CN115454018A - Automatic driving scene test case generation method and system based on complexity - Google Patents
Automatic driving scene test case generation method and system based on complexity Download PDFInfo
- Publication number
- CN115454018A CN115454018A CN202210977515.1A CN202210977515A CN115454018A CN 115454018 A CN115454018 A CN 115454018A CN 202210977515 A CN202210977515 A CN 202210977515A CN 115454018 A CN115454018 A CN 115454018A
- Authority
- CN
- China
- Prior art keywords
- data
- complexity
- test case
- target data
- automatic driving
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0243—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults model based detection method, e.g. first-principles knowledge model
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/20—Pc systems
- G05B2219/24—Pc safety
- G05B2219/24065—Real time diagnostics
Abstract
The invention provides a method and a system for generating an automatic driving scene test case based on complexity, wherein the method comprises the following steps: acquiring target data; according to the target data, obtaining complexity data corresponding to the target data; the complexity data are sorted, and the sorted complexity data are classified; converting target data corresponding to the complexity data of the selected category by using Bayesian theorem based on the classified complexity data to obtain discrete data; and performing kernel density estimation on the discrete data to obtain a test case corresponding to the class complexity data. According to the method, the complexity data are classified, so that data conversion is conveniently carried out on the target data corresponding to the complexity data of different levels, and the nuclear density estimation is carried out on the converted data, so that the corresponding test cases can be conveniently generated according to the complexity of different levels, the test efficiency and the test accuracy are improved, and the safer automatic driving automobile is helped to be produced.
Description
Technical Field
The invention relates to the technical field of intelligent driving, in particular to a method and a system for generating an automatic driving scene test case based on complexity.
Background
An autopilot system refers to a system that assists a person in driving by a machine, and in special cases completely replaces human driving. The vehicle is controlled by an automatic driving system, which can avoid the defects of a human driver in some cases. A large number of autopilot tests are required for vehicles in an autopilot system to ensure the safety of autopilot. In addition to the stability of the autonomous driving system itself, the safety of autonomous driving is affected by complicated traffic conditions, unexpected behavior of the vehicle, and the like.
When the traditional test cases are used for compiling the test cases, a large number of test cases are generated only by adjusting parameters, complex scenes such as complex traffic conditions and unexpected behaviors of vehicles are not considered, and therefore the test accuracy of the generated test cases is poor.
Disclosure of Invention
The invention provides a method and a system for generating an automatic driving scene test case based on complexity, which are used for solving the defect of poor automatic driving scene test accuracy caused by the fact that the complexity is not considered in the generation of a test case in the prior art and improving the test efficiency and the test accuracy.
The invention provides a method for generating an automatic driving scene test case based on complexity, which comprises the following steps: acquiring target data; obtaining complexity data corresponding to the target data according to the target data; sorting the complexity data, and classifying the sorted complexity data; converting target data corresponding to the complexity data of the selected category by using Bayesian theorem based on the classified complexity data to obtain discrete data; and performing kernel density estimation on the discrete data to obtain a test case corresponding to the class complexity data.
According to the method for generating the test case of the automatic driving scene based on the complexity, provided by the invention, the discrete data is subjected to nuclear density estimation to obtain the test case corresponding to the class complexity data, and the method comprises the following steps: selecting a corresponding kernel function and bandwidth based on a preset kernel function selection rule and a bandwidth selection rule; performing kernel density estimation on the discrete data according to the kernel function and the bandwidth to obtain a density function corresponding to the category complexity data; and generating a test case corresponding to the category complexity data according to the density function.
According to the method for generating the test case of the automatic driving scene based on the complexity, which is provided by the invention, the corresponding kernel function and the bandwidth are selected based on the preset kernel function selection rule and the bandwidth selection rule, and the method comprises the following steps: obtaining a bandwidth according to the number of the discrete data, the dimensionality of the discrete data and the standard deviation of the discrete data; and selecting a Gaussian kernel based on a kernel function selection rule, and obtaining a kernel function according to the bandwidth.
According to the method for generating the test case of the automatic driving scene based on the complexity, the method for acquiring the target data comprises the following steps: acquiring original data; and marking the original data based on the variable parameters to obtain marked data.
According to the method for generating the test case of the automatic driving scene based on the complexity, after the labeled data is obtained, the method further comprises the following steps: performing interpolation processing on the labeled data to obtain constant-frequency time series data serving as target data; and/or, performing smoothing processing on the labeling data.
According to the method for generating the test case of the automatic driving scene based on the complexity, after the target data is obtained, the method further comprises the following steps: and averaging the target data based on a preset moving average window.
The invention also provides a system for generating the test case of the automatic driving scene based on the complexity, which comprises the following steps: the data acquisition module acquires target data; the complexity acquisition module is used for acquiring complexity data corresponding to the target data according to the target data; the classification module is used for sequencing the complexity data and classifying the sequenced complexity data; the data conversion module is used for converting target data corresponding to the complexity data of the selected category by using Bayesian theorem based on the classified complexity data to obtain discrete data; and the test case generation module is used for carrying out nuclear density estimation on the discrete data to obtain the test case corresponding to the class complexity data.
The invention also provides an electronic device, which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein when the processor executes the program, the steps of the automatic driving scene test case generation method based on the complexity are realized.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program that, when executed by a processor, implements the steps of the complexity-based autopilot scenario test case generation method as described in any of the above.
The invention also provides a computer program product comprising a computer program, wherein the computer program is used for realizing the steps of the automatic driving scene test case generation method based on the complexity when being executed by a processor.
According to the method and the system for generating the test case of the automatic driving scene based on the complexity, the complexity data are classified, so that data conversion is conveniently carried out on target data corresponding to the complexity data of different levels, and the nuclear density estimation is carried out on the converted data, so that the corresponding test case is conveniently generated according to the complexity of different levels, the test efficiency and the test accuracy are improved, and the safer automatic driving automobile is helped to be produced.
Drawings
In order to more clearly illustrate the technical solutions of the present invention or the prior art, the drawings needed for the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
FIG. 1 is a schematic flow chart of a method for generating test cases in an automatic driving scenario based on complexity according to the present invention;
FIG. 2 is a schematic structural diagram of a test case generation system for an automatic driving scenario based on complexity according to the present invention;
fig. 3 is a schematic structural diagram of an electronic device provided in the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 shows a schematic flow diagram of a method for generating a test case in an automatic driving scenario based on complexity according to the present invention, where the method includes:
s11, acquiring target data;
s12, obtaining complexity data corresponding to the target data according to the target data;
s13, sorting the complexity data and classifying the sorted complexity data;
s14, converting target data corresponding to the complexity data of the selected category by using Bayesian theorem based on the classified complexity data to obtain discrete data;
and S15, performing nuclear density estimation on the discrete data to obtain a test case corresponding to the class complexity data.
It should be noted that S1N in this specification does not represent the sequence of the method for generating the test case in the automatic driving scenario based on the complexity, and the method for generating the test case in the automatic driving scenario based on the complexity of the present invention is specifically described below.
Step S11, target data is acquired.
It should be noted that, in order to understand the influence of each variable parameter on the final complexity and generate a test case based on the influence relationship, the obtained target data should be simply changed only in the variable parameter, so as to subsequently generate a more accurate test case, and thus, obtaining the target data includes: acquiring original data; and marking the original data based on the variable parameters to obtain marked data.
In the present embodiment, the target data includes, but is not limited to, position and heading information of the host vehicle, speed and acceleration information of all moving targets, position and heading information relative to the host vehicle, and the like, for each sampling period.
For example, the raw data is represented as
Where n represents the number of raw data and k represents the dimensionality of the raw data. In an optional embodiment, when the original data is labeled, the method further includes: the original data is subjected to dimensionality reduction, and then the target data is expressed as
In an optional embodiment, after obtaining the annotation data, the method further includes: performing interpolation processing on the labeled data to obtain constant-frequency time series data serving as target data; and/or smoothing the marked data. It should be noted that, since there may be a high-frequency natural feature in part of the data, the interpolation may be performed based on the nearest neighbor interpolation method, so that the data is presented as a constant-frequency time series. For example, the target data is represented as
Closing k to interpolate it to obtain
Wherein, t k ＝t m ,
In an optional embodiment, since most of the information during driving comes from the scene vision, and the visual sense organ cannot continuously sense the information, after acquiring the target data, the method further includes: and averaging the target data based on a preset moving average window. It should be noted that, since human vision can sense images separated by 13 milliseconds, for convenience, the length of the moving average window needs to be less than 13 milliseconds, for example, the length can be set to 10 milliseconds, and then the corresponding target data is expressed as
And S12, obtaining complexity data corresponding to the target data according to the target data.
In this embodiment, obtaining complexity data corresponding to the target data according to the target data includes: obtaining a first complexity based on a heading angle of the target vehicle relative to the host vehicle; obtaining a second complexity based on the velocity, the maximum velocity, and the minimum velocity of the target vehicle relative to the host vehicle; obtaining a third complexity based on the velocity, the maximum distance, and the minimum distance of the target vehicle relative to the host vehicle; deriving a fourth complexity based on the deceleration of the host vehicle and the duration of the deceleration; complexity data is obtained based on the first complexity, the second complexity, the third complexity, and the fourth complexity.
The complexity data is represented as:
where θ represents a heading angle of the target vehicle relative to the host vehicle, v y Representing the velocity of the target vehicle relative to the host vehicle, d l Indicating the longitudinal distance of the target vehicle relative to the host vehicle and dcc the deceleration of the host vehicle. In a still further aspect of the present invention,
f(θ，v y ，d l )＝f 1 (θ)·f 2 (v y )·f 3 (d l )
wherein:
in addition, f is 1 (θ) is expressed as a first complexity. In addition, the closer the heading angle θ is to a right angle, the higher the complexity.
Wherein f is 2 (v y ) Representing a second complexity, v maximum Denotes the maximum speed, v minimum Representing the minimum speed. It is noted that the complexity is higher when the velocity of the host vehicle perpendicular to the lane direction position is approximately close to the maximum velocity, and follows v y As the maximum speed is approached, the complexity rises dramatically. In addition, the smaller the speed, the higher the complexity.
Wherein, f 3 (d l ) Representing a third complexity, d l_maximum Denotes the maximum distance, d l_minimum Representing the minimum distance. The complexity is higher as the distance approaches the minimum distance, and rises sharply as the distance approaches the minimum distance.
Wherein g (dcc) represents a fourth complexity,
Δ t is the average value of the deceleration of the host vehicle, and the duration of the deceleration. It should be noted that the larger the deceleration of the host vehicle, the longer the duration, and the stronger the braking demand of the vehicle at that time, the more complex the scene can be considered, reflecting the complexity of the environment of the host vehicleAnd (5) controlling. Here, the value of deceleration is more sensitive than the duration.
In an optional embodiment, after obtaining the complexity data corresponding to the target data, the method further includes: outliers in the complexity data are removed.
And S13, sorting the complexity data and classifying the sorted complexity data.
In this embodiment, a binomial system can be used to model the distribution, for example, if the complexity is divided into three levels, then 1:2:1 as a ratio, thereby classifying the data. Similarly, as the number of classes increases, higher order binomial coefficients are employed. It should be noted that the number of categories should match the number of data. For example, assuming n classes, each class should have data
Then, based on the central limit theorem, the limit is taken as normal distribution, i.e.
Where B (n, p) represents a quadratic distribution, where n represents the number of independent events and p represents the probability of an event occurring. N (mu, sigma) 2 ) Represents a normal distribution, where μ represents the mean and σ represents the standard deviation.
Similarly, for complexity data
After classification, various types of complexity data are obtained, namely
Wherein the content of the first and second substances,
in an alternative embodiment, before step S14, the method further includes: and carrying out unbiased assumption on the variable parameters to consider all parameter combinations to be equal in occurrence chance, so that the data can be transformed by using Bayes theorem in the following process.
And S14, converting the target data corresponding to the complexity data of the selected category by using Bayesian theorem based on the classified complexity data to obtain discrete data.
In this embodiment, bayesian theorem is expressed as:
wherein the content of the first and second substances,
it should be noted that when the category of the complexity data is determined, P (ranking) is constant for all target data of the complexity level. Similarly, P (parameters) is also a constant, independent of any other information, assuming that all target data for different parameter values are equally important before data analysis is performed. Combining the above assumptions, then P (parameters | ranking) is proportional to P (ranking | parameters) for the known class of complexity data.
Thus, by Bayesian theorem, directly on
Continue to give G, i.e.
Thereby obtaining G i ＝{(t 1 ，t 2 ，…，t m )|ranking＝i}
And S15, performing kernel density estimation on the discrete data to obtain a test case corresponding to the class complexity data.
In this embodiment, performing kernel density estimation on discrete data to obtain a test case corresponding to class complexity data includes: selecting a corresponding kernel function and bandwidth based on a preset kernel function selection rule and a bandwidth selection rule; performing kernel density estimation on the discrete data according to the kernel function and the bandwidth to obtain a density function corresponding to the class complexity data; and generating a test case corresponding to the category complexity data according to the density function. Specifically, the method comprises the following steps:
firstly, based on a preset kernel function selection rule and a bandwidth selection rule, selecting a corresponding kernel function and a bandwidth, specifically comprising: obtaining the bandwidth according to the number of the discrete data, the dimensionality of the discrete data and the standard deviation of the discrete data; and selecting a Gaussian kernel based on a kernel function selection rule, and acquiring a kernel function according to the bandwidth.
It should be noted that the bandwidth is expressed as:
wherein h is 1 Represents the bandwidth, n 1 Representing the number of discrete data, d 1 Represents the dimension of the discrete data, and σ represents the standard deviation of the discrete data.
Selecting a gaussian kernel based on the kernel function, which is expressed as:
where x represents a physical quantity of discrete data.
And secondly, performing kernel density estimation on the discrete data according to the kernel function and the bandwidth to obtain a density function.
Note that the kernel density estimate is expressed as:
wherein n is 2 Representing the size of the data set, d 2 Dimension representing a parameter space, K a kernel function, h 2 Indicating the bandwidth.
Then correspondingly, the kernel density estimation is performed on the discrete data, and a density function can be obtained, which is expressed as:
by
In a clear view of the above, it is known that,
therefore, it is
But also as a function of Gi generation.
And finally, generating a test case corresponding to the category complexity data according to the density function. It should be noted that, according to the density function, generating a test case corresponding to the category complexity data includes: determining the number of cases required by each category complexity level according to the category of the complexity data; and randomly extracting the test cases of the number of the cases from the density function corresponding to the complexity data of each category. At this time, the parameter values of the generated test case are the most likely example parameter values in the corresponding complexity levels.
In an optional embodiment, after obtaining the test case corresponding to the category complexity data, the method further includes: and testing by using the test case, recording the test result, marking the data, and returning to the step S11 by using the marked data to execute the steps again, so that the model is updated conveniently, and more accurate test case parameter values are generated subsequently.
In summary, the complexity data are classified, so that the data conversion is performed on the target data corresponding to the complexity data of different levels, and the kernel density estimation is performed on the converted data, so that the corresponding test cases are generated according to the complexity of different levels, the test efficiency and accuracy are improved, and the safer automatic driving automobile is produced.
The automatic driving scene test case generation system based on the complexity provided by the invention is described below, and the automatic driving scene test case generation system based on the complexity described below and the automatic driving scene test case generation method based on the complexity described above can be referred to correspondingly.
Fig. 2 shows a schematic structural diagram of a complexity-based automatic driving scenario test case generation system, where the apparatus includes:
a data acquisition module 21 that acquires target data;
the complexity obtaining module 22 is used for obtaining complexity data corresponding to the target data according to the target data;
the classification module 23 is used for sorting the complexity data and classifying the sorted complexity data;
the data conversion module 24 is used for converting the target data corresponding to the complexity data of the selected category by using Bayesian theorem based on the classified complexity data to obtain discrete data;
and the test case generation module 25 is used for carrying out kernel density estimation on the discrete data to obtain the test case corresponding to the class complexity data.
In this embodiment, in order to understand the influence of each variable parameter on the final complexity and generate a test case based on the influence relationship, the acquired target data should be simply changed in the variable parameter only, so as to generate a more accurate test case subsequently, and therefore, the data acquiring module 21 includes: a data acquisition unit that acquires original data; and the marking unit marks the original data based on the variable parameters to obtain marked data.
In an optional embodiment, the data obtaining module 21 further includes: and the dimension reduction unit is used for reducing the dimension of the original data when marking the original data.
In an optional embodiment, the data obtaining module 21 further includes: the interpolation processing unit is used for carrying out interpolation processing on the labeled data to obtain constant-frequency time series data serving as target data; and/or, the data obtaining module 21 further includes: and a smoothing unit for smoothing the annotation data. It should be noted that, since there may be high-frequency natural features in part of the data, interpolation may be performed based on the nearest neighbor interpolation method, so that the data is presented as a constant-frequency time series.
In an alternative embodiment, since most of the information during driving comes from the scene vision, and the visual sense organ cannot continuously sense the information, the system further comprises: and the data averaging module is used for averaging the target data based on a preset moving average window after the target data is acquired. It should be noted that, since human vision can perceive images at intervals of 13 ms, the length of the moving average window needs to be less than 13 ms for convenience, and may be set to 10 ms, for example.
The complexity obtaining module 22 is configured to obtain complexity data based on an intersection angle between the host vehicle and an adjacent vehicle, a vehicle speed of the host vehicle, a vertical lane direction position of the host vehicle, a driving direction of the host vehicle, an acceleration of the host vehicle, and a relative speed between the host vehicle and the adjacent vehicle.
Specifically, the complexity obtaining module 22 includes: a first complexity obtaining unit that obtains a first complexity based on a heading angle of the target vehicle relative to the host vehicle; a second complexity obtaining unit that obtains a second complexity based on a velocity, a maximum velocity, and a minimum velocity of the target vehicle relative to the host vehicle; a third complexity obtaining unit that obtains a third complexity based on a velocity, a maximum distance, and a minimum distance of the target vehicle relative to the host vehicle; a fourth complexity obtaining unit that obtains a fourth complexity based on a deceleration of the host vehicle and a duration of the deceleration; and the comprehensive complexity acquisition unit is used for acquiring complexity data based on the first complexity, the second complexity, the third complexity and the fourth complexity.
In an optional embodiment, the system further comprises: and the data removing module is used for removing the outlier data in the complexity data after the complexity data corresponding to the target data is obtained.
A classification module 23 comprising: the sorting unit sorts the complexity data; and the classification unit is used for classifying the sorted complexity data. In this embodiment, a binomial system may be used to model the distribution to classify the data. It should be noted that the number of categories should match the number of data.
In an optional embodiment, the system further comprises: and the unbiased assumption module is used for carrying out unbiased assumption on the variable parameters so as to treat all parameter combinations as equal occurrence opportunities, thereby being convenient for transforming the data by utilizing Bayes theorem subsequently.
The test case generation module 25 includes: the selection unit selects a corresponding kernel function and bandwidth based on a preset kernel function selection rule and a bandwidth selection rule; the kernel density estimation unit is used for carrying out kernel density estimation on the discrete data according to the kernel function and the bandwidth to obtain a density function corresponding to the class complexity data; and the test case generating unit generates a test case corresponding to the category complexity data according to the density function.
Specifically, the selection unit includes: the bandwidth selection subunit obtains the bandwidth according to the number of the discrete data, the dimensionality of the discrete data and the standard deviation of the discrete data; and the kernel function selection subunit selects the Gaussian kernel based on the kernel function selection rule and obtains the kernel function according to the bandwidth.
A test case generation unit comprising: the quantity determining subunit determines the number of cases required by each category complexity level according to the category of the complexity data; and the test case generation subunit randomly extracts the test cases of the number of the cases from the density function corresponding to the complexity data of each category.
In an optional embodiment, the system further comprises: and the updating module is used for testing by using the test case after obtaining the test case corresponding to the category complexity data, recording the test result, marking the data, and returning to the step S11 to execute the steps again by using the marked data, so that the model is updated conveniently, and more accurate test case parameter values are generated subsequently.
In summary, in the embodiment of the present invention, the classification module classifies the complexity data, so that the data conversion module performs data conversion on target data corresponding to different levels of complexity data, and the test case generation module performs kernel density estimation on the converted data, so that corresponding test cases are generated according to different levels of complexity, thereby improving test efficiency and accuracy, and helping to produce safer auto-driven vehicles.
Fig. 3 illustrates a physical structure diagram of an electronic device, which may include, as shown in fig. 3: a processor (processor) 31, a communication Interface (communication Interface) 32, a memory (memory) 33 and a communication bus 34, wherein the processor 31, the communication Interface 32 and the memory 33 are communicated with each other via the communication bus 34. The processor 31 may invoke logic instructions in the memory 33 to perform a complexity-based autopilot scenario test case generation method comprising: acquiring target data; according to the target data, obtaining complexity data corresponding to the target data; sorting the complexity data, and classifying the sorted complexity data; converting target data corresponding to the complexity data of the selected category by using Bayesian theorem based on the classified complexity data to obtain discrete data; and carrying out kernel density estimation on the discrete data to obtain a test case corresponding to the class complexity data.
In addition, the logic instructions in the memory 33 may be implemented in the form of software functional units and stored in a computer readable storage medium when the logic instructions are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In another aspect, the present invention further provides a computer program product, the computer program product including a computer program, the computer program being stored on a non-transitory computer-readable storage medium, wherein when the computer program is executed by a processor, the computer is capable of executing the method provided by the above methods to execute the method for generating the test case of the automatic driving scenario based on complexity, the method including: acquiring target data; according to the target data, obtaining complexity data corresponding to the target data; sorting the complexity data, and classifying the sorted complexity data; converting target data corresponding to the complexity data of the selected category by using Bayesian theorem based on the classified complexity data to obtain discrete data; and performing kernel density estimation on the discrete data to obtain a test case corresponding to the class complexity data.
In yet another aspect, the present invention also provides a non-transitory computer-readable storage medium having stored thereon a computer program, which when executed by a processor implements a method for performing complexity-based automated driving scenario test case generation provided by the above methods, the method comprising: acquiring target data; according to the target data, obtaining complexity data corresponding to the target data; the complexity data are sorted, and the sorted complexity data are classified; converting target data corresponding to the complexity data of the selected category by using Bayesian theorem based on the classified complexity data to obtain discrete data; and carrying out kernel density estimation on the discrete data to obtain a test case corresponding to the class complexity data.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one position, or may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment may be implemented by software plus a necessary general hardware platform, and may also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for generating an automatic driving scene test case based on complexity is characterized by comprising the following steps:
acquiring target data;
obtaining complexity data corresponding to the target data according to the target data;
sorting the complexity data, and classifying the sorted complexity data;
converting target data corresponding to the complexity data of the selected category by using Bayesian theorem based on the classified complexity data to obtain discrete data;
and performing kernel density estimation on the discrete data to obtain a test case corresponding to the class complexity data.
2. The method for generating the test case based on the automatic driving scene with the complexity according to claim 1, wherein the step of performing kernel density estimation on the discrete data to obtain the test case corresponding to the category complexity data comprises the steps of:
selecting a corresponding kernel function and bandwidth based on a preset kernel function selection rule and a bandwidth selection rule;
performing kernel density estimation on the discrete data according to the kernel function and the bandwidth to obtain a density function corresponding to the category complexity data;
and generating a test case corresponding to the category complexity data according to the density function.
3. The method for generating the test case based on the automatic driving scenario of the complexity according to claim 2, wherein the selecting the corresponding kernel function and the bandwidth based on the preset kernel function selection rule and the bandwidth selection rule comprises:
obtaining a bandwidth according to the number of the discrete data, the dimensionality of the discrete data and the standard deviation of the discrete data;
and selecting a Gaussian kernel based on a kernel function selection rule, and obtaining a kernel function according to the bandwidth.
4. The method for generating the test case based on the automatic driving scene of the complexity according to claim 1, wherein the obtaining the target data comprises:
acquiring original data;
and labeling the original data based on the variable parameters to obtain labeled data.
5. The method for generating the test case based on the automatic driving scenario of claim 4, wherein after obtaining the annotation data, the method further comprises:
performing interpolation processing on the labeled data to obtain constant-frequency time series data serving as target data; and/or the presence of a gas in the gas,
and carrying out smoothing processing on the labeled data.
6. The method for generating the test case based on the automatic driving scene of the complexity according to claim 1, after obtaining the target data, further comprising:
and averaging the target data based on a preset moving average window.
7. A complexity-based automatic driving scenario test case generation system is characterized by comprising:
the data acquisition module acquires target data;
the complexity obtaining module is used for obtaining complexity data corresponding to the target data according to the target data;
the classification module is used for sequencing the complexity data and classifying the sequenced complexity data;
the data conversion module is used for converting target data corresponding to the complexity data of the selected category by using Bayesian theorem based on the classified complexity data to obtain discrete data;
and the test case generation module is used for carrying out nuclear density estimation on the discrete data to obtain the test case corresponding to the class complexity data.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps of the complexity-based autopilot scenario test case generation method of any of claims 1 to 6.
9. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program, when executed by a processor, implements the steps of the complexity-based automated driving scenario test case generation method of any of claims 1 to 6.
10. A computer program product comprising a computer program, wherein the computer program, when executed by a processor, implements the steps of the complexity-based automated driving scenario test case generation method of any of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210977515.1A CN115454018A (en) | 2022-08-15 | 2022-08-15 | Automatic driving scene test case generation method and system based on complexity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210977515.1A CN115454018A (en) | 2022-08-15 | 2022-08-15 | Automatic driving scene test case generation method and system based on complexity |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115454018A true CN115454018A (en) | 2022-12-09 |
Family
ID=84299042
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210977515.1A Pending CN115454018A (en) | 2022-08-15 | 2022-08-15 | Automatic driving scene test case generation method and system based on complexity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115454018A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116934556A (en) * | 2023-09-08 | 2023-10-24 | 四川三思德科技有限公司 | Target personnel accurate control method based on big data fusion |
-
2022
- 2022-08-15 CN CN202210977515.1A patent/CN115454018A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116934556A (en) * | 2023-09-08 | 2023-10-24 | 四川三思德科技有限公司 | Target personnel accurate control method based on big data fusion |
| CN116934556B (en) * | 2023-09-08 | 2023-12-26 | 四川三思德科技有限公司 | Target personnel accurate control method based on big data fusion |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107016193B (en) | Expected following distance calculation method in driver following behavior analysis | |
| EP3217332A1 (en) | Risk prediction method | |
| US20220048536A1 (en) | Method and device for testing a driver assistance system | |
| CN109919252B (en) | Method for generating classifier by using few labeled images | |
| US20210125061A1 (en) | Device and method for the generation of synthetic data in generative networks | |
| CN111994084B (en) | Method and system for identifying driving style of driver and storage medium | |
| US20220019713A1 (en) | Estimation of probability of collision with increasing severity level for autonomous vehicles | |
| EP3907654A1 (en) | Method for explainable active learning, to be used for object detector, by using bayesian dual autoencoder and active learning device using the same | |
| US11392804B2 (en) | Device and method for generating label objects for the surroundings of a vehicle | |
| CN109871792B (en) | Pedestrian detection method and device | |
| CN114117740A (en) | Simulation test scene generation method and device based on automatic driving | |
| CN112141098B (en) | Obstacle avoidance decision method and device for intelligent driving automobile | |
| CN115454018A (en) | Automatic driving scene test case generation method and system based on complexity | |
| JP7047498B2 (en) | Learning programs, learning methods and learning devices | |
| CN113642114A (en) | Modeling method for humanoid random car following driving behavior capable of making mistakes | |
| CN113176562A (en) | Multi-target tracking method and device, electronic equipment and readable storage medium | |
| US11908178B2 (en) | Verification of computer vision models | |
| CN115168614A (en) | Autonomous vehicle view shielding area collision risk assessment method and system | |
| CN114544191A (en) | Automatic driving test scene quantitative evaluation method and related equipment | |
| CN112733784A (en) | Neural network training method for determining whether charging amount of desulfurized gypsum is appropriate | |
| CN111950644A (en) | Model training sample selection method and device and computer equipment | |
| CN112698578A (en) | Automatic driving model training method and related equipment | |
| EP4109191A1 (en) | Methods and systems for generating ground truth data | |
| JP7247993B2 (en) | RUNNING TEST PATTERN CREATION APPARATUS AND METHOD | |
| Lu et al. | A BEV Scene Classification Method based on Historical Location Points and Unsupervised Learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |