CN113326638A - Method and device for determining automatic driving test scene - Google Patents
Method and device for determining automatic driving test scene Download PDFInfo
- Publication number
- CN113326638A CN113326638A CN202110883759.9A CN202110883759A CN113326638A CN 113326638 A CN113326638 A CN 113326638A CN 202110883759 A CN202110883759 A CN 202110883759A CN 113326638 A CN113326638 A CN 113326638A
- Authority
- CN
- China
- Prior art keywords
- scene
- probability
- parameter
- historical
- sample
- 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
Images
Classifications
-
- 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
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
- G01C21/34—Route searching; Route guidance
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C25/00—Manufacturing, calibrating, cleaning, or repairing instruments or devices referred to in the other groups of this subclass
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Manufacturing & Machinery (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
Abstract
The application discloses a method and a device for determining an automatic driving test scene, wherein the method comprises the following steps: determining K scene parameters and a cumulative distribution function corresponding to each scene parameter according to a historical scene library; carrying out N equal division on the probability interval of the cumulative distribution function corresponding to each scene parameter to obtain N probability sub-layers; randomly extracting M scene parameter probability sample points from the probability sublayer corresponding to each scene parameter according to a preset matrix R; according to each scene parameter probability sample point, K scene parameter values corresponding to K scene parameters in the ith target scene sample are obtained, the K scene parameter values are combined to form the ith target scene sample, through the process, the whole scene space can be better covered with fewer scenes, the load of simulation is reduced, and the efficiency of algorithm evaluation is improved.
Description
Technical Field
The application relates to the technical field of automatic driving, in particular to a method and a device for determining an automatic driving test scene.
Background
Automatic driving relies on the cooperative cooperation of artificial intelligence, visual computing, radar, monitoring devices and global positioning systems, allowing computers to operate motor vehicles automatically and safely without any human active operation.
In order to ensure the driving safety of the automatic driving vehicle, the automatic driving scene library is required to be utilized to carry out simulation evaluation on the automatic driving algorithm, the performance of the automatic driving algorithm is evaluated, all scene parameters are required to be traversed under all scenes, which cannot be achieved theoretically, in the prior art, the scene parameters are randomly extracted from the automatic driving scene library, and then the randomly extracted scene parameters are combined to form a new test scene, so that the automatic driving algorithm is simulated and trained.
Because a large number of test scenes are needed to test the automatic driving algorithm, a scene parameter group needs to be repeatedly extracted to synthesize a new test scene, but the test scenes of randomly extracted scene parameter combinations have repeated or similar situations and cannot cover all automatic driving test participation, namely the determined efficiency of the existing automatic driving test scenes is low.
Disclosure of Invention
The embodiment of the application aims to provide a method and a device for determining an automatic driving test scene, wherein the method can more comprehensively cover the whole scene space under the condition of a small number of test scene samples, and the efficiency of the determined automatic driving test scene is increased.
In order to achieve the above purpose, the embodiments of the present application employ the following technical solutions:
in a first aspect, an embodiment of the present application provides a method for determining an automatic driving test scenario, where the method for determining includes:
determining K scene parameters and a cumulative distribution function corresponding to each scene parameter according to a historical scene library;
n equal division is carried out on the probability interval of the cumulative distribution function corresponding to each scene parameter to obtain N probability sub-layers;
randomly extracting M scene parameter probability sample points from a probability sublayer corresponding to each scene parameter according to a preset matrix R;
acquiring K scene parameter values corresponding to K scene parameters in the ith target scene sample according to each scene parameter probability sample point, and combining the K scene parameter values to form the ith target scene sample;
and determining i target scene samples, and forming a target automatic driving field library according to the i target scene samples.
Optionally, the determining method further includes:
randomly extracting N historical scene samples from a historical scene library, wherein each historical scene comprises K historical scene parameters;
and constructing a preset matrix R according to the N historical scene samples and the corresponding K historical scene parameters.
Optionally, constructing a preset matrix R according to the N historical scene samples and the corresponding K historical scene parameters includes:
and acquiring a historical scene parameter K from the K historical scene parameters, and determining the values of the historical scene parameter K in the N historical scene samples respectively.
Constructing an NxK preset matrix R by N historical scene samples according to the preset sequence of the values of the historical scene parameters K, wherein each element R in the preset matrix RikAnd the sequence number represents the sequence number of the kth historical scene parameter in the ith historical scene sample in the historical scene samples after the kth historical scene parameter is sequenced in the N historical scene samples according to the preset sequence.
Optionally, according to any one of the above determining methods for an automatic driving test scenario, according to the preset matrix R, randomly extracting M scenario parameter probability sample points from the probability sublayer corresponding to each scenario parameter includes:
acquiring a sequencing serial number of each scene parameter in the ith historical scene sample point according to the preset matrix R;
determining a probability sublayer corresponding to each scene parameter in the ith target scene sample according to the sequencing serial number of each scene parameter;
and randomly extracting M scene parameter probability sample points from the probability sublayer corresponding to each scene parameter.
Optionally, obtaining K scene parameter values corresponding to K scene parameters in the ith target scene sample according to each scene parameter probability sample point includes:
and performing corresponding inverse cumulative distribution function conversion on each scene parameter probability sample point to obtain K scene parameter values corresponding to the K scene parameters respectively, wherein the K scene parameter values are numerical values corresponding to the K scene parameters in the ith target scene sample.
In a second aspect, an embodiment of the present application further provides an apparatus for determining an automatic driving test scenario, where the apparatus includes:
the first determining module is used for determining K scene parameters and a cumulative distribution function corresponding to each scene parameter according to the historical scene library;
the first processing module is used for carrying out N equal division on the probability interval of the cumulative distribution function corresponding to each scene parameter to obtain N probability sublayers;
the first sampling module is used for randomly extracting M scene parameter probability sample points from a probability sublayer corresponding to each scene parameter according to a preset matrix R;
and the acquisition module is used for acquiring K scene parameter values corresponding to K scene parameters in the ith target scene sample according to each scene parameter probability sample point, and the ith target scene sample is formed by combining the K scene parameter values.
Optionally, the determining means further comprises:
the second sampling module is used for randomly extracting N historical scene samples from a historical scene library, and each historical scene comprises K historical scene parameters;
and the second construction module is used for constructing a preset matrix R according to the N historical scene samples and the corresponding K historical scene parameters.
Optionally, the determining means further comprises:
and the second processing module is used for performing corresponding inverse cumulative distribution function conversion on each scene parameter probability sample point to obtain K scene parameter values corresponding to the K scene parameters respectively, wherein the K scene parameter values are numerical values corresponding to the K scene parameters in the ith target scene sample.
In a third aspect, an embodiment of the present application further provides an electronic device, including: the automatic driving test scene determining method comprises a processor, a storage medium and a bus, wherein the storage medium stores machine readable instructions executable by the processor, when the electronic device runs, the processor and the storage medium are communicated through the bus, and the processor executes the machine readable instructions to execute the steps of the automatic driving test scene determining method.
In a fourth aspect, the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the determination method for an autopilot test scenario are performed as described above.
The application discloses a method and a device for determining an automatic driving test scene, wherein the method comprises the following steps: determining K scene parameters and a cumulative distribution function corresponding to each scene parameter according to a historical scene library; n equal division is carried out on the probability interval of the cumulative distribution function corresponding to each scene parameter to obtain N probability sub-layers; randomly extracting M scene parameter probability sample points from a probability sublayer corresponding to each scene parameter according to a preset matrix R; according to each scene parameter probability sample point, K scene parameter values corresponding to K scene parameters in the ith target scene sample are obtained, the K scene parameter values are combined to form the ith target scene sample, through the process, the whole scene space can be better covered with fewer scenes, the load of simulation is reduced, and the efficiency of algorithm evaluation is improved.
Additional features and advantages of the present application will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by the practice of the embodiments of the present application. The objectives and other advantages of the application may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained from the drawings without inventive effort.
Fig. 1 shows a first flowchart of a method for determining an automatic driving test scenario according to an embodiment of the present application.
Fig. 2 is a schematic flowchart illustrating a first method for determining a probability sublayer corresponding to each scene parameter according to a preset matrix R according to an embodiment of the present application.
Fig. 3 shows a first schematic structural diagram of an automatic driving test scenario determination apparatus provided in an embodiment of the present application.
Fig. 4 shows a schematic structural diagram of an electronic device provided in an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. The components of the embodiments of the present application, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present application without making any creative effort, shall fall within the protection scope of the present application.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures. Meanwhile, in the description of the present disclosure, the terms "first", "second", and the like are used only for distinguishing the description, and are not to be construed as indicating or implying relative importance.
In order to reduce the burden of the automatic driving simulation and improve the efficiency of algorithm evaluation, the present application provides a method for determining an automatic driving test scenario, and referring to fig. 1, fig. 1 shows a first flow diagram of a method for determining an automatic driving test scenario provided by an embodiment of the present application, specifically, the method for determining includes:
s110, determining K scene parameters and a cumulative distribution function corresponding to each scene parameter according to a historical scene library.
Here, the historical scene library includes automatic driving test scenes used in training the automatic driving algorithm in a plurality of historical time periods, and determines a type K of a plurality of scene parameters related to the automatic driving test scenes in the plurality of historical scene libraries, where K represents a type of the scene parameters constituting the automatic driving scene, and for example, a value of K is 4 if the scene parameters in a certain automatic driving scene include weather, a vehicle speed of a preceding vehicle, a temperature, and a vehicle speed of a following vehicle.
And acquiring the parameter value of each scene parameter in the automatic driving test scene, and determining the cumulative distribution function corresponding to each scene parameter by counting the probability distribution of the parameter values.
And S120, performing N equal division on the probability interval of the cumulative distribution function corresponding to each scene parameter to obtain N probability sublayers.
In a specific embodiment, N equal divisions are performed on the probability interval [0,1] of the cumulative distribution function corresponding to each scene parameter, each scene parameter obtains [0,1/N ], [1/N,2/N ], [ N-1/N,1] and N probability sublayers in total, wherein the value range of N is [1, N ], the value of N can be determined according to the number of the finally required target automatic driving test scenes, and the larger the value of N is, the larger the number requirement of the target automatic driving test scenes is.
After each scene parameter is equally divided by N, each scene parameter has N probability sublayers, and 3 scene parameters have N × 3 probability sublayers, wherein any probability point in each scene parameter probability interval [0,1] represents the distribution probability of the scene parameter value corresponding to the probability point in the historical scene library.
S130, randomly extracting M scene parameter probability sample points from the probability sublayer corresponding to each scene parameter according to the preset matrix R.
Here, the preset matrix R is a matrix constructed according to a historical scene library and used for describing a correlation relationship between scene parameters, the preset matrix R is used for replacing a random matrix to perform random combination on randomly extracted scene parameter values, pseudo-correlation between the scene parameters is eliminated, a specific probability sublayer of each scene parameter in K scenes subjected to random combination can be determined according to the preset matrix R, and M scene parameter probability sample points are randomly extracted from the probability sublayer corresponding to each determined scene parameter, wherein the number of M is determined according to the requirement of the number of scene samples in a final constructed target scene library, and under the condition that the number of probability sublayers N is determined, the number of target scene samples can be expanded by increasing the number of M, for example, when N =2 and M =1, 2 target scene samples are obtained after the parameters are randomly combined, and when N =2 and M =2, the parameters are randomly combined to obtain 4 target scene samples.
Referring to fig. 2, a schematic flow chart of a method for determining a probability sublayer corresponding to each scene parameter according to a preset matrix R according to an embodiment of the present application is provided, and specifically, the method includes:
s1301, randomly extracting N historical scene samples from the historical scene library.
Where each historical scene sample contains K historical scene parameters, e.g. x is randomly extracted from the historical scene library1（x11，y12，z13）、x2（x21，y22，z 23）、x3（x31，y32，z 33）、x4（x41，y42，z 43) The 4 historical scene samples comprise three historical scene parameter values of a vehicle speed x, a weather y and a temperature z, wherein x is11、x21 、x31 、x41Is the corresponding value of the vehicle speed x, y12、y22 、y32 、y42Is the weather y corresponding value, z13、z23 、z33 、z43Is the temperature z corresponding value.
S1302, obtaining a historical scene parameter K from the K historical scene parameters, and determining corresponding values of the historical scene parameter K in the N historical scene samples.
k represents a specific scene parameter, for example, k is the vehicle speed, and then the historical scene parameter of the vehicle speed x is selected from three historical scene parameters of the vehicle speed x, the weather y and the temperature z, and the vehicle speed x is determined in the historical scene sample x1、x2、x3、x4The corresponding values in the 4 historical scene samples, if x1=（35，40，19）、x2=（28，75，32）、x3=（40，30，17）、x4= (55, 55, 29), then vehicle speed x corresponds to value x in 4 historical scene samples11、x21 、x31 、x4135, 28, 40, 55 respectively.
S1303, constructing an N multiplied by K preset matrix R by the N historical scene samples according to the preset sequence of the values of the historical scene parameters K.
Wherein each element R in the predetermined matrix RikThe sequence number is a sequence number which represents the kth historical scene parameter in the ith historical scene sample in the historical scene samples and is sequenced in the N historical scene samples according to the preset sequence, and the value of i is [1, N]K is [1, K ]]。
The preset order is a pre-designated order, and may be a sequence from small to large or a sequence from large to small, for example, the vehicle speed x is set to be the vehicle speed x11、x21 、x31 、x41The corresponding values 35, 28, 40, 55 are x in order from small to large1、x2、x3、x4The following 4 × 3 matrix a was constructed:
in formula (1), a represents a historical scene sample x to be randomly extracted1、x2、x3、x4And sequencing the vehicle speed values from small to large to obtain a matrix A, wherein each row in the matrix A is from top to bottom. Respectively corresponding to historical scene samples x2、x1、x3、x4From left to right, the matrix A represents the values of the vehicle speed x, the weather y and the temperature z in 4 historical scene samples, namely x2Corresponds to the first sample, x, in the matrix A1Corresponding to the second sample, x, in the matrix A3Corresponds to the third sample, x, in the matrix A4Corresponding to the fourth sample in matrix a.
Constructing an NxK preset matrix R according to a sequencing sequence number obtained by sequencing each historical scene parameter in the historical scene sample according to a preset sequence in the formula (1):
in formula (2), R represents a preset matrix R, wherein each element in the preset matrix R is represented by RikAnd i represents the number of rows corresponding to each element in the preset matrix R, and k represents the scene parameter corresponding to each element in the preset matrix R, for example, each column in the preset matrix R sequentially represents the vehicle speed x, the weather y and the temperature z from left to right.
And S1304, acquiring the sequence number of each scene parameter in the ith historical scene sample according to the preset matrix R.
For example, if i =2 and k =3, r is given by the formula (2)23And =2, which indicates that the third scene parameter temperature z in the 2 nd historical scene sample in the sorted historical scene samples is 2 in the sorted sequence number from small to large in all temperature values in the 4 historical scene samples.
And S1305, determining the probability sublayer corresponding to each scene parameter in the ith target scene sample according to the sequencing serial number of each scene parameter.
For example, according to r23=2, it may be determined that the probability sublayer for the scene parameter temperature in the 2 nd target scene sample is 2, i.e., [1/N,2/N]。
S1306, randomly extracting M scene parameter probability sample points from the probability sub-layer corresponding to each scene parameter.
For example, each scene parameter is divided into 4 probability sub-layers, M =1, r23=2, then the second probability sublayer determined is [1/4, 1/2%]From the probability sublayer [1/4,1/2]Randomly draw 1 scene probability sample point 3/8.
Returning to fig. 1, S140, obtaining K scene parameter values corresponding to K scene parameters in the ith target scene sample according to each scene parameter probability sample point, and combining the K scene parameter values to form the ith scene sample.
And performing corresponding inverse cumulative distribution function conversion on each scene parameter probability sample point to obtain K scene parameter values corresponding to the K scene parameters respectively, wherein the K scene parameter values are numerical values corresponding to the K scene parameters in the ith target scene sample.
For example, when a scene parameter vehicle speed corresponding value in the 1 st target scene sample is obtained, i =1 and k =1 are determined, and a corresponding element R is searched according to a preset matrix R11=1, then in the first probability sublayer of vehicle speed [0, 1/N%]Randomly extracting a probability sample point of scene parameter temperature, carrying out corresponding inverse cumulative distribution function conversion on the probability sample point to obtain a temperature value under the temperature probability, taking the temperature value as a value of a scene parameter vehicle speed x in a 1 st target scene sample, sequentially obtaining values of other scene parameter temperatures and weather in the 1 st target scene sample according to the method, and forming the 1 st target scene sample by the values.
In a specific embodiment of the present application, the value corresponding to each scene parameter in the ith target scene sample may be determined by using the following formula:
in the formula (3), wherein,
in the ith target scene sampleA value corresponding to the kth scene parameter, i represents a target scene sample, and k represents a scene parameter; n represents the number of probability sublayers divided by each scene parameter;
representing a probability sublayer corresponding to a kth scene parameter in an ith target scene sample determined according to a preset matrix;
indicating that the k scene parameter in the i target scene sample is [0,1]]Random variables distributed uniformly above;
the sample points which represent the random extraction of the kth scene parameter in the ith target scene sample in which probability sub-layer;
and performing an inversion operation on the probability sample point of the kth scene parameter to obtain a value corresponding to the kth scene parameter in the ith target scene sample.
For example, when i =1, k =1,
the value corresponding to the 1 st scene parameter in the 1 st target scene sample is represented, if the preset matrix is formula (2), the formula (2) can show that,
is 1, that is, the 1 st scene parameter in the 1 st target scene sample randomly takes a point in the first probability sublayer, if the probability sublayer N is 4,
the random value 1/4 is obtained by calculation according to the formula (3)
Represented in the 1 st probability sublayer [0,1/4 ]]The probability sample point 1/16 is randomly obtained, and the cumulative distribution function corresponding to the parameter is used for negation to obtain the value corresponding to the 1 st scene parameter in the 1 st target scene sample, and under the condition that the preset matrix R is different, the value range of the probability sample point is influenced.
Fig. 3 is a schematic structural diagram of a first apparatus for determining an automatic driving test scenario provided in an embodiment of the present application, specifically, the first apparatus includes: a first determination module 210, a first processing module 220, a first sampling module 230, and an acquisition module 240.
A first determining module 210, configured to determine, according to a historical scene library, K scene parameters and a cumulative distribution function corresponding to each scene parameter;
the first processing module 220 is configured to equally divide the probability interval of the cumulative distribution function corresponding to each scene parameter by N to obtain N probability sublayers;
a first sampling module 230, configured to randomly extract M scene parameter probability sample points from a probability sublayer corresponding to each scene parameter according to a preset matrix R;
an obtaining module 240, configured to obtain, according to each scene parameter probability sample point, K scene parameter values corresponding to K scene parameters in an ith target scene sample, and combine the K scene parameter values to form an ith target scene sample;
the determination means further comprises:
and the second sampling module is used for randomly extracting N historical scene samples from the historical scene library, wherein each historical scene comprises K historical scene parameters.
And the second construction module is used for constructing a preset matrix R according to the N historical scene samples and the corresponding K historical scene parameters.
Wherein the second building block is further configured to:
and acquiring a historical scene parameter K from the K historical scene parameters, and determining the values of the historical scene parameter K in the N historical scene samples respectively.
N historical scene samples are sequenced in a preset order of the value of the historical scene parameter k,constructing an NxK preset matrix R, wherein each element R in the preset matrix RikAnd the sequence number represents the sequence number of the kth historical scene parameter in the ith historical scene sample in the historical scene samples after the kth historical scene parameter is sequenced in the N historical scene samples according to the preset sequence.
According to a preset matrix R, randomly extracting M scene parameter probability sample points from a probability sublayer corresponding to each scene parameter, wherein the method comprises the following steps:
acquiring the sequence number of each scene parameter in the ith historical scene sample according to the preset matrix R;
determining a probability sublayer corresponding to each scene parameter in the ith target scene sample according to the sequencing serial number of each scene parameter;
and randomly extracting M scene parameter probability sample points from the probability sublayer corresponding to each scene parameter.
And the second processing module is used for performing corresponding inverse cumulative distribution function conversion on each scene parameter probability sample point to obtain K scene parameter values corresponding to the K scene parameters respectively, wherein the K scene parameter values are numerical values corresponding to the K scene parameters in the ith target scene sample.
Referring to fig. 4, fig. 4 shows a schematic structural diagram of an electronic device provided in an embodiment of the present application, and as shown in fig. 4, the electronic device 300 includes a processor 310, a memory 320, and a bus 330.
The memory 320 stores machine-readable instructions executable by the processor 310, when the electronic device 300 runs, the processor 310 communicates with the memory 320 through the bus 330, and when the machine-readable instructions are executed by the processor 310, the steps of the method in the method embodiment shown in fig. 1 and fig. 2 can be executed.
An embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the method steps in the method embodiment shown in fig. 1 may be executed.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
Finally, it should be noted that: the above examples are only specific embodiments of the present application, and are not intended to limit the technical solutions of the present application, and the scope of the present application is not limited thereto, although the present application is described in detail with reference to the foregoing examples, those skilled in the art should understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope disclosed in the present application; such modifications, changes or substitutions do not depart from the spirit and scope of the exemplary embodiments of the present application, and are intended to be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A method for determining an automatic driving test scenario, the method comprising:
determining K scene parameters and a cumulative distribution function corresponding to each scene parameter according to a historical scene library;
carrying out N equal division on the probability interval of the cumulative distribution function corresponding to each scene parameter to obtain N probability sub-layers;
randomly extracting M scene parameter probability sample points from the probability sublayer corresponding to each scene parameter according to a preset matrix R;
acquiring K scene parameter values corresponding to K scene parameters in the ith target scene sample according to each scene parameter probability sample point;
and forming the ith target scene sample by combining the K scene parameter values.
2. The determination method according to claim 1, characterized in that the determination method further comprises:
randomly extracting N historical scene samples from the historical scene library, wherein each historical scene sample comprises K historical scene parameters;
and constructing a preset matrix R according to the N historical scene samples and the corresponding K historical scene parameters.
3. The method according to claim 2, wherein constructing a preset matrix R according to the N historical scene samples and the corresponding K historical scene parameters comprises:
acquiring a historical scene parameter K from the K historical scene parameters, and determining the values of the historical scene parameter K in the N historical scene samples respectively;
and constructing an N × K preset matrix R by using the N history scene samples according to a preset sequence of the values of the history scene parameters K, wherein each element rik in the preset matrix R represents a sequence number of the kth history scene parameter in the ith history scene sample in the history scene samples after being sequenced according to the preset sequence in the N history scene samples.
4. The method according to any one of claims 1 to 3, wherein the randomly extracting M scene parameter probability sample points from the probability sublayer corresponding to each scene parameter according to the preset matrix R comprises:
acquiring the sequence number of each scene parameter in the ith historical scene sample according to the preset matrix R;
determining a probability sublayer corresponding to each scene parameter in the ith target scene sample according to the sequencing serial number of each scene parameter;
and randomly extracting M scene parameter probability sample points from the probability sub-layer corresponding to each scene parameter.
5. The method according to claim 1, wherein the obtaining K scene parameter values corresponding to K scene parameters in an i-th target scene sample according to the each scene parameter probability sample point includes:
and performing corresponding inverse cumulative distribution function conversion on each scene parameter probability sample point to obtain K scene parameter values corresponding to K scene parameters respectively, wherein the K scene parameter values are numerical values corresponding to K scene parameters in the ith target scene sample.
6. An apparatus for determining an autopilot test scenario, the apparatus comprising:
the first determining module is used for determining K scene parameters and a cumulative distribution function corresponding to each scene parameter according to the historical scene library;
the first processing module is used for carrying out N equal division on the probability interval of the cumulative distribution function corresponding to each scene parameter to obtain N probability sublayers;
the first sampling module is used for randomly extracting M scene parameter probability sample points from the probability sublayer corresponding to each scene parameter according to a preset matrix R;
and the acquisition module is used for acquiring K scene parameter values corresponding to K scene parameters in the ith target scene sample according to each scene parameter probability sample point, and the ith target scene sample is formed by combining the K scene parameter values.
7. The determination apparatus according to claim 6, wherein the determination apparatus further comprises:
the second sampling module is used for randomly extracting N historical scene samples from the historical scene library, and each historical scene comprises K historical scene parameters;
and the second construction module is used for constructing a preset matrix R according to the N historical scene samples and the corresponding K historical scene parameters.
8. The determination apparatus according to claim 6, wherein the determination apparatus further comprises:
and the second processing module is used for performing corresponding inverse cumulative distribution function conversion on each scene parameter probability sample point to obtain K scene parameter values corresponding to the K scene parameters respectively, wherein the K scene parameter values are numerical values corresponding to the K scene parameters in the ith target scene sample.
9. An electronic device comprising an updater, a memory and a bus, the memory storing machine readable instructions executable by the updater, the updater and the memory communicating via the bus when the electronic device is running, the machine readable instructions being executable by the updater to perform the steps of the method for determining an autopilot test scenario of any of claims 1-5.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a computer program which, when being executed by an updater, performs the steps of the method of determining an autopilot test scenario according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110883759.9A CN113326638B (en) | 2021-08-03 | 2021-08-03 | Method and device for determining automatic driving test scene |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110883759.9A CN113326638B (en) | 2021-08-03 | 2021-08-03 | Method and device for determining automatic driving test scene |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113326638A true CN113326638A (en) | 2021-08-31 |
| CN113326638B CN113326638B (en) | 2021-11-09 |
Family
ID=77426826
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110883759.9A Active CN113326638B (en) | 2021-08-03 | 2021-08-03 | Method and device for determining automatic driving test scene |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113326638B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115828638A (en) * | 2023-01-09 | 2023-03-21 | 西安深信科创信息技术有限公司 | Automatic driving test scene script generation method and device and electronic equipment |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150109414A1 (en) * | 2013-10-17 | 2015-04-23 | Amit Adam | Probabilistic time of flight imaging |
| CN109816149A (en) * | 2018-12-28 | 2019-05-28 | 中国电力科学研究院有限公司 | A kind of wind power plant is contributed scene generating method and device at random |
| CN110188797A (en) * | 2019-04-26 | 2019-08-30 | 同济大学 | A kind of intelligent automobile method for rapidly testing based on Bayes's optimization |
| CN111476629A (en) * | 2020-03-06 | 2020-07-31 | 北京三快在线科技有限公司 | Data prediction method and device, electronic equipment and storage medium |
| CN112036001A (en) * | 2020-07-01 | 2020-12-04 | 长安大学 | Automatic driving test scene construction method, device and equipment and readable storage medium |
| CN112466119A (en) * | 2020-11-26 | 2021-03-09 | 清华大学 | Method and system for predicting vehicle following speed of vehicle by using vehicle-road cooperative data |
| CN112583414A (en) * | 2020-12-11 | 2021-03-30 | 北京百度网讯科技有限公司 | Scene processing method, device, equipment, storage medium and product |
| US20210165481A1 (en) * | 2021-02-10 | 2021-06-03 | Intel Corporation | Method and system of interactive storytelling with probability-based personalized views |
-
2021
- 2021-08-03 CN CN202110883759.9A patent/CN113326638B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150109414A1 (en) * | 2013-10-17 | 2015-04-23 | Amit Adam | Probabilistic time of flight imaging |
| CN109816149A (en) * | 2018-12-28 | 2019-05-28 | 中国电力科学研究院有限公司 | A kind of wind power plant is contributed scene generating method and device at random |
| CN110188797A (en) * | 2019-04-26 | 2019-08-30 | 同济大学 | A kind of intelligent automobile method for rapidly testing based on Bayes's optimization |
| CN111476629A (en) * | 2020-03-06 | 2020-07-31 | 北京三快在线科技有限公司 | Data prediction method and device, electronic equipment and storage medium |
| CN112036001A (en) * | 2020-07-01 | 2020-12-04 | 长安大学 | Automatic driving test scene construction method, device and equipment and readable storage medium |
| CN112466119A (en) * | 2020-11-26 | 2021-03-09 | 清华大学 | Method and system for predicting vehicle following speed of vehicle by using vehicle-road cooperative data |
| CN112583414A (en) * | 2020-12-11 | 2021-03-30 | 北京百度网讯科技有限公司 | Scene processing method, device, equipment, storage medium and product |
| US20210165481A1 (en) * | 2021-02-10 | 2021-06-03 | Intel Corporation | Method and system of interactive storytelling with probability-based personalized views |
Non-Patent Citations (4)
| Title |
|---|
| STEVEYG: "自动驾驶仿真超全综述：从仿真场景、系统到评价", 《CSDN：HTTPS://BLOG.CSDN.NET/STEVEYG/ARTICLE/DETAILS/115009464》 * |
| 张明芳 等: "人机混驾环境下无信号交叉口自动驾驶汽车左转运动规划研究", 《中国公路学报 第34卷 第7期 2021年7月》 * |
| 朱向雷 等: "自动驾驶智能系统测试研究综述", 《软件学报》 * |
| 秀秀的奇妙旅行: "自动生成用于测试和评估自动驾驶汽车的各种挑战性场景", 《CSDN：HTTPS://BLOG.CSDN.NET/YUNXIU988622/ARTICLE/DETAILS/109577538》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115828638A (en) * | 2023-01-09 | 2023-03-21 | 西安深信科创信息技术有限公司 | Automatic driving test scene script generation method and device and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113326638B (en) | 2021-11-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111199270B (en) | Regional wave height forecasting method and terminal based on deep learning | |
| CN110674604A (en) | Transformer DGA data prediction method based on multi-dimensional time sequence frame convolution LSTM | |
| CN109978134B (en) | Fault prediction method based on fast integrated convolutional neural network | |
| CN111723732A (en) | Optical remote sensing image change detection method, storage medium and computing device | |
| CN111458142A (en) | Sliding bearing fault diagnosis method based on generation of countermeasure network and convolutional neural network | |
| CN113326638B (en) | Method and device for determining automatic driving test scene | |
| CN113326210B (en) | Method and device for determining automatic driving test scene | |
| CN112001110A (en) | Structural damage identification monitoring method based on vibration signal space real-time recursive graph convolutional neural network | |
| CN114556248A (en) | Method for determining a sensor configuration | |
| CN114139293A (en) | Rolling bearing fault diagnosis method and system | |
| CN115859816A (en) | Wind power minute-level prediction method and system based on CNN-LSTM algorithm | |
| CN114419464A (en) | Twin network change detection model based on deep learning | |
| CN114120063A (en) | Unsupervised fine-grained image classification model training method and classification method based on clustering | |
| CN112232171A (en) | Remote sensing image information extraction method and device based on random forest and storage medium | |
| CN112784499A (en) | Electric reactor magnetic field simulation method based on deep learning instead of finite element analysis | |
| CN117150402A (en) | Power data anomaly detection method and model based on generation type countermeasure network | |
| Linnenbrink et al. | kNNDM: k-fold Nearest Neighbour Distance Matching Cross-Validation for map accuracy estimation | |
| CN116595363A (en) | Prediction method, apparatus, device, storage medium, and computer program product | |
| Yang et al. | Predictor selection for CNN-based statistical downscaling of monthly precipitation | |
| CN116400168A (en) | Power grid fault diagnosis method and system based on depth feature clustering | |
| CN111241749B (en) | Permanent magnet synchronous motor chaos prediction method based on reserve pool calculation | |
| CN115034426A (en) | Rolling load prediction method based on phase space reconstruction and multi-model fusion Stacking integrated learning mode | |
| CN114239945A (en) | Short-term power load prediction method, device, equipment and storage medium | |
| CN113052060A (en) | Bearing residual life prediction method and device based on data enhancement and electronic equipment | |
| CN113408210A (en) | Deep learning based non-intrusive load splitting method, system, medium, and apparatus |
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 | ||
| CP03 | Change of name, title or address |
Address after: Room 401, 4th Floor, No. 66 Zizhuyuan Road, Haidian District, Beijing, 100048 Patentee after: Beijing Saimu Technology Co.,Ltd. Address before: 100082 1501 ZhongGuanCun international innovation building, Haidian District, Beijing Patentee before: Beijing saimu Technology Co.,Ltd. |
|
| CP03 | Change of name, title or address |