula (13) that RV vehicle conflicts time difference such as following formula with LV vehicle
It is shown:
Δ T=Tr(t)-Tl(t) (20)
Wherein, Δ T indicates conflicting the time difference for target lane rear car RV and lane-change vehicle LV.
As Δ T>0, indicate that LV reaches potential conflict point before RV, when Δ T<0, expression RV reaches potential before LV
Conflict point.
1) LV security gain
A.LV lane-change
LV selects security gain and conflict time difference Δ T-phase when lane-change strategy to close.In view of safety is to automatic Pilot vehicle
Importance, when carrying out the design of security gain function, value range without as speed income and comfort income,
It is limited in the range of -1 to 1.As Δ T < 0, LV reaches conflict point after RV, and lane-change behavior has very big safety hidden at this time
Suffer from, the security gain of LV in this case is regarded as herein bear it is infinite.As Δ T > 0, LV reaches conflict point punching before RV at this time
It is not linear relationship between prominent time difference and security gain, with the increase of conflict time difference Δ T, security gain can also increase
Add, but increased rate is gradually reduced；When the value of Δ T continues to increase to safety critical point TMWhen, security gain reaches maximum,
Financial value is 0, and security gain keeps stablizing later.Since conflict relationship, row is not present in two vehicles before lane-change game does not occur
The state of sailing be it is safe, so security gain will not be positive value, therefore the security gain maximum value of this paper is taken as 0.This process
It can be described with logarithmic function, then the expression formula of safety income is as follows,
Wherein,Indicate the security gain in LV when selecting lane-change strategy.
B LV not lane-change
LV is not in the case of lane-change, and conflict relationship is not present in two vehicles, and without security risk, the safety of LV not will receive
It influences, so the security gain of LV still keeps maximum value.
Wherein,Indicate security gain of the LV when selecting not lane-change strategy.
2) RV security gain
A.LV lane-change RV is not avoided
When LV lane-change, RV takes the security gain of not avoidance strategy same and Δ T-phase pass.When Δ T > 0, RV is arrived after LV
Up to potential conflict point, continuing selection in this case, avoidance strategy does not have very big security risk then, so RV in this case
Security gain be taken as minus infinity.As Δ T < 0, RV reaches potential conflict point before LV at this time.The safety of RV is received at this time
Beneficial then be with the reduction of Δ T, then safety is higher, and rate is gradually reduced.When Δ T reaches-TMAfterwards, security gain reaches most
Height, security gain keeps stablizing later.Security gain expression formula is as follows,
Wherein,When indicating LV lane-change, RV selects the security gain of not avoidance strategy.
B.LV lane-change RV evacuation
When LV lane-change, RV takes the security gain of avoidance strategy then opposite with above formula.When Δ T < 0, indicates that RV takes and keep away
Strategy is allowed still first to reach potential conflict point than LV, RV selection avoidance strategy is unreasonable in this case, also dangerous
, so the security gain of RV is taken as minus infinity.When Δ T > 0, RV reaches potential conflict point after LV, at this time the safety of RV
For income then with the increase of Δ T, safety is higher.When Δ T reaches TMLater, security gain reaches highest.As follows,
Wherein,When indicating LV lane-change, RV selects the security gain of avoidance strategy.
C.LV not lane-change
Not in the case of lane-change, the safety of RV is equally unaffected LV, so the security gain of RV is still kept most
Big value.
Wherein,Indicate security gain of the RV in LV not lane-change.
(4) total revenue
After speed income, security gain and the comfort income that LV and RV has been determined, next to pass through these three
Income calculation goes out the total revenue of vehicle.LV and RV total revenue expression formula is as follows,
Wherein,Indicate the speed income of LV,Indicate the security gain of LV,Indicate the comfort income of LV,
α1, β1, γ1Weight parameter between tri- incomes of LV is respectively indicated, f (*) is indicated to the normalized function of each income.
Wherein,Indicate the speed income of RV,Indicate the security gain of RV,Indicate that the comfort of RV is received
Benefit, α2, β2, γ2Weight parameter between tri- incomes of RV is respectively indicated, δ indicates that the mankind drive the income random number of vehicle RV.
4, model solution
Last solution includes that final strategy and acceleration select two parts in the present invention.LV first and RV finds out above-mentioned respectively
Then corresponding vehicle acceleration under four kinds of strategy combinations determines LV and RV respectively in four kinds of strategy combinations by acceleration
Under aggregate earnings value, finally by gain matrix obtain two vehicles it is respective it is final strategy and the strategy combination under two vehicles acceleration
Degree.
(1) acceleration selects
1) lane-change vehicle acceleration selects
Lane-change vehicle LV is not only influenced by RV vehicle when selecting acceleration, also has relationship with PV and FV vehicle, because
Mainly to study the conflict between lane-change vehicle LV and target lane rear car RV, RV only considered in LV security gain to the shadow of LV
It rings, so acceleration cannot be calculated by financial value.Therefore LV is to find out to select under two kinds of strategies of lane-change and non-lane-change respectively
The acceleration and then calculating financial value selected.Acceleration of the lane-change vehicle LV under two kinds of strategies of lane-change and non-lane-change is described below
Degree selection.
A.LV selects lane-change strategy
Under lane-change strategy, longitudinal acceleration selection is influenced by FV and RV two cars LV simultaneously, when LV selection is changed
Road strategy and when preparing lane-change, guarantee the safety between LV and vehicle FV and RV simultaneously.One is introduced based on desired vehicle
When head away from lane-change Fast track surgery, the model think LV acceleration during lane-change selection be LV pursue its with vehicle FV and
RV keeps the process of expectation time headway, and acceleration is the difference of the true time headway and expectation time headway by LV and FV, RV
It determines.Model expression is as follows:
Wherein,Represent acceleration of the LV under lane-change strategy, hf(t) time headway of t moment LV and FV, h are representedr
(t) time headway of t moment LV and RV is represented,The time headway of t moment driver desired FV and LV is represented,
The time headway of t moment driver desired RV and LV is represented, k indicates lane-change vehicle in total acceleration to target lane front truck
The considerations of degree, a1, b1, c1, a2, b2, c2It is parameter.
B.LV selects not lane-change strategy
Lane-change vehicle LV selection not lane-change strategy when, that is, LV continues in former lane with speeding front truck PV when driving, with
During speeding, vehicle LV and PV keep a safe distance, which can guarantee that LV and PV are not when PV carries out emergency brake
Rear-end collision occurs.So introducing the safe distance rule of Gipps, acceleration is calculated.According to Gipps model, in order not to PV
It collides, the safety that LV need to be kept is as follows with speed of speeding:
Wherein,It is longitudinal safe speed of LV vehicle relative vehicle PV, bpAnd blIt is that vehicle PV and LV are respective
Maximum brake acceleration.xl(t) be t moment lane-change vehicle LV lengthwise position, xp(t) be t moment front truck PV lengthwise position, τ
It is the reaction time.
Then the acceleration of LV is not shown below under lane-change strategy:
Wherein,Indicate the acceleration that LV will be selected under not lane-change strategy.
2) rear car acceleration in target lane selects
A.RV selects avoidance strategy
Acceleration under the avoidance strategy of RV vehicle can pass through expected evacuation speed required by formula (11)It finds out.From RV
When vehicle selects avoidance strategy, RV vehicle needs to guarantee it to be not more thanAverage speed travel to potential conflict point,
It can guarantee the safety of game both sides' vehicle in this way.Then RV reach it is potential punching point speed be,
In formula,Indicate that RV vehicle reaches the speed of potential conflict point, vr(t) indicate RV vehicle t moment speed,Indicate the expected evacuation speed of RV.
To which by kinematics formula, evacuation acceleration of the RV vehicle under avoidance strategy is shown below,
In formula,Indicate the evacuation acceleration of RV, LrIndicate RV at a distance from potential conflict point.
B.RV selects not avoidance strategy
When RV vehicle takes not avoidance strategy, it will continue with front truck FV traveling of speeding, during with speeding and front vehicles
A safe distance is kept, also according to Gipps model, finding out RV need to keep safe with speed of speeding:
Wherein,It is longitudinal safe speed of RV vehicle relative vehicle FV, brAnd bfIt is that vehicle RV and FV are respective
Maximum brake acceleration.xr(t) be t moment lane-change vehicle RV lengthwise position, xf(t) be t moment front truck FV lengthwise position.
Then the acceleration of RV is not shown below under avoidance strategy,
Wherein,Indicate acceleration of the RV not under avoidance strategy.
(2) automatic driving vehicle policy selection
After completing the acceleration ladder-like selection model under each strategy of lane-change vehicle LV and target lane rear car RV, from
The dynamic policy selection problem for driving vehicle in lane-change then becomes our critical issues to be solved.
In the present invention, automatic driving vehicle is to construct an automatic driving vehicle nesting lane-change game framework first,
The game framework is able to reflect out the lane-change gambling process between automatic driving vehicle LV and mankind's driving vehicle RV.Then, certainly
Dynamic driving vehicle simulates the gambling process of two vehicles using the acceleration selection algorithm of lane-change vehicle LV and target lane rear car RV,
It obtains the information of vehicles in each game stage in automatic driving vehicle nesting lane-change game framework, and predicts that following each game stage is double
The total revenue of square vehicle.Finally, automatic driving vehicle utilizes two vehicle incomes in nested lane-change game framework, by inversely concluding
Method obtains the optimal strategy that should be selected the current generation.Detailed process is as shown in Figure 5.
The process that the reverse induction of automatic Pilot lane-change vehicle LV solves optimal policy is described in detail below.Fig. 6 is simulated
Automatic driving vehicle LV and the mankind drive the gambling process between vehicle RV, for convenience of algorithm, each stage vehicle receipts of game are illustrated
Benefit is provided with specific value.
When RV is when STEP2 selects not avoidance strategy, LV, in STEP3, actively makes a concession as automatic driving vehicle,
Not lane-change strategy is selected, the income of LV is 3 at this time.When LV is when STEP2 selects lane-change strategy, it is contemplated that RV vehicle can select to make
The bigger not avoidance strategy of its STEP2 income, so it is 3 that LV, which selects the income of lane-change strategy in STEP2, and LV is selected in STEP2
The income for selecting not lane-change strategy is 5, so LV is not lane-change strategy, income 5 in the strategy of STEP2.Similarly, when LV exists
When STEP1 selects lane-change strategy, it is contemplated that RV vehicle can select the not avoidance strategy for keeping it bigger in STEP1 income, so LV
In the income of STEP1 selection lane-change strategy are as follows: RV is not when STEP2 selection avoids, strategy and income (i.e. LV of the LV in STEP2
Not lane-change, income 5).And LV selects not lane-change strategy, income 2 in STEP1, so the final strategy of LV is in STEP2
Select not lane-change.Automatic driving vehicle LV predicts its own meeting lane-change failure in current lane-change game, so, automatic Pilot
The optimal policy of vehicle is not lane-change.
(3) mankind drive vehicle policy selection
Mankind's driving vehicle is different from automatic driving vehicle, and automatic driving vehicle can be driven by driving vehicle to the mankind
The general rule for sailing behavior is modeled, and simulates entire lane-change gambling process, Jin Erxuan by lane-change nesting game playing algorithm
Select the optimal policy of oneself.And the mankind drive vehicle when carrying out policy selection, it can be to will be selected in the game current generation
The income of strategy be compared, and the therefrom bigger strategy of selection income.
5, the simulating, verifying of technical effect
Simulating scenes are described first.Then this model and existing model are compared, elaborates two moulds
Difference between type, and simulation analysis has been carried out to the safety of model, this model is demonstrated with enough safeties.Most
Afterwards, simulation analysis has been carried out to model, to conflict time difference of this model under various scenes, lane-change success rate, game number
It is statisticallyd analyze with acceleration of gravity, to demonstrate the reasonability of model.
(1) simulating scenes
Before model emulation, it is necessary first to a scene is constructed, as shown in fig. 7, lane-change vehicle LV is automatic Pilot vehicle
, target lane rear car RV is that the mankind drive vehicle.Assuming that current lane Maximum speed limit 100km/h, target lane Maximum speed limit
120km/h, vehicle original state are all to drive at a constant speed, and LV initial velocity is 90km/h, and RV initial velocity is 110km/h, at the beginning of PV
Beginning speed is 90km/h, and FV initial velocity is 120km/h, and it is 0m, PV and FV initial position pair that RV initial position, which corresponds to x-axis coordinate,
Answering x-axis coordinate is 120m, simulates different lane-change scenes by the variation of LV initial position.Speed corresponding to LV and RV vehicle
Income weight α=0.3, security gain weight beta=0.5, comfort income weight γ=0.2.Criticality safety conflict time difference TM
=3s.
(2) model compares
Opponent-automatically mixing drive lane-change environment existing research in, model not can reflect mankind's vehicle RV with
The practical lane-change gambling process of automatic driving vehicle LV, LV combine game theory think of in its automatic Pilot policy selection algorithm
Think, predicts the policy selection process of RV.And in actual lane-change gambling process, LV is but absolutely not in view of RV vehicle
Strategy, and only by calculating the aggressiveness of RV previous step, and automatic driving vehicle policy selection model is introduced into as parameter
In financial value calculate.Although this method may be more accurate in the judgement for mankind's vehicle financial value, because of people
Class vehicle inherently has very big randomness, and aggressiveness is only used as a parameter of vehicle revenue function, to the mankind's
Aggressiveness is estimated there are in the case where error or under the containing of other income, and LV does not make in mankind's vehicle probably and keeping away
In the case where allowing, lane-change is carried out, to cause security risk.
And the present invention has carried out more deep analysis to the lane-change gambling process of automatic driving vehicle and mankind's vehicle, it is right
The multistep dynamic game frame for establishing LV and RV of RV.The frame embodies in practical lane-change game, and the strategy of LV and RV select
Select process.Automatic driving vehicle and mankind's vehicle can all make itself strategy according to the strategy that opponent vehicle previous step selects and select
It selects, and only in the case where side's vehicle is made a concession, lane-change game can just terminate, and in addition side's vehicle then can be selected directly
Select uncompromising strategy, in addition, before vehicle does not determine that game terminates stand fast strategy all have it is exploratory.Institute
With automatic driving vehicle can select to be that direct lane-change is also to continue with for the strategy of previous step mankind's vehicle rich in this model
It plays chess, there is higher safety.
Simulation analysis is carried out to the safety of model and existing model below.Rushing using the LV and RV at lane-change moment herein
The prominent time difference analyzes safety, by carrying out 100 emulation to different lane-change scenes respectively, finally counts two models and changes
The conflict time difference at road moment, it is as shown in Figure 8 to draw out box figure.As shown in Figure 8, the top of existing model and this paper model
Edge, upper quartile and median all relatively, and both greater than critical conflict time difference TM.But the lower edge of existing model
And lower quartile, all 0 between 1s, for this explanation if LV lane-change at this time, two vehicles almost reach potential conflict simultaneously
Point, RV do not make the behavior of evacuation, but the strategy of LV is but determined as direct lane-change in existing model.And this paper model
Lower edge is slightly less than 3s, and lower quartile illustrates that at the time of LV selects lane-change be that can guarantee both sides' vehicle on the position of 4s
Safety.
The lane-change situation of two models corresponding to the box figure lower edge conflict time difference is as shown in Figure 9 and Figure 10.By Fig. 9
It can be seen that RV does not make avoidance strategy, and two vehicles conflict, and the time difference is very small, and LV reaches potential conflict point when LV carries out lane-change
When, RV with speeding apart from very little, there are bigger security risks for two vehicles.And LV is judging RV selection avoidance strategy in Figure 10
Later, lane-change behavior has then been selected, so the two vehicles conflict time difference is bigger, has been terminated since the lane-change moment to lane-change, RV
It is safe with distance of speeding compared with LV is kept always.It can be seen that this model has higher safety.
(3) model analysis
15 different lane-change scenes in LV initial position 0-80m are had chosen, each lane-change scene has been carried out 100 times
Emulation exports related data, conflict time difference, lane-change success rate including game start time and lane-change moment, game number
And acceleration of gravity, and it is as shown in figure 11 to draw simulation analysis figure.When Figure 11 (a) reflects game start time and lane-change respectively
Quarter LV conflicts the time difference with RV's, can be seen that conflict time difference all very littles of game start time in figure, if direct lane-change, has
Very big security risk.And after game, the conflict time difference at lane-change moment will be higher by much compared to game start time,
Illustrate that LV by selected lane-change behavior after game is safe.Figure 11 (b) and Figure 11 (c) reflect LV initial bit respectively
Relationship between the variation set and lane-change success rate and game number.When LV initial position is between 0-20m, two spacing at this time
Close, lane-change success rate is 0%, vehicle game number all 1 time, shows that LV actively selects not lane-change in game for the first time
Strategy.When LV initial position is between 20-60m, the lane-change success rate of LV is about 80%, the game 2 of fraction also occurs
Secondary situation illustrates that LV game for the first time under these lane-change scenes can all select lane-change strategy, since RV is mankind's vehicle, one
It in a little situations is influenced that LV will not be avoided by enchancement factor, LV lane-change is caused to fail.When LV initial position is between 60-80m
When, the lane-change success rate of LV is about 100%, and for game number other than occurring 1 time and 2 times, there are also small parts to occur 3 times
With 4 situations, this explanation is increasing with the spacing of LV and RV when game starts, and LV has bigger lane-change space, institute
Also increasingly to occupy advantage in game, LV also will appear the game of more numbers, and the lane-change success rate of LV also significantly mentions
It is high.Figure 11 (d) reflects the relationship between the variation of LV initial position and the acceleration of gravity of two vehicles game for the first time.LV lane-change accelerates
The trend being gradually increasing is presented in degree, and then downward trend is not presented as the spacing of LV and PV reduces in lane-change acceleration to LV, RV
Avoid acceleration then with LV and RV spacing increase and present rise trend, RV do not avoid acceleration then with RV and FV spacing
It is related, so the trend of variation is not presented.
Claims (8)
1. automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment, characterized by the following steps:
Step 1: LV generates lane-changing intention；
Step 2: judging whether that meeting lane-change game starts condition: if it is, entering step three；If it is not, then whether judging LV
Meet lane-change condition: if so, nine are entered step, if it is not, then waiting generate lane-changing intention next time, subsequently into step 1；
Step 3: LV building Dynamic Game Model and two vehicle incomes of calculating；
Step 4: finding out the optimal policy that LV should be selected；
Step 5: judging whether to meet lane-change game termination condition: if so, ten are entered step, if it is not, then entering step six；
Step 6: LV opens turn signal and exploratory lateral shift；
Step 7: LV selects lane-change acceleration；
Step 8: judging whether RV selects to avoid: if it is not, then return step three；If so, entering step nine；
Step 9: LV starts lane-change, until lane-change terminates；
Step 10: LV continues in former lane with speeding, this lane-changing intention terminates.
2. automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment according to claim 1, special
Sign is: the lane-change game starts condition, and to include: LV keep safe distance with PV and LV is no more than with the time difference that conflicts of PV
The threshold value of setting, in which:
(1) safe distance of LV and PV is calculated using following formula:
In formula, GlFor the safe distance of lane-change vehicle LV and current lane front truck PV, xlIt (t) is the position t moment lane-change vehicle LV,
xpIt (t) is the position t moment current lane front truck PV, lpFor the length of wagon of current lane front truck PV, blFor the maximum deceleration of LV
Degree, τlFor the reaction time of LV, vl(t) speed for being t moment LV, vp(t) speed for being t moment PV；
(2) calculation method of LV and the time difference that conflicts of PV are as follows:
1) LV is calculated as follows and reaches the distance travelled in potential conflict point process from lane-change starting point:
Wherein, LlThe operating range of potential conflict point is changed lane to from current location for LV；
2) distance that RV drives to potential conflict point from current location is calculated as follows:
Lr=xc+d
Wherein, LrIndicate the operating range of RV point from current location to potential conflict, xcIndicate that the left front angle point distance of lane-change vehicle is latent
In the fore-and-aft distance of conflict point, d indicates LV and the space headway of RV in the longitudinal direction；
3)
Wherein: vr(t) and vl(t) speed of the current t moment of RV and LV is respectively indicated.
3. automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment according to claim 2, special
Sign is: the position of the potential conflict point determines with the following method:
(1) the ordinate y of potential conflict point is calculatedc:
yc=ye-wcar
Wherein, yeFor the ordinate of lane-change geometric locus terminal, wcarIndicate vehicle width；
(2) lane-change equation of locus is established:
Wherein, x and y is the horizontal and vertical position of vehicle LV headstock left end, xeAnd yeRespectively the transverse direction of lane-change final on trajectory and
Longitudinal coordinate；
(3) by ycValue bring lane-change trajectory curve equation into and can solve the abscissa x of potential conflict pointc, finally obtain potential
Position (the x of conflict pointc, yc)。
4. automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment according to claim 2, special
Sign is: the lane-change game termination condition include: when automatic Pilot lane-change vehicle LV game step number reaches the step number of setting or
Person is during lane-change as automatic driving vehicle LV and when being unsatisfactory for safe distance with distance of speeding of current lane front truck PV.
5. automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment according to claim 1, special
Sign is: building Dynamic Game Model and the method for calculating two vehicle incomes described in step 3 are as follows: first calculates LV and RV in four kinds of plans
Corresponding vehicle acceleration under slightly combining, then determines that LV and RV are total under four kinds of strategy combinations respectively by acceleration
Financial value.
6. automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment according to claim 5, special
Sign is: the method for calculating LV and RV corresponding vehicle acceleration under four kinds of strategy combinations are as follows:
(1) acceleration under LV selection lane-change strategy is calculated:
Wherein,Represent acceleration of the LV under lane-change strategy, hf(t) time headway of t moment LV and FV, h are representedr(t)
The time headway of t moment LV and RV is represented,The time headway of t moment driver desired FV and LV is represented,It represents
The time headway of t moment driver desired RV and LV, k indicate that lane-change vehicle examines target lane front truck in total acceleration
Worry degree, a1, b1, c1, a2, b2, c2It is parameter；
(2) calculating LV selects the acceleration under not lane-change strategy:
Wherein,Indicate the acceleration that LV will be selected under not lane-change strategy,It is the vertical of LV vehicle relative vehicle PV
To safe speed, it is calculated as follows:
Wherein, bpAnd blIt is the respective maximum brake acceleration of vehicle PV and LV, xl(t) be t moment lane-change vehicle LV longitudinal position
It sets, xp(t) be t moment front truck PV lengthwise position, τ is the reaction time；
(3) acceleration under RV selection avoidance strategy is calculated:
In formula,Indicate the evacuation acceleration of RV, LrRV is at a distance from potential conflict point for expression, vr(t) indicate RV vehicle in t
The speed at quarter,It indicates that RV vehicle reaches the speed of potential conflict point, is calculated as follows:
In formula,Indicate the expected evacuation speed of RV；
(4) calculating RV selects the acceleration under not avoidance strategy:
Wherein,Indicate acceleration of the RV not under avoidance strategy,It is longitudinal safety of RV vehicle relative vehicle FV
Speed is calculated as follows:
In formula, brAnd bfIt is the respective maximum brake acceleration of vehicle RV and FV, xr(t) be t moment lane-change vehicle RV longitudinal position
It sets, xf(t) be t moment front truck FV lengthwise position.
7. automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment according to claim 6, special
Sign is: the method for calculating the LV and RV aggregate earnings value under four kinds of strategy combinations respectively are as follows:
(1) calculating speed income:
1) speed income of the LV under not lane-change strategy:
Wherein,Indicate speed income of the LV under not lane-change strategy, vp(t) speed of the PV in t moment, v are indicatedl(t) it indicates
Speed of the LV in t moment；
2) speed income of the LV under lane-change strategy:
Wherein,Indicate speed income of the LV under lane-change strategy, vf(t) speed of the FV in t moment, v are indicatedl(t) LV is indicated
In the speed of t moment；
3) speed income of the RV under avoidance strategy:
Wherein,Indicate the speed income under RV avoidance strategy,The expected evacuation speed for indicating RV, is calculated as follows:
Wherein, TlIndicate that LV executes the running time that lane-change reaches potential conflict point from current location；
4) speed income of the RV under not avoidance strategy:
Wherein,Indicate speed income of the RV not under avoidance strategy, vf(t) and vr(t) distinguish target lane front truck FV and RV
Speed；
(2) comfort income is calculated:
1) the comfort income of LV:
Wherein,Indicate the comfort income of LV, alThe acceleration of (t- λ) expression long LV of previous step；
2) the comfort income of RV:
Wherein,Indicate the comfort income of RV, arThe acceleration of (t- λ) expression long RV of previous step；
(3) security gain is calculated:
1) security gain of the LV when selecting lane-change strategy:
Wherein,Indicate that the security gain in LV when selecting lane-change strategy, Δ T indicate target lane rear car RV and lane-change vehicle
The conflict time difference of LV, TMIndicate safety critical point；
2) security gain of the LV when selecting not lane-change strategy:
Wherein,Indicate security gain of the LV when selecting not lane-change strategy；
3) when LV lane-change, RV selects the security gain of not avoidance strategy:
Wherein,When indicating LV lane-change, RV selects the security gain of not avoidance strategy；
4) when LV lane-change, RV selects the security gain of avoidance strategy:
Wherein,When indicating LV lane-change, RV selects the security gain of avoidance strategy；
5) security gain of the RV in LV not lane-change:
Wherein,Indicate security gain of the RV in LV not lane-change；
(4) aggregate earnings value is calculated:
1) total revenue of LV is calculated:
Wherein,Indicate the speed income of LV,Indicate the security gain of LV,Indicate the comfort income of LV, α1,
β1, γ1Weight parameter between tri- incomes of LV is respectively indicated, f (*) is indicated to the normalized function of each income；
2) total revenue of RV is calculated:
Wherein,Indicate the speed income of RV,Indicate the security gain of RV,Indicate the comfort income of RV, α2,
β2, γ2Weight parameter between tri- incomes of RV is respectively indicated, δ indicates that the mankind drive the income random number of vehicle RV.
8. automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment according to claim 1, special
Sign is: finding out the optimal policy that LV should be selected using reverse induction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910603412.7A CN110298131B (en) | 2019-07-05 | 2019-07-05 | Method for establishing automatic driving lane change decision model in hybrid driving environment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910603412.7A CN110298131B (en) | 2019-07-05 | 2019-07-05 | Method for establishing automatic driving lane change decision model in hybrid driving environment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110298131A true CN110298131A (en) | 2019-10-01 |
| CN110298131B CN110298131B (en) | 2021-07-13 |
Family
ID=68030488
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910603412.7A Active CN110298131B (en) | 2019-07-05 | 2019-07-05 | Method for establishing automatic driving lane change decision model in hybrid driving environment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110298131B (en) |
Cited By (39)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110884423A (en) * | 2019-12-20 | 2020-03-17 | 吉林大学 | Automatic control system and method for steering lamp under autonomous lane change of vehicle |
| CN110962853A (en) * | 2019-11-25 | 2020-04-07 | 浙江工业大学 | Vehicle game lane change cooperation method in Internet of vehicles environment |
| CN111081065A (en) * | 2019-12-13 | 2020-04-28 | 北京理工大学 | Intelligent vehicle collaborative lane change decision model under road section mixed traveling condition |
| CN111199284A (en) * | 2019-12-17 | 2020-05-26 | 天津职业技术师范大学(中国职业培训指导教师进修中心) | Vehicle-vehicle interaction model under condition of manned and unmanned mixed driving |
| CN111267846A (en) * | 2020-02-11 | 2020-06-12 | 南京航空航天大学 | Game theory-based peripheral vehicle interaction behavior prediction method |
| CN111361564A (en) * | 2020-04-29 | 2020-07-03 | 吉林大学 | Lane change system considering benefit maximization and comprehensive decision method |
| CN111391848A (en) * | 2020-03-02 | 2020-07-10 | 吉林大学 | Automatic driving vehicle lane changing method |
| CN111439264A (en) * | 2020-04-03 | 2020-07-24 | 东南大学 | Implementation method of lane change control model based on man-machine hybrid driving |
| CN111739342A (en) * | 2020-03-26 | 2020-10-02 | 腾讯科技（深圳）有限公司 | Method, device, medium, and vehicle for avoiding vehicle ahead of side |
| CN111994090A (en) * | 2020-09-02 | 2020-11-27 | 中国科学技术大学 | Method and system for identifying lane-changing cut-in intention of driver based on hybrid strategy game |
| CN111994089A (en) * | 2020-09-02 | 2020-11-27 | 中国科学技术大学 | Driver lane change intention identification method and system based on hybrid strategy game |
| CN111994088A (en) * | 2020-09-02 | 2020-11-27 | 中国科学技术大学 | Driver lane change intention identification method and system based on hybrid strategy game |
| CN112348198A (en) * | 2020-10-30 | 2021-02-09 | 上海对外经贸大学 | Coordination method of machine behaviors of man-machine hybrid decision in conflict |
| CN112373485A (en) * | 2020-11-03 | 2021-02-19 | 南京航空航天大学 | Decision planning method for automatic driving vehicle considering interactive game |
| CN112590791A (en) * | 2020-12-16 | 2021-04-02 | 东南大学 | Intelligent vehicle lane change gap selection method and device based on game theory |
| CN112721929A (en) * | 2021-01-11 | 2021-04-30 | 成都语动未来科技有限公司 | Decision-making method for lane changing behavior of automatic driving vehicle based on search technology |
| CN112907967A (en) * | 2021-01-29 | 2021-06-04 | 吉林大学 | Intelligent vehicle lane change decision-making method based on incomplete information game |
| CN113076897A (en) * | 2021-04-09 | 2021-07-06 | 广州机械科学研究院有限公司 | Game dynamic driving safety measurement and control method and regulation and control terminal of intelligent networked automobile |
| CN113255101A (en) * | 2021-04-22 | 2021-08-13 | 东南大学 | Method and device for calibrating lane change simulation model of vehicle at intersection entrance lane |
| CN113313941A (en) * | 2021-05-25 | 2021-08-27 | 北京航空航天大学 | Vehicle track prediction method based on memory network and encoder-decoder model |
| CN113335282A (en) * | 2021-06-01 | 2021-09-03 | 南京航空航天大学 | Path changing decision-making method based on game theory |
| CN113361613A (en) * | 2021-06-15 | 2021-09-07 | 东南大学 | Method and device for classifying ramp vehicle lane change simulation models based on trajectory data |
| CN113495563A (en) * | 2021-06-10 | 2021-10-12 | 吉林大学 | Traffic vehicle lane change decision planning method for automatic driving virtual test |
| CN113823118A (en) * | 2021-02-19 | 2021-12-21 | 石家庄铁道大学 | Intelligent network vehicle lane changing method combining urgency degree and game theory |
| CN113920740A (en) * | 2021-11-16 | 2022-01-11 | 重庆邮电大学 | Vehicle-road cooperative driving system and method combining vehicle association degree and game theory |
| CN113954869A (en) * | 2020-06-01 | 2022-01-21 | 北京航迹科技有限公司 | Road condition reminding method and device, electronic equipment and storage medium |
| CN113963535A (en) * | 2021-09-30 | 2022-01-21 | 华为技术有限公司 | Driving decision determination method and device and electronic equipment storage medium |
| WO2022052856A1 (en) * | 2020-09-10 | 2022-03-17 | 腾讯科技（深圳）有限公司 | Vehicle-based data processing method and apparatus, computer, and storage medium |
| CN114360319A (en) * | 2022-01-17 | 2022-04-15 | 中山大学 | Remote driving speed optimization method of driving simulator |
| CN114506324A (en) * | 2020-10-23 | 2022-05-17 | 上海汽车集团股份有限公司 | Lane decision method and related device |
| CN114882705A (en) * | 2022-05-30 | 2022-08-09 | 武汉理工大学 | Freight vehicle interactive game lane change decision-making method based on lane change decision-making system |
| CN115412883A (en) * | 2022-08-31 | 2022-11-29 | 重庆交通大学 | Intelligent network connection over-the-horizon driving auxiliary system based on 5G position sharing |
| CN115456392A (en) * | 2022-09-06 | 2022-12-09 | 长安大学 | High-speed multi-vehicle multi-driving behavior conflict collaborative decision-making method and device |
| CN115547047A (en) * | 2022-09-30 | 2022-12-30 | 中汽院智能网联科技有限公司 | Intelligent internet vehicle following model based on attention model |
| CN115906265A (en) * | 2022-12-27 | 2023-04-04 | 中交第二公路勘察设计研究院有限公司 | Near main line outlet marking optimization method based on lane changing behavior characteristics |
| WO2023087157A1 (en) * | 2021-11-16 | 2023-05-25 | 华为技术有限公司 | Intelligent driving method and vehicle applying same |
| CN113954828B (en) * | 2021-10-26 | 2023-08-29 | 江苏科创车联网产业研究院有限公司 | Automatic driving vehicle cruise control method and device and electronic equipment |
| CN117284297A (en) * | 2023-11-27 | 2023-12-26 | 福思(杭州)智能科技有限公司 | Vehicle control method and device and domain controller |
| CN114882705B (en) * | 2022-05-30 | 2024-04-26 | 武汉理工大学 | Freight vehicle interactive game lane change decision method based on lane change decision system |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2854075A2 (en) * | 2013-09-26 | 2015-04-01 | Denso Corporation | Vehicle control system and image sensor |
| CN106379237A (en) * | 2016-09-30 | 2017-02-08 | 西南交通大学 | Augmented reality-based lane changing whole-process driver assistant system of vehicle |
| CN107452201A (en) * | 2017-07-24 | 2017-12-08 | 重庆大学 | Rear car determines method and with speeding on as modeling method with acceleration of speeding when a kind of consideration front truck lane-change is sailed out of |
| CN108387242A (en) * | 2018-02-07 | 2018-08-10 | 西南交通大学 | Automatic Pilot lane-change prepares and executes integrated method for planning track |
| CN108595823A (en) * | 2018-04-20 | 2018-09-28 | 大连理工大学 | A kind of computational methods of Autonomous Vehicles lane-change strategy that combining driving style and theory of games |
| CN108583578A (en) * | 2018-04-26 | 2018-09-28 | 北京领骏科技有限公司 | The track decision-making technique based on multiobjective decision-making matrix for automatic driving vehicle |
| CN109540159A (en) * | 2018-10-11 | 2019-03-29 | 同济大学 | A kind of quick complete automatic Pilot method for planning track |
| CN109855639A (en) * | 2019-01-15 | 2019-06-07 | 天津大学 | Unmanned method for planning track based on forecasting-obstacle and MPC algorithm |
-
2019
- 2019-07-05 CN CN201910603412.7A patent/CN110298131B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2854075A2 (en) * | 2013-09-26 | 2015-04-01 | Denso Corporation | Vehicle control system and image sensor |
| CN106379237A (en) * | 2016-09-30 | 2017-02-08 | 西南交通大学 | Augmented reality-based lane changing whole-process driver assistant system of vehicle |
| CN107452201A (en) * | 2017-07-24 | 2017-12-08 | 重庆大学 | Rear car determines method and with speeding on as modeling method with acceleration of speeding when a kind of consideration front truck lane-change is sailed out of |
| CN108387242A (en) * | 2018-02-07 | 2018-08-10 | 西南交通大学 | Automatic Pilot lane-change prepares and executes integrated method for planning track |
| CN108595823A (en) * | 2018-04-20 | 2018-09-28 | 大连理工大学 | A kind of computational methods of Autonomous Vehicles lane-change strategy that combining driving style and theory of games |
| CN108583578A (en) * | 2018-04-26 | 2018-09-28 | 北京领骏科技有限公司 | The track decision-making technique based on multiobjective decision-making matrix for automatic driving vehicle |
| CN109540159A (en) * | 2018-10-11 | 2019-03-29 | 同济大学 | A kind of quick complete automatic Pilot method for planning track |
| CN109855639A (en) * | 2019-01-15 | 2019-06-07 | 天津大学 | Unmanned method for planning track based on forecasting-obstacle and MPC algorithm |
Non-Patent Citations (4)
| Title |
|---|
| YANG DA 等: ""A dynamic lane-changing trajectory planning model for automated vehicles"", 《TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES》 * |
| 聂建强: ""高速公路车辆自主性换道行为建模研究"", 《中国博士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 薛春铭 等: ""基于博弈论的人类驾驶与无人驾驶协作换道模型"", 《计算机工程》 * |
| 薛春铭: ""基于博弈的车辆协作换道策略研究"", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (56)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110962853A (en) * | 2019-11-25 | 2020-04-07 | 浙江工业大学 | Vehicle game lane change cooperation method in Internet of vehicles environment |
| CN111081065A (en) * | 2019-12-13 | 2020-04-28 | 北京理工大学 | Intelligent vehicle collaborative lane change decision model under road section mixed traveling condition |
| CN111081065B (en) * | 2019-12-13 | 2021-03-30 | 北京理工大学 | Intelligent vehicle collaborative lane change decision model under road section mixed traveling condition |
| CN111199284A (en) * | 2019-12-17 | 2020-05-26 | 天津职业技术师范大学(中国职业培训指导教师进修中心) | Vehicle-vehicle interaction model under condition of manned and unmanned mixed driving |
| CN110884423A (en) * | 2019-12-20 | 2020-03-17 | 吉林大学 | Automatic control system and method for steering lamp under autonomous lane change of vehicle |
| CN110884423B (en) * | 2019-12-20 | 2023-09-08 | 吉林大学 | Automatic control system and method for steering lamp under automatic lane change of vehicle |
| CN111267846A (en) * | 2020-02-11 | 2020-06-12 | 南京航空航天大学 | Game theory-based peripheral vehicle interaction behavior prediction method |
| CN111267846B (en) * | 2020-02-11 | 2021-05-11 | 南京航空航天大学 | Game theory-based peripheral vehicle interaction behavior prediction method |
| CN111391848B (en) * | 2020-03-02 | 2022-03-08 | 吉林大学 | Automatic driving vehicle lane changing method |
| CN111391848A (en) * | 2020-03-02 | 2020-07-10 | 吉林大学 | Automatic driving vehicle lane changing method |
| CN111739342A (en) * | 2020-03-26 | 2020-10-02 | 腾讯科技（深圳）有限公司 | Method, device, medium, and vehicle for avoiding vehicle ahead of side |
| CN111739342B (en) * | 2020-03-26 | 2022-03-01 | 腾讯科技（深圳）有限公司 | Method, device, medium, and vehicle for avoiding vehicle ahead of side |
| CN111439264A (en) * | 2020-04-03 | 2020-07-24 | 东南大学 | Implementation method of lane change control model based on man-machine hybrid driving |
| CN111361564B (en) * | 2020-04-29 | 2023-07-07 | 吉林大学 | Lane changing system considering benefit maximization and comprehensive decision method |
| CN111361564A (en) * | 2020-04-29 | 2020-07-03 | 吉林大学 | Lane change system considering benefit maximization and comprehensive decision method |
| CN113954869A (en) * | 2020-06-01 | 2022-01-21 | 北京航迹科技有限公司 | Road condition reminding method and device, electronic equipment and storage medium |
| CN113954869B (en) * | 2020-06-01 | 2024-04-09 | 北京航迹科技有限公司 | Road condition reminding method and device, electronic equipment and storage medium |
| CN111994089A (en) * | 2020-09-02 | 2020-11-27 | 中国科学技术大学 | Driver lane change intention identification method and system based on hybrid strategy game |
| CN111994090A (en) * | 2020-09-02 | 2020-11-27 | 中国科学技术大学 | Method and system for identifying lane-changing cut-in intention of driver based on hybrid strategy game |
| CN111994088A (en) * | 2020-09-02 | 2020-11-27 | 中国科学技术大学 | Driver lane change intention identification method and system based on hybrid strategy game |
| CN111994090B (en) * | 2020-09-02 | 2021-11-02 | 中国科学技术大学 | Method and system for identifying lane-changing cut-in intention of driver based on hybrid strategy game |
| WO2022052856A1 (en) * | 2020-09-10 | 2022-03-17 | 腾讯科技（深圳）有限公司 | Vehicle-based data processing method and apparatus, computer, and storage medium |
| CN114506324A (en) * | 2020-10-23 | 2022-05-17 | 上海汽车集团股份有限公司 | Lane decision method and related device |
| CN114506324B (en) * | 2020-10-23 | 2024-03-15 | 上海汽车集团股份有限公司 | Lane decision method and related device |
| CN112348198A (en) * | 2020-10-30 | 2021-02-09 | 上海对外经贸大学 | Coordination method of machine behaviors of man-machine hybrid decision in conflict |
| CN112373485A (en) * | 2020-11-03 | 2021-02-19 | 南京航空航天大学 | Decision planning method for automatic driving vehicle considering interactive game |
| CN112590791A (en) * | 2020-12-16 | 2021-04-02 | 东南大学 | Intelligent vehicle lane change gap selection method and device based on game theory |
| CN112590791B (en) * | 2020-12-16 | 2022-03-11 | 东南大学 | Intelligent vehicle lane change gap selection method and device based on game theory |
| CN112721929B (en) * | 2021-01-11 | 2022-11-22 | 成都语动未来科技有限公司 | Decision-making method for lane changing behavior of automatic driving vehicle based on search technology |
| CN112721929A (en) * | 2021-01-11 | 2021-04-30 | 成都语动未来科技有限公司 | Decision-making method for lane changing behavior of automatic driving vehicle based on search technology |
| CN112907967A (en) * | 2021-01-29 | 2021-06-04 | 吉林大学 | Intelligent vehicle lane change decision-making method based on incomplete information game |
| CN113823118B (en) * | 2021-02-19 | 2022-07-08 | 石家庄铁道大学 | Intelligent networking vehicle lane changing method combining urgency degree and game theory |
| CN113823118A (en) * | 2021-02-19 | 2021-12-21 | 石家庄铁道大学 | Intelligent network vehicle lane changing method combining urgency degree and game theory |
| CN113076897A (en) * | 2021-04-09 | 2021-07-06 | 广州机械科学研究院有限公司 | Game dynamic driving safety measurement and control method and regulation and control terminal of intelligent networked automobile |
| CN113255101A (en) * | 2021-04-22 | 2021-08-13 | 东南大学 | Method and device for calibrating lane change simulation model of vehicle at intersection entrance lane |
| CN113313941A (en) * | 2021-05-25 | 2021-08-27 | 北京航空航天大学 | Vehicle track prediction method based on memory network and encoder-decoder model |
| CN113313941B (en) * | 2021-05-25 | 2022-06-24 | 北京航空航天大学 | Vehicle track prediction method based on memory network and encoder-decoder model |
| CN113335282A (en) * | 2021-06-01 | 2021-09-03 | 南京航空航天大学 | Path changing decision-making method based on game theory |
| CN113495563B (en) * | 2021-06-10 | 2022-09-20 | 吉林大学 | Traffic vehicle lane change decision planning method for automatic driving virtual test |
| CN113495563A (en) * | 2021-06-10 | 2021-10-12 | 吉林大学 | Traffic vehicle lane change decision planning method for automatic driving virtual test |
| CN113361613A (en) * | 2021-06-15 | 2021-09-07 | 东南大学 | Method and device for classifying ramp vehicle lane change simulation models based on trajectory data |
| CN113963535A (en) * | 2021-09-30 | 2022-01-21 | 华为技术有限公司 | Driving decision determination method and device and electronic equipment storage medium |
| CN113954828B (en) * | 2021-10-26 | 2023-08-29 | 江苏科创车联网产业研究院有限公司 | Automatic driving vehicle cruise control method and device and electronic equipment |
| CN113920740B (en) * | 2021-11-16 | 2023-12-29 | 北京白龙马云行科技有限公司 | Vehicle-road cooperative driving system and method combining vehicle association degree and game theory |
| WO2023087157A1 (en) * | 2021-11-16 | 2023-05-25 | 华为技术有限公司 | Intelligent driving method and vehicle applying same |
| CN113920740A (en) * | 2021-11-16 | 2022-01-11 | 重庆邮电大学 | Vehicle-road cooperative driving system and method combining vehicle association degree and game theory |
| CN114360319A (en) * | 2022-01-17 | 2022-04-15 | 中山大学 | Remote driving speed optimization method of driving simulator |
| CN114882705A (en) * | 2022-05-30 | 2022-08-09 | 武汉理工大学 | Freight vehicle interactive game lane change decision-making method based on lane change decision-making system |
| CN114882705B (en) * | 2022-05-30 | 2024-04-26 | 武汉理工大学 | Freight vehicle interactive game lane change decision method based on lane change decision system |
| CN115412883A (en) * | 2022-08-31 | 2022-11-29 | 重庆交通大学 | Intelligent network connection over-the-horizon driving auxiliary system based on 5G position sharing |
| CN115456392B (en) * | 2022-09-06 | 2023-09-05 | 长安大学 | High-speed multi-vehicle multi-driving behavior conflict collaborative decision-making method and device |
| CN115456392A (en) * | 2022-09-06 | 2022-12-09 | 长安大学 | High-speed multi-vehicle multi-driving behavior conflict collaborative decision-making method and device |
| CN115547047A (en) * | 2022-09-30 | 2022-12-30 | 中汽院智能网联科技有限公司 | Intelligent internet vehicle following model based on attention model |
| CN115906265A (en) * | 2022-12-27 | 2023-04-04 | 中交第二公路勘察设计研究院有限公司 | Near main line outlet marking optimization method based on lane changing behavior characteristics |
| CN117284297A (en) * | 2023-11-27 | 2023-12-26 | 福思(杭州)智能科技有限公司 | Vehicle control method and device and domain controller |
| CN117284297B (en) * | 2023-11-27 | 2024-02-27 | 福思(杭州)智能科技有限公司 | Vehicle control method and device and domain controller |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110298131B (en) | 2021-07-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110298131A (en) | Automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment | |
| CN110362910B (en) | Game theory-based automatic driving vehicle lane change conflict coordination model establishment method | |
| CN110298122B (en) | Unmanned vehicle urban intersection left-turn decision-making method based on conflict resolution | |
| CN109345020B (en) | Non-signalized intersection vehicle driving behavior prediction method under complete information | |
| CN109035862B (en) | Multi-vehicle cooperative lane change control method based on vehicle-to-vehicle communication | |
| CN108806252B (en) | A kind of Mixed Freeway Traffic Flows collaboration optimal control method | |
| CN111439264B (en) | Implementation method of lane change control model based on man-machine hybrid driving | |
| CN109709956A (en) | A kind of automatic driving vehicle speed control multiple-objection optimization with algorithm of speeding | |
| CN106997689A (en) | V2P collision free methods based on crossing | |
| Ding et al. | Multivehicle coordinated lane change strategy in the roundabout under internet of vehicles based on game theory and cognitive computing | |
| CN113276884A (en) | Intelligent vehicle interactive decision passing method and system with variable game mode | |
| CN116740945B (en) | Method and system for multi-vehicle collaborative grouping intersection of expressway confluence region in mixed running environment | |
| Ji et al. | Estimating the social gap with a game theory model of lane changing | |
| CN112201070A (en) | Deep learning-based automatic driving expressway bottleneck section behavior decision method | |
| CN112330135A (en) | Urban traffic jam space evolution method based on improved cellular automaton model | |
| CN115457782A (en) | Deep reinforcement learning-based conflict-free cooperation method for intersection of automatic driving vehicles | |
| CN115601958A (en) | Internet-of-vehicles traffic flow modeling method based on continuous cellular automaton | |
| Akti et al. | A game-theoretical approach for lane-changing maneuvers on freeway merging segments | |
| CN113823076A (en) | Instant-stop and instant-walking road section blockage relieving method based on networked vehicle coordination control | |
| CN117173910A (en) | Intelligent motorcade collaborative import method and device and electronic equipment | |
| Fu et al. | Cooperative decision-making of multiple autonomous vehicles in a connected mixed traffic environment: A coalition game-based model | |
| CN116090336A (en) | Virtual marshalling train reference curve calculation method based on improved reinforcement learning algorithm | |
| CN115188214A (en) | Two-lane mixed traffic cooperative control method, automobile and readable storage medium | |
| Hassanin et al. | Efficiency performance and safety evaluation of the responsibility-sensitive safety in freeway car-following scenarios using automated longitudinal controls | |
| Gu et al. | Mandatory Lane-Changing Decision-Making in Dense Traffic for Autonomous Vehicles based on Deep Reinforcement Learning |
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 |