CN110008577A - The automatic lane-change Function Appraising method of vehicle based on worst global danger level search - Google Patents
The automatic lane-change Function Appraising method of vehicle based on worst global danger level search Download PDFInfo
- Publication number
- CN110008577A CN110008577A CN201910255614.7A CN201910255614A CN110008577A CN 110008577 A CN110008577 A CN 110008577A CN 201910255614 A CN201910255614 A CN 201910255614A CN 110008577 A CN110008577 A CN 110008577A
- Authority
- CN
- China
- Prior art keywords
- lane
- dan
- vehicle
- danger level
- measured
- 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
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/15—Vehicle, aircraft or watercraft design
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Aviation & Aerospace Engineering (AREA)
- Automation & Control Theory (AREA)
- Traffic Control Systems (AREA)
- Control Of Driving Devices And Active Controlling Of Vehicle (AREA)
Abstract
The present invention discloses a kind of automatic lane-change Function Appraising method of vehicle based on worst global danger level search, this method is by building emulation platform, analysis to automatic lane-change function and lane-change scene to be measured, and then global search is carried out to most dangerous lane-change scene, search is with worst global danger level MSdanFor output, the corresponding most dangerous lane-change scene of worst global danger level is reproduced by actual place, automatic Pilot lane-change function to be measured is verified, if automatic Pilot lane-change function to be measured can meet service requirement under most dangerous lane-change scene, assert that the automatic Pilot lane-change function to be measured passes through assessment.The present invention can improve the safety in the automatic lane-change function real vehicle operational process of automatic Pilot, to play certain progradation to the development of automatic Pilot technology and industrialization while reducing emulation with field test amount.
Description
Technical field
The present invention relates to field of intelligent transportation technology, and in particular to a kind of automatic driving vehicle is based on worst global dangerous
The assessment method of automatic lane-change function in the case of degree.
Background technique
Automatic driving vehicle is developed so far a faced main problem is how to land, this not only needs further complete
Kind automatic Pilot technology, and solve the problems, such as how commercialized automatic Pilot is, it is often more important that how to ensure automatic driving vehicle
Safety behind upper road.Automatic lane-change function is most basic one of the function of automatic driving vehicle, but relative to traditional vehicle
Active safety system, since automatic lane-change function institute's duration is longer, and the number for the scene variable element being related to is significantly
Increase, while when speed is higher, faulty operation once occurs in automatic lane-change function, will bring biggish danger.Therefore, to certainly
The test assessment for moving function is not only more complicated, but also needs stringenter.As traditionally vehicle active safety is tested
Method tests the automatic lane-change function of automatic Pilot, in order to reach higher safety, needs to carry out the field of huge amount
Ground test, the time it is excessively huge with monetary cost and can not be practical.But still lack in existing research to automatic driving vehicle certainly
The theoretical research for moving the test assessment of function, is especially a lack of how research ensures to move certainly while reducing test volume
The safety of road function.
Summary of the invention
It is directed to the prior art, the present invention provides a kind of automatic lane-change function of vehicle based on worst global danger level search
Assessment method.
This method determines the functional boundary of automatic Pilot lane-change function to be measured first, while determining and to be measured driving automatically equipped with this
Sail the configuration of lane-change function vehicle sensors, on influential vehicle parameter of automatic lane-change function tool etc..It builds on this basis
The emulation platform of automatic driving vehicle model to be measured, model of place and automatic Pilot algorithm black-box model to be measured.Further, may be used
Targetedly place is carried out to the vehicle comprising sensor to test, and the parameter of vehicle emulation platform is adjusted to realize, it is ensured that
The accuracy of emulation platform is in certain range.
By the way that automatic lane-change function ODD corresponding with the function, (Operational Design Domain design can be grasped
Make range) and application scenarios analysis, determine each bound of parameter in scene.Initial parameter is selected according to lane-change scene analysis
It selects, the most dangerous scene problem of search is then converted into optimization problem and is solved.The input of the optimization problem is lane-change scene
Parameter and vehicle initial parameter, optimization aim are global danger level Sdan, global danger level SdanRelated Computational Methods are embedded into imitative
True platform obtains corresponding global danger level after the emulation platform that each run determines scenario parameters and vehicle initial parameter
Sdan.According to optimization algorithm to worst global danger level MS can be reacheddanScenario parameters and vehicle initial parameter searched
Rope, to obtain worst global danger level MSdanCorresponding vehicle lane-changing scene.Since the scene is corresponding automatic lane-change function
The worst scene that can be met within the scope of ODD, therefore the worst scene by that can cover it in difficulty to a certain extent
His scene to reduce field test amount, and demonstrates the performance of corresponding automatic lane-change function in extreme circumstances, improves
Safety on automatic driving vehicle behind road.
The technical solution used in the present invention is as follows: a kind of automatic lane-change function of vehicle based on worst global danger level search
Energy assessment method, this method comprises the following steps:
S1: building emulation platform, and loading vehicles model, model of place determine Vehicle dynamic parameters, formal parameter, biography
The key parameters such as sensor parameter, and including road parameters, weather parameters, traffic instruction parameter, interference number of vehicles, interference
The environmental parameter of each moment speed of vehicle etc.;
Further, it can be tested by place, above-mentioned modeling parameters are adjusted, it is ensured that the accuracy of emulation platform.
S2: by the analysis to automatic lane-change function, functional boundary and lane-change scene, the boundary of environmental parameter is determined.
S3: one initial lane-change scene of selection, and then global search is carried out to most dangerous lane-change scene, search for most dangerous scene
It is obtained in a manner of optimization problem solving, the input of the optimization problem is environmental parameter and lane-change function and boundary, and it is complete for exporting
Office danger level Sdan, wherein defining worst global danger level MSdanCorresponding scene is most dangerous lane-change scene.
S4: reproducing most dangerous lane-change scene by actual place, verify to automatic Pilot lane-change function to be measured, if
Automatic Pilot lane-change function to be measured can meet service requirement under the most dangerous lane-change scene, then can assert the automatic Pilot to be measured
Lane-change function under other scenes also can maximum probability meet the requirements, it is believed that the automatic Pilot lane-change function to be measured passes through survey
It comments.
Further, global danger level S is solveddanMethod are as follows:
1) one Che-vehicle overall situation danger level S is defineddan-vm, wherein most cart-vehicle overall situation danger level MSdan-vmBy vehicle to be measured
With it is each interference vehicle each moment collide danger level maximum value determine, be expressed as follows:
MSdan-vm=max { sup (Sdan-1(t)),sup(Sdan-2(t)),...,sup(Sdan-i(t))} (1)
Wherein, Sdan-iIt (t) is the danger level of vehicle to be measured and i-th interference vehicle in t moment, sup characterization takes Sdan-i
(t) dividing value on, Sdan-i(t) be expressed as follows:
In formula, (xi(t),yi(t))、ViIt (t) is the position coordinates and speed of i-th vehicle t moment in interference vehicle, (xv
(t),yv(t))、VvIt (t) is the position coordinates and speed of vehicle t moment to be measured, W is vehicle width, WLFor road width, CVFor
One constant coefficient, be arranged the coefficient purpose be ensure (- | xv(t)-xi(t)|+(Vi(t)-Vv(t))+CV) and (- | xv(t)-xi
(t)|-(Vi(t)-Vv(t))+CV) it is not negative value, CVTake a sufficiently large value;
2) a deviation overall situation danger level S is defineddan-lm, wherein maximum deviation overall situation danger level MSdan-lmBy to
Measuring car is determined relative to lane boundary in the maximum value that each moment deviates danger level, is expressed as follows:
MSdan-lm=max { sup (Sdan-llane(t)),sup(Sdan-rlane(t))} (4)
In formula, Sdan-llane(t) and Sdan-rlaneIt (t) is respectively vehicle to be measured and any related lane left and right side boundary
The deviation danger level of t moment, Sdan-llane(t) and Sdan-rlane(t) be expressed as follows:
In formula, ylFor the y-coordinate on the lane l left/right boundary, ClFor a constant coefficient, the purpose which is set be ensure (- |
xv(t)-xi(t)|+Vv(t)+Cl) it is not negative value, ClTake a sufficiently large value;
3) global danger level S is defineddan, wherein worst overall situation danger level MSdanIt is expressed as follows:
MSdan=max { MSdan-vm,MSdan-lm} (6)。
The invention is characterized in that the technical program, it can be while reducing emulation with field test amount, raising is driven automatically
The safety in automatic lane-change function real vehicle operational process is sailed, to play centainly to the development of automatic Pilot technology and industrialization
Progradation.
Detailed description of the invention
Attached drawing is used to provide further understanding of the present invention, and constitutes part of specification, with reality of the invention
It applies example and is used together to explain the present invention, be not construed as limiting the invention.In the accompanying drawings:
Fig. 1 shows the lane-change scene of automatic lane-change function to be measured of the invention;
Fig. 2 shows the foundation of lane-change scene coordinate system of the invention and each vehicle initial position parameters;
Fig. 3 shows the automatic search platform and search routine figure of the invention based on worst global danger level.
Specific embodiment
In the following description, a large amount of concrete details are given so as to provide a more thorough understanding of the present invention.So
And it is obvious to the skilled person that the present invention may not need one or more of these details and be able to
Implement.With reference to the accompanying drawing, illustrate embodiments of the present invention.
Test Test and analysis will be carried out by the two-way traffic lane-change function to a kind of automatic driving vehicle herein, to the present invention
The method of proposition is further detailed and verifies.
Two-way traffic lane-change function description is as shown in Figure 1 with lane-change scene.In figure, VUT indicates automatic driving vehicle to be measured,
C11、C12、C21、C22To interfere vehicle, other vehicles in vehicle original to be measured lane and target lane are represented, due to these vehicles
Certain interference effect will be played to the lane-change process of tested vehicle, so claiming these vehicles to be referred to as interferes vehicle.
The target of VUT is to change to C from speed way11、C12Between low speed carriage way position.It is relevant to lane-change function its
Its factor is Vehicle body parameter, including vehicle width W, Vehicle length L, the wide W in laneL, it is assumed here that the formal parameter of 5 vehicles
Unanimously.Meanwhile the speed of 5 vehicles is respectively VT、VC11、VC12、VC21And VC22, VTFor vehicle speed to be measured, other are corresponding license number
Speed.
In order to further indicate that vehicle relative positional relationship, the coordinate system for being fixed on road as shown in Figure 2 is established.Figure
In (xi,yi) be i vehicle position coordinates, (x0i,y0i) it is initial position of the i vehicle at the t=0 moment, remember VUT at the t=0 moment
Initial position is (0,0).
According to the lane-change duration under the scene, the speed change parameter of each vehicle was defined as in 5 seconds here, can be changed
The vehicle acceleration a that parameter is each secondiAnd the initial position of t=0 moment vehicle.Due to vehicle degree of danger generally with
Speed is directly proportional, determines that the initial velocity of vehicle is the speed upper limit value in corresponding lane here.According to above-mentioned definition, vehicle is interfered
The dynamic environment parameter of i are as follows:
[x0i v0i ai| t=0s ai| t=1s ai| t=2s ai| t=3s ai| t=4s ai| t=5s]
Therefore, under this lane-change scene, four interference vehicles include 28 dynamic environment parameters altogether.In order to quickly search
Rigging has the scene of worst global danger level, is built here comprising model of place, vehicle dynamic model and automatic lane-change to be measured
The emulation platform of algorithm black box.It is flat that emulation is integrated to according to the global danger level computing module of above-mentioned formula (1)~(6) foundation simultaneously
Platform.Under each group of environmental parameter, after being run in emulation platform, it can be calculated by global danger level computing module and correspond to this
The global danger level of one group of parameter.Using above-mentioned 28 environmental parameters as optimized variable, global danger level SdanIt is maximum that (maximum is corresponding
Worst scene) it is optimization aim, in conjunction with optimization tool, searched for automatically with the worst global danger level of emulation platform building jointly flat
Platform, such as Fig. 3.
The most dangerous scene for automatic lane-change function to be measured is obtained by above-mentioned automatic search platform, by surveying in place
Identical parameters scene is built in examination, and carries out place experiment of the automatic lane-change vehicle to be measured under the scene, obtains automatic lane-change
The worst or relatively poor scene that may occur in the operation of function real vehicle, as the worst scene can meet to automatic lane-change function
Service requirement, then can determine whether the automatic lane-change function other scenes operation can maximum probability meet the requirements, it is believed that should
Automatic lane-change function passes through assessment.
The foregoing is only a preferred embodiment of the present invention, but scope of protection of the present invention is not limited thereto,
In the technical scope disclosed by the present invention, any changes or substitutions that can be easily thought of by anyone skilled in the art,
It should be covered by the protection scope of the present invention.
Claims (4)
1. a kind of automatic lane-change Function Appraising method of vehicle based on worst global danger level search, this method includes following step
It is rapid:
S1: building emulation platform, establishes auto model, model of place;
S2: by the analysis of functional boundary and lane-change scene to automatic lane-change function to be measured, the boundary of environmental parameter is determined;
S3: one initial lane-change scene of selection, and determine initial environment parameter, and then the overall situation is carried out to most dangerous lane-change scene and is searched
Rope, search is with worst global danger level MSdanAs output;
S4: most dangerous lane-change scene is reproduced by actual place, automatic Pilot lane-change function to be measured is verified, if to be measured
Automatic Pilot lane-change function can meet service requirement under most dangerous lane-change scene, then assert the automatic Pilot lane-change function to be measured
Pass through assessment.
2. the vehicle automatic lane-change Function Appraising method according to claim 1 based on worst global danger level search,
It is characterized in that:
Worst overall situation danger level MSdanWith most cart-vehicle overall situation danger level MSdan-vmWith maximum deviation overall situation danger level
MSdan-lmMaximum value in the two is as selected value.
3. the vehicle automatic lane-change Function Appraising method according to claim 2 based on worst global danger level search,
It is characterized in that: most cart-vehicle overall situation danger level MSdan-vmMethod for solving are as follows:
Define Che-vehicle overall situation danger level Sdan-vm, most cart-vehicle overall situation danger level MSdan-vmBy vehicle to be measured and each interference vehicle
Each moment collide danger level maximum value determine, be expressed as follows:
MSdan-vm=max { sup (Sdan-1(t)),sup(Sdan-2(t)),...,sup(Sdan-i(t))}(1)
In formula, Sdan-iIt (t) is the danger level of vehicle to be measured and i-th interference vehicle in t moment, sup characterization takes Sdan-i(t) upper bound
Value, Sdan-i(t) be expressed as follows:
In formula, (xi(t),yi(t))、ViIt (t) is the position coordinates and speed of i-th vehicle t moment in interference vehicle, (xv(t),yv
(t))、VvIt (t) is the position coordinates and speed of vehicle t moment to be measured, W is vehicle width, WLFor road width, CVOften it is for one
Number, it is ensured that (- | xv(t)-xi(t)|+(Vi(t)-Vv(t))+CV) and (- | xv(t)-xi(t)|-(Vi(t)-Vv(t))+CV) be not
Negative value.
4. the vehicle automatic lane-change Function Appraising method according to claim 2 or 3 based on worst global danger level search,
It is characterized by: maximum deviation overall situation danger level MSdan-lmMethod for solving are as follows:
Define deviation overall situation danger level Sdan-lm, middle maximum deviation overall situation danger level MSdan-lmIt is opposite by vehicle to be measured
It determines, is expressed as follows in the maximum value that each moment deviates danger level in lane boundary:
MSdan-lm=max { sup (Sdan-llane(t)),sup(Sdan-rlane(t))} (4)
In formula, Sdan-llane(t) and Sdan-rlane(t) be respectively vehicle to be measured to any related lane left and right side boundary t when
The deviation danger level at quarter, Sdan-llane(t) and Sdan-rlane(t) be expressed as follows:
In formula, (xi(t),yi(t))、ViIt (t) is the position coordinates and speed of i-th vehicle t moment in interference vehicle, (xv(t),yv
(t))、VvIt (t) is the position coordinates and speed of vehicle t moment to be measured, W is vehicle width, WLFor road width, ylFor the lane l
The corresponding y-coordinate in left/right boundary, ClFor a constant coefficient, it is ensured that (- | xv(t)-xi(t)|+Vv(t)+Cl) it is not negative value.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910255614.7A CN110008577B (en) | 2019-04-01 | 2019-04-01 | Vehicle automatic lane changing function evaluation method based on worst global risk degree search |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910255614.7A CN110008577B (en) | 2019-04-01 | 2019-04-01 | Vehicle automatic lane changing function evaluation method based on worst global risk degree search |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110008577A true CN110008577A (en) | 2019-07-12 |
CN110008577B CN110008577B (en) | 2020-12-11 |
Family
ID=67169283
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910255614.7A Expired - Fee Related CN110008577B (en) | 2019-04-01 | 2019-04-01 | Vehicle automatic lane changing function evaluation method based on worst global risk degree search |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110008577B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111310302A (en) * | 2020-01-16 | 2020-06-19 | 中国信息通信研究院 | Test scene generation method and device |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20180061236A1 (en) * | 2015-03-18 | 2018-03-01 | Nec Corporation | Driving control device, driving control method, and vehicle-to-vehicle communication system |
CN108491600A (en) * | 2018-03-12 | 2018-09-04 | 郑州大学 | A kind of vehicle movement emulation mode based on people's this autonomy traffic |
CN108519773A (en) * | 2018-03-07 | 2018-09-11 | 西安交通大学 | The paths planning method of automatic driving vehicle under a kind of structured environment |
CN109353337A (en) * | 2018-08-10 | 2019-02-19 | 重庆大数据研究院有限公司 | A kind of intelligent vehicle lane-change stage collision probability safety predicting method |
-
2019
- 2019-04-01 CN CN201910255614.7A patent/CN110008577B/en not_active Expired - Fee Related
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20180061236A1 (en) * | 2015-03-18 | 2018-03-01 | Nec Corporation | Driving control device, driving control method, and vehicle-to-vehicle communication system |
CN108519773A (en) * | 2018-03-07 | 2018-09-11 | 西安交通大学 | The paths planning method of automatic driving vehicle under a kind of structured environment |
CN108491600A (en) * | 2018-03-12 | 2018-09-04 | 郑州大学 | A kind of vehicle movement emulation mode based on people's this autonomy traffic |
CN109353337A (en) * | 2018-08-10 | 2019-02-19 | 重庆大数据研究院有限公司 | A kind of intelligent vehicle lane-change stage collision probability safety predicting method |
Non-Patent Citations (2)
Title |
---|
DANCHEN ZHAO 等: "Autonomous Driving Simulation for Unmanned Vehicles", 《2015 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION》 * |
杨刚 等: "基于车车通信的车辆并行协同自动换道控制", 《公路交通科技》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111310302A (en) * | 2020-01-16 | 2020-06-19 | 中国信息通信研究院 | Test scene generation method and device |
CN111310302B (en) * | 2020-01-16 | 2022-06-17 | 中国信息通信研究院 | Test scene generation method and device |
Also Published As
Publication number | Publication date |
---|---|
CN110008577B (en) | 2020-12-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113010967B (en) | Intelligent automobile in-loop simulation test method based on mixed traffic flow model | |
CN105718710B (en) | A kind of driving behavior analysis method and apparatus | |
CN111309600A (en) | Virtual scene injection automatic driving test method and electronic equipment | |
CN103256939B (en) | Intelligent vehicle utilizes and becomes the method that granularity right of way radar map carries out information fusion | |
CN109100155A (en) | A kind of unmanned vehicle is in ring high-speed simulation test macro and method | |
CN102788704B (en) | Based on vehicle handling stability detection system and the detection method of pilot model | |
CN107992016A (en) | A kind of automatic driving vehicle analog detection method | |
CN107991898A (en) | A kind of automatic driving vehicle simulating test device and electronic equipment | |
CN109992884A (en) | Automatic driving vehicle Function Appraising method based on worst scene search | |
CN106840097A (en) | A kind of road grade method of estimation based on adaptive extended kalman filtering | |
CN109948289B (en) | Vehicle autonomous parking function evaluation method based on maximum global encroachment rate search | |
RU2626424C1 (en) | Device and method for evaluation of vehicle attitude angle and position | |
CN109726426A (en) | A kind of Vehicular automatic driving virtual environment building method | |
CN110803172B (en) | Modeling method of brake system of automatic driving vehicle and vehicle brake system | |
CN105835890A (en) | Vehicle sensor compensation | |
CN113252366A (en) | Automatic parking evaluation method and device for vehicle, vehicle and storage medium | |
CN106546260A (en) | A kind of correcting method and system of traverse measurement data | |
CN101118500A (en) | Software emulation method of self-determination driving vehicle running process | |
CN110775066B (en) | Modeling method of steering system of automatic driving vehicle and vehicle steering system | |
CN110008577A (en) | The automatic lane-change Function Appraising method of vehicle based on worst global danger level search | |
CN109910008B (en) | Obstacle avoidance early warning system and method for data type laser radar robot | |
CN109871659A (en) | A kind of emulation mode and system of ultrasonic radar | |
CN111123729A (en) | Method and device for vehicle driving simulation optimization, storage medium and terminal equipment | |
CN110160802A (en) | Automobile climbing vehicle speed measurement method and electronic equipment | |
Clark et al. | Simulation-based reduction of operational and cybersecurity risks in autonomous vehicles |
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 | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20201211 |