CN111915888B - Method for calculating complexity of traffic participants in automatic driving test scene - Google Patents
Method for calculating complexity of traffic participants in automatic driving test scene Download PDFInfo
- Publication number
- CN111915888B CN111915888B CN202010672885.5A CN202010672885A CN111915888B CN 111915888 B CN111915888 B CN 111915888B CN 202010672885 A CN202010672885 A CN 202010672885A CN 111915888 B CN111915888 B CN 111915888B
- Authority
- CN
- China
- Prior art keywords
- type
- longitudinal
- traffic
- complexity
- traffic participant
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000012360 testing method Methods 0.000 title claims abstract description 71
- 238000000034 method Methods 0.000 title claims abstract description 25
- 238000005070 sampling Methods 0.000 claims abstract description 62
- 238000010845 search algorithm Methods 0.000 claims abstract description 8
- 238000005457 optimization Methods 0.000 claims abstract description 7
- 230000033001 locomotion Effects 0.000 claims description 16
- 230000001133 acceleration Effects 0.000 claims description 14
- 238000004364 calculation method Methods 0.000 claims description 13
- 238000013461 design Methods 0.000 abstract description 3
- 238000011156 evaluation Methods 0.000 description 5
- 238000011160 research Methods 0.000 description 4
- 238000011002 quantification Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
-
- 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
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
- G08G1/0129—Traffic data processing for creating historical data or processing based on historical data
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention relates to a method for calculating the complexity of traffic participants in an automatic driving test scene, which comprises the following steps: s1: obtaining dynamic parameter threshold values R of all types of traffic participants in test scenetype(ii) a S2: according to the initial state of each traffic participant and the dynamic parameter threshold value RtypeAnd calculating to obtain a corresponding longitudinal sampling distance set Stype(ii) a S3: obtaining a longitudinal sampling distance set S through an optimization search algorithmtypeThe optimal longitudinal control quantity corresponding to each sampling distance; s4: calculating the reachable domain omega of each traffic participant within the predicted time t according to the sampling distance and the optimal longitudinal control quantitytype(ii) a S5: according to the reachable domain omega of each traffic participanttypeCompared with the prior art, the method and the device for calculating the complexity H of the traffic participants in the test scene provide basis for the state and parameter design of the dynamic traffic participants in the automatic driving test scene, and have the advantages of improving the test efficiency of the automatic driving automobile and the like.
Description
Technical Field
The invention relates to the field of testing and evaluation in the technology of automatically driving automobiles, in particular to a method for calculating the complexity of a traffic participant in an automatically driving test scene.
Background
In recent years, the technology of autonomous driving cars is one of the hot spots of current research, both in academic and industrial areas. However, with the rapid development of the automatic driving technology, the frequent occurrence of dangerous accidents gradually highlights some safety-oriented problems. One of the means to solve these problems is to perform accurate, sufficient, and complete tests on an autonomous vehicle. A complete test process, which comprises the steps of carrying out multiple tests on a tested object in the same environment, quantizing the test result by using standard evaluation indexes, and then feeding back to a test scene designer; and the designer iterates and updates the arrangement and the layout of the test scene according to the evaluation result so as to accelerate the test process of the automatic driving automobile and improve the test efficiency.
However, at present, the evaluation of the automatic driving test scenario lacks a unified standard, which includes the complexity evaluation of a typical test scenario of each vehicle driven by an individual vehicle, especially the related research of a complexity quantification method of dynamic traffic participants in the test scenario, resulting in poor lateral comparability of the dynamic traffic participants and the test scenario under different parameter configurations. Therefore, the research of the complexity quantification of the dynamic traffic participants in the automatic driving test scene is a big vacancy in the field.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provide a method for calculating the complexity of traffic participants in an automatic driving test scene, fill the blank of the research field of the complexity quantification of dynamic traffic participants in the automatic driving test scene, provide a basis for the design and arrangement of the dynamic traffic participants in the automatic driving test scene, enable different complexity scenes facing the same tested object or different tested objects under the same complexity to have a transverse contrast condition, improve the efficiency of testing an automatic driving automobile and accelerate the landing of the automatic driving automobile.
The purpose of the invention can be realized by the following technical scheme:
a method for calculating the complexity of traffic participants in an automatic driving test scene comprises the following steps:
s1: obtaining dynamic parameter threshold values R of all types of traffic participants in test scenetype；
S2: according to the initial state of each traffic participant and the dynamic parameter threshold value RtypeAnd calculating to obtain a corresponding longitudinal sampling distance set Stype；
S3: obtaining a longitudinal sampling distance set S through an optimization search algorithmtypeThe optimal longitudinal control quantity corresponding to each sampling distance;
s4: calculating the reachable domain omega of each traffic participant within the predicted time t according to the sampling distance and the optimal longitudinal control quantitytype；
S5: according to the reachable domain omega of each traffic participanttypeAnd calculating the complexity H of the traffic participants of the test scene.
Further, the kinetic parameter threshold value RtypeThe expression of (a) is:
Rtype＝(A,D)
wherein type is the type of the traffic participant, A is the longitudinal maximum acceleration of the traffic participant, and D is the longitudinal minimum deceleration of the traffic participant.
Further preferably, the initial state of the transportation participant comprises an initial position S0Initial velocity v0And an initial acceleration a0。
Further, the longitudinal control amount includes acceleration and deceleration of the vehicle in the longitudinal direction.
Still further preferably, said set of longitudinal sampling distances S is characterized bytypeIncluding the initial state and dynamic parameter threshold value R of the corresponding traffic participant within the prediction time ttypeThe reachable position and the initial position S0The longitudinal distance of (a).
Further, the step S3 specifically includes:
s31: initializing the jth sampling distance
Longitudinal control amount of
S32: calculating the longitudinal control quantity of the corresponding traffic participant
Next, the longitudinal distance S that can be reached within the time t is predictedi；
S33: judging the calculated longitudinal distance SiWhether or not to equal the sampling distance
Corresponding preset longitudinal distance SjIf yes, outputting the current longitudinal control quantity
As the sampling distance
Is controlled by the optimum longitudinal control quantity Uj *Otherwise, go to step S34;
s34: to make the sampling distance
Optimizing search for direction with maximized transverse distance X, and updating longitudinal control quantity
And returns to perform step S32;
s35: and repeatedly executing the steps S31-S34 until the optimal longitudinal control quantity corresponding to all the sampling distances is obtained.
Further, the optimal longitudinal control quantity comprises n longitudinal discrete control quantities a in the prediction time tiThe step S4 specifically includes:
s41: for longitudinal discrete control quantity aiIterative search is carried out, and the jth sampling distance of the corresponding traffic participant is calculated
Next, the maximum lateral distance X that can be achieved by the motion trajectory within the prediction time t is predictedmaxDetermining the track boundary at the sampling distance;
s42: repeating the step S41 to obtain the track boundary at all the sampling distancesAnd obtaining the reachable domain omega of the corresponding traffic participant within the predicted time t according to the track boundary under each sampling distancetype；
S43: repeating the steps S41-S42 to obtain the reachable domain omega of each traffic participant within the predicted time ttype。
Further, the step S5 specifically includes:
s51: calculate reachable Domain omegatypeArea and the reachable region omega selected by the corresponding traffic participant is obtainedtypeThe probability of a certain track is recorded as the track probability
S52: calculating motion trail complexity H of each traffic participant in test scene based on information entropy theorytxpe；
S53: judging whether the number of the traffic participants in the test scene is more than one, if so, executing a step S54, otherwise, outputting the motion trail complexity H of the traffic participantstypeThe complexity H of the traffic participants as a test scenario;
s54: the reachable domain omega of all traffic participants in the test scenetypeSumming the sets and subtracting the reachable domain omegatypeObtaining the total reachable domain of the test scene according to the area occupied by the sizes of all the traffic participants;
s55: obtaining the total complexity H of the test scene by using the total reachable domain calculationtotalTraffic participant complexity H as a test scenario.
Furthermore, the motion trail complexity H of the traffic participanttypeThe calculation formula of (A) is as follows:
wherein,
in the reachable domain omega for traffic participantstypeSelecting from the inner m tracksProbability of x trajectory.
The total complexity HtotalThe calculation formula of (A) is as follows:
wherein,
the reachable domain of the tau-th traffic participant is shown, and M is the number of the traffic participants in the test scene.
Preferably, the type of transportation participant comprises a car, a bus, a motorcycle, a bicycle and a pedestrian.
Compared with the prior art, the invention has the following advantages:
1) the method for quantizing the complexity of the dynamic traffic participants in the automatic driving test scene can fill the gap of the method for quantizing the complexity of the current automatic driving test scene, provides a basis for the design and arrangement of the dynamic traffic participants in the automatic driving test scene, enables different complexity scenes facing the same tested object or different tested objects under the same complexity to have transverse comparison conditions, and improves the efficiency of testing the automatic driving automobile, so that the problems of a tested automobile system can be quickly found, the iterative update of the automatic driving automobile is promoted, and the landing of the automatic driving automobile is accelerated;
2) the method utilizes the information entropy theory to calculate the motion trail complexity of the traffic participants in the automatic driving test scene, and improves the accuracy of the calculation of the test scene complexity;
3) the method is based on vehicle kinematics, the transverse and longitudinal motions of the vehicle are described by using the control quantity, and the calculated vehicle reachable region can relatively comprehensively reflect the driving possibility of the vehicle in a certain prediction time in the future based on the current state, so that the complexity calculation of the traffic participants determined by the reachable region is relatively reasonable;
4) the algorithm used by the invention is developed and completed on MATLAB, and can also be used for calculating the vehicle reachable domain and complexity under the determined vehicle control quantity obtained by the subsequent simulation of vehicle motion control.
Drawings
FIG. 1 is a schematic flow chart of the method of the present invention;
FIG. 2 is a schematic overall flow chart of the present invention;
FIG. 3 is a flow chart of an optimized search algorithm of the present invention;
FIG. 4 is a schematic diagram of reachable domains of traffic participants in a test scenario of an embodiment, where (4a) is the maximum reachable domain of a bus, and (4b) is the maximum reachable domain of a car;
FIG. 5 is a schematic diagram of the total reachable region of two vehicles in the test scenario of the embodiment.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments. It is to be understood that the embodiments described are only a few embodiments of the present invention, and not all embodiments. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, shall fall within the scope of protection of the present invention.
As shown in fig. 1, the present invention provides a method for calculating complexity of traffic participants in an automatic driving test scenario, comprising the following steps:
s1: obtaining dynamic parameter threshold values R of all types of traffic participants in test scenetype；
S2: according to the initial state of each traffic participant and the dynamic parameter threshold value RtypeAnd calculating to obtain a corresponding longitudinal sampling distance set Stype；
S3: obtaining a longitudinal sampling distance set S through an optimization search algorithmtypeThe optimal longitudinal control quantity corresponding to each sampling distance;
s4: calculating the reachable domain omega of each traffic participant within the predicted time t according to the sampling distance and the optimal longitudinal control quantitytype；
S5: according to the reachable domain omega of each traffic participanttypeCalculating traffic participant complexity H of test scenario。
(1) Step S1
S1: obtaining dynamic parameter threshold values R of all types of traffic participants in test scenetypeThreshold value R of kinetic parametertypeThe expression of (a) is:
Rtype＝(A,D)
wherein type is the type of the traffic participant, A is the longitudinal maximum acceleration of the traffic participant, and D is the longitudinal minimum deceleration of the traffic participant.
Firstly, determining the dynamic parameter threshold value range R of different types of traffic participantstype(A, D), wherein the dynamic parameters are the longitudinal maximum acceleration A and the minimum deceleration D of the traffic participant, namely the dynamic parameter threshold range RtypeThe set of (c) may be:
wherein A iscarFor maximum acceleration of the car, DcarFor minimum deceleration of cars, AbusMaximum acceleration of the passenger car, DbusFor minimum deceleration of passenger cars, AmotoFor maximum acceleration of the motorcycle, DmotoFor minimum deceleration of motorcycle, AbikeMaximum acceleration of the bicycle, DbikeFor minimum deceleration of a bicycle, the present invention takes into account the size and dynamics of the different traffic participants, classifying the traffic participants into types including, but not limited to, cars, passenger cars, motorcycles, bicycles, pedestrians, etc.
(2) Step S2
S2: according to the initial state of each traffic participant and the dynamic parameter threshold value RtypeAnd calculating to obtain a corresponding longitudinal sampling distance set Stype。
The initial state of the vehicle includes an initial position S0Initial velocity v0And an initial acceleration a0A set of determined longitudinal sampling distances StypeThe element in (1) is a sampling distance, namely the longitudinal distance between an initial position point and an reachable pointSeparation:
m is the number of sampling distances, and the distance interval between each sampling distance can be set to 1m or other values.
(3) Step S3-step S4
S3: obtaining a longitudinal sampling distance set S through an optimization search algorithmtypeCorresponding to each sampling distance.
S4: calculating the reachable domain omega of each traffic participant within the predicted time t according to the sampling distance and the optimal longitudinal control quantitytype。
Dynamic parameter threshold R from using an optimized search algorithmtypeSelecting within range at each sampling distance
Lower optimal longitudinal control quantity Uj *：
Wherein the optimum longitudinal control quantity Uj *Longitudinal discrete control quantity a contained in predicted time ti(i is 1,2, …, n), where n is the vertical discrete control quantity a in the prediction time tiThe number of (a) and each discrete control amount aiThe time interval therebetween is Δ t.
By controlling the discrete control quantity aiIterative search is carried out to calculate the sampling distance of the traffic participants within the predicted time t
The maximum lateral distance X of the track can be reached, further the reachable domain determined by the track boundary under the sampling distance is determined, and finally the reachable domain omega of the traffic participant in the prediction time t is obtainedtype：
The optimization search algorithm specifically comprises the following steps:
s31: initializing the jth sampling distance
Longitudinal control amount of
S32: calculating the longitudinal control quantity of the corresponding traffic participant
Next, the longitudinal distance S that can be reached within the time t is predictedi；
S33: judging the calculated longitudinal distance SiWhether or not to equal the sampling distance
Corresponding preset longitudinal distance SjIf yes, outputting the current longitudinal control quantity
As the sampling distance
Is controlled by the optimum longitudinal control quantity Uj *Otherwise, go to step S34;
s34: to make the sampling distance
Optimizing search for direction with maximized transverse distance X, and updating longitudinal control quantity
And returns to perform step S32;
s35: and repeatedly executing the steps S31-S34 until the optimal longitudinal control quantity corresponding to all the sampling distances is obtained.
The iterative search specifically comprises the following steps:
s41: for longitudinal discrete control quantity aiIterative search is carried out, and the jth sampling distance of the corresponding traffic participant is calculated
Next, the maximum lateral distance X that can be achieved by the motion trajectory within the prediction time t is predictedmaxDetermining the track boundary at the sampling distance;
s42: repeatedly executing the step S41 to obtain the track boundaries under all the sampling distances, and obtaining the reachable domain omega of the corresponding traffic participant within the prediction time t according to the track boundaries under all the sampling distancestype；
S43: repeating the steps S41-S42 to obtain the reachable domain omega of each traffic participant within the predicted time ttype。
(4) Step S5
S5: according to the reachable domain omega of each traffic participanttypeAnd calculating the complexity H of the traffic participants of the test scene.
For convenience of calculation, the trajectory of the traffic participant is considered to be in the reachable domain ΩtypeThe interior is uniformly distributed, and the vehicle selects the track probability of a certain track in the reachable domain
Wherein<Ωtype>To pass the reachable domain area of the traffic participant under the constraint of the kinetic parameters,
when the complexity H of the traffic participants of the test scene is calculated based on the information entropy theory, two situations are specifically included:
(1) only one traffic participant exists in the scene, and the motion trail complexity H of the traffic participanttypeI.e. the traffic participant complexity H of the test scenario.
At the moment, the probability parameter P in the information entropy calculation formula is the probability that the vehicle selects a certain track x in the reachable domain, namely
I.e. the complexity of the movement path of the traffic participant HtypeThe calculation formula of (A) is as follows:
(2) a plurality of traffic participants are arranged in the scene
At the moment, the reachable domains of M different traffic participants in the same scene need to be merged, the area occupied by the size of the traffic participants in the reachable domains is subtracted, the total reachable domain in the scene is obtained, and finally the total complexity H of the test scene is obtainedtotalComplexity H of traffic participants as test scenario, total complexity H of test scenariototalThe calculation formula of (A) is as follows:
wherein,
the reachable domain of the tau-th traffic participant is shown, and M is the number of the traffic participants in the test scene.
Example 1
In this embodiment, the traffic participant types in the test scenario include car (car) and bus (bus), and it is determined that the dynamic parameter thresholds of the car (car) and the bus (bus) are respectively:
threshold value R of dynamic parameter of carcar(-7.5,6.9), bus dynamics parameter threshold, Rbus＝(-6.0,4.0)。
The initial state parameter configuration of two kinds of vehicles is the same, specifically:
S0＝0；
a0＝0.5；(ms-2)
v0＝15；(ms-1)
then, calculating according to the initial state of the vehicle and the dynamic threshold range to obtain the small steamThe final longitudinal position of the full-force deceleration of the car is 15, the final longitudinal position of the full-force acceleration of the car is 162, and therefore the car sampling distance set ScarSample distance set S of 16,17, …,162 for a similarly large busbus＝{19,20,…,125}。
The optimal longitudinal control quantity corresponding to each sampling distance is obtained through optimization search, the coordinate parameter of the vehicle under each timestamp is determined according to the optimal longitudinal control quantity, the maximum reachable domain boundaries of the car and the bus obtained after connection of each coordinate point are respectively shown in fig. 4, and the reachable domain range is the range of the region which the vehicle can reach when the vehicle starts from the lane center line position of the lower lane and changes lanes to the upper lane.
In this embodiment, each longitudinal sampling distance
The determined reachable domain boundary is composed of 50 points, and the time interval between each coordinate point is 0.1s, so that the predicted time t is 5s in the embodiment, that is, the reachable domain of the vehicle in the future of 5s can be obtained.
After the sampling distance sets and the optimal longitudinal control quantity of the two vehicles are obtained, the reachable domain areas of the two vehicles can be calculated: calculating the reachable area omega for each longitudinal sampling distancetypeObtaining the maximum reachable domain area omega of the car through iterationcar＝830.5701m2Maximum reachable area omega of busbus＝616.2915m2And calculating the motion track complexity H of the carcar9.70BIT, motion trajectory complexity H of motor coachbus＝9.27BIT。
When the car and the bus are located in the same test scene, as shown in fig. 5, the total area of the reachable region in the test scene is the area of the reachable region of the car minus the area occupied by the bus, the size of the bus selected in this embodiment is 5990mm, the width is 2050mm, that is, the total reachable region area is Ωtotal＝818.2736m2Finally, the total complexity H of the scene is obtained through calculationtotal＝9.67BIT。
While the invention has been described with reference to specific embodiments, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (9)
1. A method for calculating the complexity of traffic participants in an automatic driving test scene is characterized by comprising the following steps:
s1: obtaining dynamic parameter threshold values R of all types of traffic participants in test scenetype；
S2: according to the initial state of each traffic participant and the dynamic parameter threshold value RtypeAnd calculating to obtain a corresponding longitudinal sampling distance set Stype；
S3: obtaining a longitudinal sampling distance set S through an optimization search algorithmtypeThe optimal longitudinal control quantity corresponding to each sampling distance;
s4: calculating the reachable domain omega of each traffic participant within the predicted time t according to the sampling distance and the optimal longitudinal control quantitytype；
S5: according to the reachable domain omega of each traffic participanttypeCalculating the complexity H of the traffic participants of the test scene;
wherein, step S5 specifically includes:
s51: calculate reachable Domain omegatypeArea and the reachable region omega selected by the corresponding traffic participant is obtainedtypeThe probability of a certain track is recorded as the track probability
S52: calculating motion trail complexity H of each traffic participant in test scene based on information entropy theorytype；
S53: judging whether the number of the traffic participants in the test scene is more than one, if so, executing a step S54, otherwise, outputting the motion trail complexity H of the traffic participantstypeAs test scenariosTraffic participant complexity H;
s54: the reachable domain omega of all traffic participants in the test scenetypeSumming the sets and subtracting the reachable domain omegatypeObtaining the total reachable domain of the test scene according to the area occupied by the sizes of all the traffic participants;
s55: obtaining the total complexity H of the test scene by using the total reachable domain calculationtotalTraffic participant complexity H as a test scenario.
2. The method as claimed in claim 1, wherein the dynamic parameter threshold R is a threshold of a dynamic parametertypeThe expression of (a) is:
Rtype＝(A，D)
wherein type is the type of the traffic participant, A is the longitudinal maximum acceleration of the traffic participant, and D is the longitudinal minimum deceleration of the traffic participant.
3. The method as claimed in claim 2, wherein the initial state of the traffic participant comprises an initial position S0Initial velocity v0And an initial acceleration a0。
4. The method of claim 1, wherein the longitudinal control variables comprise acceleration and deceleration of the vehicle in a longitudinal direction.
5. The method as claimed in claim 3, wherein the set of vertical sampling distances S is a set of vertical sampling distances StypeIncluding the initial state and dynamic parameter threshold value R of the corresponding traffic participant within the prediction time ttypeThe reachable position and the initial position S0The longitudinal distance of (a).
6. The method for calculating the complexity of the traffic participants in the automatic driving test scenario as claimed in claim 1, wherein the step S3 specifically includes:
s31: initializing the jth sampling distance
Longitudinal control amount of
S32: calculating the longitudinal control quantity of the corresponding traffic participant
Next, the longitudinal distance S that can be reached within the time t is predictedi；
S33: judging the calculated longitudinal distance SiWhether or not to equal the sampling distance
Corresponding preset longitudinal distance SjIf yes, outputting the current longitudinal control quantity
As the sampling distance
Is controlled by the optimum longitudinal control quantity Uj *Otherwise, go to step S34;
s34: to make the sampling distance
Optimizing search for direction with maximized transverse distance X, and updating longitudinal control quantity
And returns to perform step S32;
s35: and repeatedly executing the steps S31-S34 until the optimal longitudinal control quantity corresponding to all the sampling distances is obtained.
7. The method as claimed in claim 6, wherein the optimal longitudinal control quantity comprises n longitudinal discrete control quantities a within the predicted time tiThe step S4 specifically includes:
s41: for longitudinal discrete control quantity aiIterative search is carried out, and the jth sampling distance of the corresponding traffic participant is calculated
Next, the maximum lateral distance X that can be achieved by the motion trajectory within the prediction time t is predictedmaxDetermining the track boundary at the sampling distance;
s42: repeatedly executing the step S41 to obtain the track boundaries under all the sampling distances, and obtaining the reachable domain omega of the corresponding traffic participant within the prediction time t according to the track boundaries under all the sampling distancestype；
S43: repeating the steps S41-S42 to obtain the reachable domain omega of each traffic participant within the predicted time ttype。
8. The method as claimed in claim 1, wherein the complexity of the motion trajectory of the traffic participant is HtypeThe calculation formula of (A) is as follows:
wherein
In the reachable domain omega for traffic participantstypeProbability of selecting x track from the inner m tracks;
the above-mentionedTotal complexity H oftotalThe calculation formula of (A) is as follows:
wherein,
the reachable domain of the tau-th traffic participant is shown, and M is the number of the traffic participants in the test scene.
9. The method of claim 1, wherein the traffic participants are of the type selected from the group consisting of cars, coaches, motorcycles, bicycles, and pedestrians.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010672885.5A CN111915888B (en) | 2020-07-14 | 2020-07-14 | Method for calculating complexity of traffic participants in automatic driving test scene |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010672885.5A CN111915888B (en) | 2020-07-14 | 2020-07-14 | Method for calculating complexity of traffic participants in automatic driving test scene |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111915888A CN111915888A (en) | 2020-11-10 |
| CN111915888B true CN111915888B (en) | 2021-09-03 |
Family
ID=73280005
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010672885.5A Active CN111915888B (en) | 2020-07-14 | 2020-07-14 | Method for calculating complexity of traffic participants in automatic driving test scene |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111915888B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114299712B (en) * | 2021-11-26 | 2024-03-01 | 腾讯科技（深圳）有限公司 | Data processing method, device, equipment and readable storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103234763A (en) * | 2013-04-09 | 2013-08-07 | 北京理工大学 | System and method for quantitatively evaluating unmanned vehicles |
| CN106067247A (en) * | 2015-04-23 | 2016-11-02 | 福特全球技术公司 | Traffic complex degree is estimated |
| CN109284889A (en) * | 2018-07-19 | 2019-01-29 | 长安大学 | A kind of unmanned vehicle reaction agility evaluating method and device |
| CN110751847A (en) * | 2019-10-15 | 2020-02-04 | 清华大学 | Decision-making method and system for automatically driving vehicle behaviors |
| CN110992695A (en) * | 2019-12-13 | 2020-04-10 | 北京理工大学 | Intelligent vehicle urban intersection traffic decision multi-objective optimization model based on conflict resolution |
| CN111027430A (en) * | 2019-11-29 | 2020-04-17 | 西安交通大学 | Traffic scene complexity calculation method for intelligent evaluation of unmanned vehicles |
| CN111402587A (en) * | 2020-03-30 | 2020-07-10 | 上海汽车集团股份有限公司 | Test method and test system for intelligent driving |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10580301B2 (en) * | 2016-03-16 | 2020-03-03 | Kiomars Anvari | Moving vehicles in a smart environment |
| CN110553853B (en) * | 2019-08-06 | 2020-11-20 | 清华大学 | Automatic driving function test and evaluation method based on poor scene search under field |
-
2020
- 2020-07-14 CN CN202010672885.5A patent/CN111915888B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103234763A (en) * | 2013-04-09 | 2013-08-07 | 北京理工大学 | System and method for quantitatively evaluating unmanned vehicles |
| CN106067247A (en) * | 2015-04-23 | 2016-11-02 | 福特全球技术公司 | Traffic complex degree is estimated |
| CN109284889A (en) * | 2018-07-19 | 2019-01-29 | 长安大学 | A kind of unmanned vehicle reaction agility evaluating method and device |
| CN110751847A (en) * | 2019-10-15 | 2020-02-04 | 清华大学 | Decision-making method and system for automatically driving vehicle behaviors |
| CN111027430A (en) * | 2019-11-29 | 2020-04-17 | 西安交通大学 | Traffic scene complexity calculation method for intelligent evaluation of unmanned vehicles |
| CN110992695A (en) * | 2019-12-13 | 2020-04-10 | 北京理工大学 | Intelligent vehicle urban intersection traffic decision multi-objective optimization model based on conflict resolution |
| CN111402587A (en) * | 2020-03-30 | 2020-07-10 | 上海汽车集团股份有限公司 | Test method and test system for intelligent driving |
Non-Patent Citations (2)
| Title |
|---|
| 基于场景的自动驾驶虚拟测试研究进展;朱冰 等;《中国公路学报》;20190630;第32卷(第6期);第1-19页 * |
| 基于模糊熵的混沌交通流复杂性分析;李一凡;《中国优秀硕士学位论文全文数据库》;20190215(第2期);第19-27页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111915888A (en) | 2020-11-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110834644B (en) | Vehicle control method and device, vehicle to be controlled and storage medium | |
| CN108919795B (en) | Automatic driving automobile lane change decision method and device | |
| CN111746559B (en) | Method and system for predicting lane changing intention of front vehicle | |
| CN110843789B (en) | Vehicle lane change intention prediction method based on time sequence convolution network | |
| CN111667512B (en) | Multi-target vehicle track prediction method based on improved Kalman filtering | |
| CN112078592B (en) | Method and device for predicting vehicle behavior and/or vehicle track | |
| CN108877213A (en) | The recognition methods and system of traffic accident | |
| CN113239986A (en) | Training method and device for vehicle track evaluation network model and storage medium | |
| CN111915888B (en) | Method for calculating complexity of traffic participants in automatic driving test scene | |
| CN105469052A (en) | Vehicle detection and tracking method and device | |
| CN112258893A (en) | Intelligent vehicle lane change collision probability assessment method based on track prediction | |
| CN115279643A (en) | On-board active learning method and apparatus for training a perception network of an autonomous vehicle | |
| CN115523934A (en) | Vehicle track prediction method and system based on deep learning | |
| CN114446046A (en) | LSTM model-based weak traffic participant track prediction method | |
| CN110843767A (en) | Spline theory-based automatic parking method and system | |
| CN117392855A (en) | Vehicle overrun overload probability identification method and system based on satellite positioning data | |
| CN110696828B (en) | Forward target selection method and device and vehicle-mounted equipment | |
| Benterki et al. | Driving intention prediction and state recognition on highway | |
| CN109670085B (en) | Method for judging vehicle collision accident grade | |
| CN116278771A (en) | Vehicle energy consumption prediction method, system and equipment | |
| CN115204755B (en) | Service area access rate measuring method and device, electronic equipment and readable storage medium | |
| CN115451987A (en) | Path planning learning method for automatic driving automobile | |
| CN112406874B (en) | Electric automobile remote charging auxiliary decision-making method | |
| CN117445901A (en) | Parking path planning method and path node prediction model training method | |
| CN116223056B (en) | Virtual collision test method, apparatus, device, storage medium, and program product |
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 |