CN116088476A - Method and device for realizing self-adaptive acceleration test evaluation - Google Patents
Method and device for realizing self-adaptive acceleration test evaluation Download PDFInfo
- Publication number
- CN116088476A CN116088476A CN202310070263.9A CN202310070263A CN116088476A CN 116088476 A CN116088476 A CN 116088476A CN 202310070263 A CN202310070263 A CN 202310070263A CN 116088476 A CN116088476 A CN 116088476A
- Authority
- CN
- China
- Prior art keywords
- test
- self
- adaptive acceleration
- adaptive
- ith
- 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
- 238000012360 testing method Methods 0.000 title claims abstract description 343
- 230000001133 acceleration Effects 0.000 title claims abstract description 145
- 238000000034 method Methods 0.000 title claims abstract description 61
- 238000011156 evaluation Methods 0.000 title claims abstract description 58
- 238000005070 sampling Methods 0.000 claims abstract description 97
- 230000003044 adaptive effect Effects 0.000 claims abstract description 60
- 230000006870 function Effects 0.000 claims description 69
- 238000004590 computer program Methods 0.000 claims description 14
- 238000003860 storage Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 10
- 238000009826 distribution Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 7
- 238000005457 optimization Methods 0.000 claims description 4
- 238000006467 substitution reaction Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000010998 test method Methods 0.000 description 4
- 238000012614 Monte-Carlo sampling Methods 0.000 description 3
- 238000005315 distribution function Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000005520 cutting process Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011076 safety test Methods 0.000 description 1
- 230000007723 transport mechanism Effects 0.000 description 1
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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
Disclosed herein is a method and apparatus for implementing adaptive acceleration test evaluation, comprising: for the self-adaptive acceleration test from the second time, determining an importance sampling function of the i-th self-adaptive acceleration test according to the test result of the self-adaptive acceleration test before i-1 times; determining a test scene X according to the determined importance sampling function i The method comprises the steps of carrying out a first treatment on the surface of the In test scene X i The self-adaptive acceleration test is carried out to obtain the test result of the ith time; determining control parameters of a self-adaptive acceleration evaluation method based on control variables according to the obtained test results; calculating a test index of the ith self-adaptive acceleration test according to the determined control parameters; and when the convergence of the test result is judged according to the test index, the test index is used as a test evaluation result. According to the embodiment of the invention, the generation of the test scene is accelerated by combining the self-adaptive acceleration test of multiple importance sampling and the self-adaptive acceleration evaluation based on the control variable, and the efficiency of the test evaluation of the automatic driving automobile is improved.
Description
Technical Field
The present disclosure relates to, but is not limited to, unmanned techniques, and more particularly to a method and apparatus for performing adaptive acceleration test evaluation.
Background
Currently, the automatic driving automobile faces serious safety problems, and the automatic driving automobile developed by related companies suffers serious traffic accidents, and the safety problems fundamentally prevent the automatic driving automobile from being applied in a large scale and being commercially landed, so that the automatic driving automobile needs to be subjected to safety test evaluation.
The basic flow of the test and evaluation of the automatic driving automobile is as follows: generating a series of test scenes (namely, a general dynamic description of the comprehensive interaction process of elements such as other vehicles, roads, traffic facilities, meteorological conditions and the like in the driving environment of the automatic driving automobile in a certain time and space range, wherein the test scenes are organic combinations of the driving scene and the driving environment of the automatic driving automobile, and comprise various entity elements and also cover actions executed by the entities and connection relations among the entities, such as expressway driving scenes, following scenes, cutting scenes, turning scenes and the like); testing the automatic driving automobile and collecting test results; and evaluating the safety of the automatic driving automobile according to the collected test results to obtain estimated values of the test indexes such as accident rate and the like.
At present, because safety key scenes (such as collision accident scenes and the like) in a natural driving environment are very rare, the adopted real road test method is very inefficient. In order to solve the inefficiency of real road test, the accelerating test scene generating method in the related technology improves the efficiency of test evaluation of the automatic driving automobile by improving the sampling probability of the key scenes, and how to generate the safety key scenes is a leading edge scientific problem to be solved urgently.
Disclosure of Invention
The following is a summary of the subject matter described in detail herein. This summary is not intended to limit the scope of the claims.
The embodiment of the invention provides a method and a device for realizing self-adaptive acceleration test evaluation, which can accelerate test scene generation and improve the efficiency of automatic driving automobile test evaluation.
The embodiment of the invention provides a method for realizing self-adaptive acceleration test evaluation, which comprises the following steps:
according to the test results of the i-1 th and the previous self-adaptive acceleration tests, determining an importance sampling function of the i-th self-adaptive acceleration test;
sampling and determining a test scene X of the ith self-adaptive test according to the determined importance sampling function of the ith self-adaptive acceleration test i ；
Test scenario X in the determined ith adaptive test i The method comprises the steps of performing self-adaptive acceleration test on an automatic driving automobile to obtain a test result of an ith self-adaptive acceleration test;
determining control parameters of a control variable-based adaptive acceleration evaluation method according to the obtained test result of the ith adaptive acceleration test;
calculating a test index of the ith self-adaptive acceleration test according to the determined control parameters;
when the convergence of the test result is judged according to the calculated test index of the ith self-adaptive acceleration test, the determined test index is determined to be a test evaluation result;
wherein i is an integer greater than 1 but less than or equal to n, n being the number of times of the self-adaptive acceleration test when the test result converges; the test scene X i ～q (i) ，q (i) For the importance sampling function of the ith self-adaptive acceleration test, an initial importance sampling function q (1) =q; the control parameter is a parameter contained in the control variable of the adaptive acceleration evaluation.
In an exemplary embodiment, before the determining the importance sampling function of the ith adaptive acceleration test, the method further includes:
the natural probability distribution p of the test scene is determined.
In an exemplary embodiment, before the determining the importance sampling function of the ith adaptive acceleration test, the method further includes:
constructing an initial importance sampling function.
In one illustrative example, the test index is determined by the following formula calculation:
wherein beta is i To control parameters, q α Representation of q (i) Carrying out summation operation, wherein the value of i is 1-n; q α (X i ) For solving for X i Substituted q α Is a value of (2);
representing that beta is summed, and i is 1-n; q (i) (X i ) Representing X i Substitution of q (i) The function value after that; />
For test scene X i A corresponding test result; p (X) i ) Representing test scenario X i Is a natural probability distribution of (c).
In one illustrative example, the control parameters are based on minimization
The variance of (c) is determined by the following formula calculation:
in one illustrative example, the determining the importance sampling function of the ith adaptive acceleration test includes:
and according to the test results of the i-1 th and previous self-adaptive acceleration tests, determining an importance sampling function of the i-th self-adaptive acceleration test by adopting a Bayesian optimization algorithm BOA.
In one illustrative example, the determination of whether the test results converge is made by:
calculating the variance of the test index;
and when the variance of the test index is smaller than or equal to a preset variance threshold, determining that the test result is converged.
On the other hand, the embodiment of the invention also provides a computer storage medium, wherein a computer program is stored in the computer storage medium, and the computer program realizes the method for realizing the self-adaptive acceleration test evaluation when being executed by a processor.
In still another aspect, an embodiment of the present invention further provides a terminal, including: a memory and a processor, the memory storing a computer program; wherein,,
the processor is configured to execute the computer program in the memory;
the computer program, when executed by the processor, implements a method of implementing adaptive accelerated test evaluation as described above.
In still another aspect, an embodiment of the present invention further provides an apparatus for implementing adaptive acceleration test evaluation, including: the method comprises the steps of determining a sampling function unit, a sampling unit, a testing unit, a parameter determining unit, a calculating unit and a processing unit; wherein,,
the determining sampling function unit is set to: according to the test results of the i-1 th and the previous self-adaptive acceleration tests, determining an importance sampling function of the i-th self-adaptive acceleration test;
the sampling unit is set as: sampling and determining a test scene X of the ith self-adaptive test according to the determined importance sampling function of the ith self-adaptive acceleration test i ；
The test unit is arranged as follows: at the time of determinationTest scene X of ith self-adaptive test i The method comprises the steps of performing self-adaptive acceleration test on an automatic driving automobile to obtain a test result of an ith self-adaptive acceleration test;
the determination parameter unit is set to: determining control parameters of a control variable-based adaptive acceleration evaluation method according to the obtained test result of the ith adaptive acceleration test;
the calculation unit is configured to: calculating a test index of the ith self-adaptive acceleration test according to the determined control parameters;
the processing unit is configured to: when the convergence of the test result is judged according to the calculated test index of the ith self-adaptive acceleration test, the determined test index is determined to be a test evaluation result;
wherein i is an integer greater than 1 but less than or equal to n, n being the number of times of the self-adaptive acceleration test when the test result converges; the test scene X i ～q (i) ，q (i) For the importance sampling function of the ith self-adaptive acceleration test, an initial importance sampling function q (1) =q; the control parameter is a parameter contained in the control variable of the adaptive acceleration evaluation.
The technical scheme of the application comprises the following steps: according to the test results of the i-1 th and the previous self-adaptive acceleration tests, determining an importance sampling function of the i-th self-adaptive acceleration test; sampling and determining a test scene X of the ith self-adaptive test according to the determined importance sampling function of the ith self-adaptive acceleration test i The method comprises the steps of carrying out a first treatment on the surface of the In test scene X i The method comprises the steps of performing self-adaptive acceleration test on an automatic driving automobile to obtain a test result of an ith self-adaptive acceleration test; determining control parameters of a control variable-based adaptive acceleration evaluation method according to the obtained test result of the ith adaptive acceleration test; calculating a test index of the ith self-adaptive acceleration test according to the determined control parameters; when the convergence of the test result is judged according to the calculated test index of the ith self-adaptive acceleration test, the determined test index is determined to be a test evaluation result; wherein i is an integer greater than 1 but less than or equal to n, n being the adaptive acceleration test when the test result convergesThe number of times; the test scene X i ～q (i) ，q (i) For the importance sampling function of the ith self-adaptive acceleration test, an initial importance sampling function q (1) =q; the control parameter is a parameter contained in the control variable of the adaptive acceleration evaluation. According to the embodiment of the invention, the generation of the test scene is accelerated by combining the self-adaptive acceleration test of multiple importance sampling and the self-adaptive acceleration evaluation based on the control variable, and the efficiency of the test evaluation of the automatic driving automobile is improved.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
Drawings
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate and do not limit the invention.
FIG. 1 is a flow chart of a method for implementing adaptive acceleration test evaluation according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of an exemplary cut scene for the application of the present invention;
FIG. 3 is a schematic diagram of the natural probability distribution of an exemplary cut scene for which the present invention is applied;
FIG. 4 is a schematic diagram of an exemplary importance sampling function applied by the present invention;
FIG. 5 is a comparison chart of an exemplary test index for the application of the present invention;
FIG. 6 is a graph comparing another test index of an application example of the present invention;
FIG. 7 is a graph comparing yet another test index for an application example of the present invention;
FIG. 8 is a graph comparing yet another test index for an application example of the present invention;
FIG. 9 is a graph of estimation error for an exemplary test index used in the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in detail hereinafter with reference to the accompanying drawings. It should be noted that, in the case of no conflict, the embodiments and features in the embodiments may be arbitrarily combined with each other.
The steps illustrated in the flowchart of the figures may be performed in a computer system, such as a set of computer-executable instructions. Also, while a logical order is depicted in the flowchart, in some cases, the steps depicted or described may be performed in a different order than presented herein.
The embodiment of the invention records the test index (such as accident rate) of the automatic driving automobile as
Wherein (1)>
As a probability measure +.>
Indicating expectations, A indicating events of interest (e.g. collision accidents), p indicating the natural probability distribution function of the test scene X, +.>
An indication function for event a;
the embodiment of the invention is based on the Monte Carlo sampling principle, and the expression of the test index determined by referring to the related technology is as follows:
wherein,,
for the estimated value of the test index based on Monte Carlo sampling principle,/for the test index>
Representing a given test scenario X i Conditional probability of event a occurrence (also referred to as test result), X i P represents the test scene X i The probability distribution function p is sampled and determined.
The probability of occurrence of accident events of the automatic driving automobile is extremely low, so that a testing method based on Monte Carlo sampling needs a large amount of measurement to obtain an estimated value of the accident rate, and the time cost and the economic cost consumed by the large amount of measurement are not acceptable; the importance sampling method can alleviate the problem, and the basic idea is to improve the sampling probability of key scenes which are easy to occur accidents, so that the testing efficiency is improved; the expression of the test index determined based on the importance sampling method is as follows:
wherein,,
for the estimated value of the test index based on the importance sampling principle, p (X i ) Representing test scenario X i Q is an importance sampling function constructed with reference to the related art, q (X i ) Representing test scenario X i A corresponding importance sampling function.
The applicant analysis of the present application found that, due to the high complexity of the automatic driving vehicle system and the black box nature of the neural network algorithm employed, the importance sampling function q was constructed with a theoretically optimal importance sampling function q * There is a great gap, so that the method based on importance sampling has efficiency improvement bottleneck, and the self-adaptive acceleration test method aims at solving the problem; after the test results are obtained, the importance sampling function used for the test can be adjusted and optimized according to the information provided by the test results, so that the evaluation efficiency of the automatic driving automobile is improved.
FIG. 1 is a flowchart of a method for implementing adaptive acceleration test evaluation according to an embodiment of the present invention, as shown in FIG. 1, including:
102, sampling and determining a test scene X of the ith self-adaptive test according to the determined importance sampling function of the ith self-adaptive acceleration test i ；
104, determining control parameters of a self-adaptive acceleration evaluation method based on control variables according to the obtained test result of the ith self-adaptive acceleration test;
wherein i is an integer greater than 1 but less than or equal to n, n being the number of times of self-adaptive acceleration test when the test result converges; test scene X i ～q (i) ，q (i) For the importance sampling function of the ith self-adaptive acceleration test, an initial importance sampling function q (1) =q; the control parameter is a parameter contained in a control variable of the adaptive acceleration evaluation.
The embodiment of the invention executes the processing of the steps 101-106 on the ith self-adaptive acceleration test from the second self-adaptive acceleration test until the convergence of the test result is judged according to the calculated test index, and the processing is ended. According to the embodiment of the invention, the generation of the test scene is accelerated by combining the self-adaptive acceleration test of multiple importance sampling and the self-adaptive acceleration evaluation based on the control variable, and the efficiency of the test evaluation of the automatic driving automobile is improved.
In an exemplary embodiment, the adaptive acceleration test for multiple importance sampling and the adaptive acceleration evaluation based on the control variable in the embodiment of the present invention are both algorithms existing in the related art, and the two algorithms may be combined and may also be implemented by referring to the related art, which is not described herein.
In one illustrative example, embodiment q of the present invention (i) And determining according to the test results of the i-1 th adaptive acceleration tests before the i-th adaptive acceleration test, so that the importance sampling function of the i-th adaptive acceleration test is sequentially determined according to the order of the values from small to large. The importance sampling function of the ith self-adaptive acceleration test is successively determined according to the sequence from small to large, the test before the nth self-adaptive acceleration test is the test with the non-converged test result, and the test result is converged when the nth self-adaptive test is executed.
In an exemplary embodiment, before determining the importance sampling function of the ith adaptive acceleration test, the method according to the embodiment of the present invention further includes:
the natural probability distribution function p of the test scene is determined and can be obtained from a natural driving data set.
In an exemplary embodiment, before determining the importance sampling function of the ith adaptive acceleration test, the method according to the embodiment of the present invention further includes:
an initial importance sampling function is constructed.
In an illustrative example, embodiments of the present invention may refer to S.Feng, Y.Feng, C.Yu, Y.Zhang and h.x.liu, "online autopilot test scenario library generation, part 1: method (Testing Scenario Library Generation for Connected and Automated Vehicles, part I: method) "IEEE Transactions on Intelligent Transportation Systems, vol.22, no.3, pp.1573-1582,March 2021,doi:10.1109/TITS.2020.2972211 and S.Feng, Y.Feng, H.Sun, S.Bao, Y.Zhang and H.X.Liu," Online autopilot automobile test scene library generation, part 2: case study (Testing Scenario Library Generation for Connected and Automated Vehicles, part II: case Studies) "IEEE Transactions on Intelligent Transportation Systems, vol.22, no.9, pp.5635-5647, sept.2021, doi:10.1109/TITS.2020.2988309, an importance sampling function was constructed.
In one illustrative example, the test index in an embodiment of the present invention is determined by the following formula calculation:
wherein beta is i To control parameters, q α Representation of q (i) Carrying out summation operation, wherein the value of i is 1-n; q α (X i ) For solving for X i Substituted q α Is a value of (2);
representing that beta is summed, and i is 1-n; q (i) (X i ) Representing X i Substitution of q (i) The function value after that; />
For test scene X i A corresponding test result; p (X) i ) Representing test scenario X i Is a natural probability distribution of (c).
The solution of the control parameters in the embodiment of the invention is a multiple linear regression problem, and can be solved by referring to the analysis of the related principles.
In one illustrative example, embodiments of the present invention derive an expression of a test index from an adaptive acceleration test comprising multiple importance samples of n tests in combination with an adaptive acceleration evaluation based on control variables.
In one illustrative example, the control parameters in embodiments of the present invention are based on minimization
Is determined by variance calculation.
In one illustrative example, the control parameters in embodiments of the present invention are based on minimization
The variance of (c) is determined by the following formula calculation:
in one illustrative example, an embodiment of the present invention determines an importance sampling function for an ith adaptive acceleration test, comprising:
and according to the test results of the i-1 th and previous self-adaptive acceleration tests, determining an importance sampling function of the i-th self-adaptive acceleration test by adopting a Bayesian optimization algorithm BOA.
In an illustrative example, the bayesian optimization algorithm of the embodiment of the present invention can be found in S.Feng, Y.Feng, H.Sun, Y.Zhang and h.x.liu, "online autopilot test scenario library generation: an adaptive framework (Testing Scenario Library Generation for Connected and Automated Vehicles: an Adaptive Framework) "IEEE Transactions on Intelligent TransportationSystems, vol.23, no.2, pp.1213-1222, feb.2022, doi:10.1109/TITS.2020.3023668.
In one illustrative example, an embodiment of the present invention determines whether the test results converge by:
calculating the variance of the test index;
and when the variance of the test index is smaller than or equal to a preset variance threshold, determining that the test result converges.
In one illustrative example, the adaptive test model of the present embodiment is implemented as follows:
acquiring natural probability distribution p of a test scene; here, the natural probability distribution may be determined by a related art with reference to a method of the adaptive acceleration test.
Constructing an initial importance sampling function q (1) The method comprises the steps of carrying out a first treatment on the surface of the In an exemplary embodiment, the embodiment of the present invention may construct an initial importance sampling function q with reference to a proxy model in the related art (1) The method comprises the steps of carrying out a first treatment on the surface of the For example, by means of an intelligent driving model (IDM, intelligent Driver Model), an initial importance sample is constructedFunction q (1) ；
For i=1:n, the cycle proceeds:
sampling test scene X i ～q (i) ；
According to the sampled test scene X i Testing the automatic driving automobile;
obtaining test results
Determining control parameters according to the test result;
calculating a test index according to the determined control parameter;
calculating the variance of the test index, and outputting the test index when the calculated variance is less than or equal to a preset variance threshold value; when the calculated variance is judged to be larger than the preset variance threshold, the cyclic processing is re-executed, and the importance sampling function is updated, namely the updated importance sampling function is determined according to the obtained test result, so as to obtain q (i+1) 。
The embodiment of the invention also provides a computer storage medium, wherein a computer program is stored in the computer storage medium, and the method for realizing the self-adaptive acceleration test evaluation is realized when the computer program is executed by a processor.
The embodiment of the invention also provides a terminal, which comprises: a memory and a processor, the memory storing a computer program; wherein,,
the processor is configured to execute the computer program in the memory;
the computer program, when executed by a processor, implements a method for implementing adaptive accelerated test evaluation as described above.
The embodiment of the invention also provides a device for realizing the self-adaptive acceleration test evaluation, which comprises: the method comprises the steps of determining a sampling function unit, a sampling unit, a testing unit, a parameter determining unit, a calculating unit and a processing unit; wherein,,
the determining sampling function unit is set to: according to the test results of the i-1 th and the previous self-adaptive acceleration tests, determining an importance sampling function of the i-th self-adaptive acceleration test;
the sampling unit is set as: sampling and determining a test scene X of the ith self-adaptive test according to the determined importance sampling function of the ith self-adaptive acceleration test i ；
The test unit is arranged as follows: test scenario X in the determined ith adaptive test i The method comprises the steps of performing self-adaptive acceleration test on an automatic driving automobile to obtain a test result of an ith self-adaptive acceleration test;
the determination parameter unit is set to: determining control parameters of a control variable-based adaptive acceleration evaluation method according to the obtained test result of the ith adaptive acceleration test;
the calculation unit is configured to: calculating a test index of the ith self-adaptive acceleration test according to the determined control parameters;
the processing unit is configured to: when the convergence of the test result is judged according to the calculated test index of the ith self-adaptive acceleration test, the determined test index is determined to be a test evaluation result;
wherein i is an integer greater than 1 but less than or equal to n, n being the number of times of the self-adaptive acceleration test when the test result converges; the test scene X i ～q (i) ，q (i) For the importance sampling function of the ith self-adaptive acceleration test, an initial importance sampling function q (1) =q; the control parameter is a parameter contained in the control variable of the adaptive acceleration evaluation.
The following briefly describes embodiments of the present invention by using examples, which are only used to state embodiments of the present invention and are not used to limit the scope of the present invention.
Application instance
In this application example, taking a cut scene of an automatic driving car as an example, further describing an embodiment of the present invention, fig. 2 is a schematic diagram of the cut scene of the application example of the present invention, as shown in fig. 2, x is a driving direction of the car, record the detected automatic driving car as AV, record a background car as BV, a distance between the BV and the AV in the driving direction of the car is R, and the BV may collide with the AV after changing lanes.
Cutting yardJing Dingyi is
Wherein R represents a relative distance (Range), a method of producing the same, and a method of producing the same>
The natural probability distribution p of the cut scene is shown in fig. 3, the abscissa is the relative velocity in meters per second, and the ordinate is the relative distance in meters;
constructing an initial importance sampling function q based on a full-speed differential model of an autonomous car (1) As shown in fig. 4, the abscissa is relative velocity in meters per second and the ordinate is relative distance in meters.
The application example self-adaptive acceleration test evaluation method (Multiple IS+CV) tests and evaluates the automatic driving automobile, and compared with an Importance Sampling (IS) method and a self-adaptive acceleration test method based on Multiple importance sampling (Multiple IS), the evaluation results of the test indexes are shown in the following figures 5-8; the four graphs show experimental results of 50, 100, 150 and 200 times of test respectively, wherein a curve 1 represents a curve of an Importance Sampling (IS) method, a curve 2 represents a curve of a Multiple importance sampling (Multiple IS) method, a curve 3 represents a curve of an adaptive acceleration test evaluation method (Multiple is+cv), and a black dotted line represents a true value of a test index; from the above graph, as the test times gradually increase, the estimation results of the three methods are all converged to the actual values of the test indexes.
As shown in FIG. 9, referring to FIG. 9, curve 1 represents the curve of the Importance Sampling (IS) method, curve 2 represents the curve of the Multiple importance sampling (Multiple IS) method, curve 3 represents the curve of the adaptive acceleration test evaluation method (Multiple IS+CV), and under the condition that the same estimation error IS reached, the number of tests required by the adaptive acceleration test evaluation method provided by the embodiment of the invention IS far smaller than that of the importance sampling method and the adaptive acceleration test method based on Multiple importance sampling, so that the number of tests can be reduced by about 95% and 80%, respectively, and the test efficiency IS greatly improved.
Those of ordinary skill in the art will appreciate that all or some of the steps, systems, functional modules/units in the apparatus, and methods disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. In a hardware implementation, the division between the functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component may have multiple functions, or one function or step may be performed cooperatively by several physical components. Some or all of the components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as known to those skilled in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computer. Furthermore, as is well known to those of ordinary skill in the art, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media.
Claims (10)
1. A method of implementing adaptive acceleration test evaluation, comprising:
according to the test results of the i-1 th and the previous self-adaptive acceleration tests, determining an importance sampling function of the i-th self-adaptive acceleration test;
sampling and determining a test scene X of the ith self-adaptive test according to the determined importance sampling function of the ith self-adaptive acceleration test i ；
Test scenario X in the determined ith adaptive test i The method comprises the steps of performing self-adaptive acceleration test on an automatic driving automobile to obtain a test result of an ith self-adaptive acceleration test;
determining control parameters of a control variable-based adaptive acceleration evaluation method according to the obtained test result of the ith adaptive acceleration test;
calculating a test index of the ith self-adaptive acceleration test according to the determined control parameters;
when the convergence of the test result is judged according to the calculated test index of the ith self-adaptive acceleration test, the determined test index is determined to be a test evaluation result;
wherein i is an integer greater than 1 but less than or equal to n, n being the number of times of the self-adaptive acceleration test when the test result converges; the test scene X i ～q (i) ，q (i) For the importance sampling function of the ith self-adaptive acceleration test, an initial importance sampling function q (1) =q; the control parameter is a parameter contained in the control variable of the adaptive acceleration evaluation.
2. The method of claim 1, wherein prior to determining the importance sampling function for the ith adaptive acceleration test, the method further comprises:
the natural probability distribution p of the test scene is determined.
3. The method of claim 1, wherein prior to determining the importance sampling function for the ith adaptive acceleration test, the method further comprises:
constructing an initial importance sampling function.
4. A method according to any one of claims 1-3, wherein the test index is determined by calculation of the following formula:
wherein beta is i To control parameters, q α Representation of q (i) Carrying out summation operation, wherein the value of i is 1-n; q α (X i ) For solving for X i Substituted q α Is a value of (2);
representing that beta is summed, and i is 1-n; q (i) (X i ) Representing X i Substitution of q (i) The function value after that; />
For test scene X i A corresponding test result; p (X) i ) Representing test scenario X i Is a natural probability distribution of (c).
5. The method of claim 4, wherein the control parameter is based on minimization
The variance of (c) is determined by the following formula calculation:
6. a method according to any of claims 1-3, wherein said determining an importance sampling function of an ith adaptive acceleration test comprises:
and according to the test results of the i-1 th and previous self-adaptive acceleration tests, determining an importance sampling function of the i-th self-adaptive acceleration test by adopting a Bayesian optimization algorithm BOA.
7. A method according to any one of claims 1-3, characterized in that it is determined whether the test result converges by:
calculating the variance of the test index;
and when the variance of the test index is smaller than or equal to a preset variance threshold, determining that the test result is converged.
8. A computer storage medium having stored therein a computer program which, when executed by a processor, implements the method of implementing adaptive accelerated test assessment of any of claims 1-7.
9. A terminal, comprising: a memory and a processor, the memory storing a computer program; wherein,,
the processor is configured to execute the computer program in the memory;
the computer program, when executed by the processor, implements a method of implementing adaptive acceleration test evaluation according to any one of claims 1-7.
10. An apparatus for implementing adaptive acceleration test evaluation, comprising: the method comprises the steps of determining a sampling function unit, a sampling unit, a testing unit, a parameter determining unit, a calculating unit and a processing unit; wherein,,
the determining sampling function unit is set to: according to the test results of the i-1 th and the previous self-adaptive acceleration tests, determining an importance sampling function of the i-th self-adaptive acceleration test;
the sampling unit is set as: sampling and determining a test scene X of the ith self-adaptive test according to the determined importance sampling function of the ith self-adaptive acceleration test i ；
The test unit is arranged as follows: test scenario X in the determined ith adaptive test i The method comprises the steps of performing self-adaptive acceleration test on an automatic driving automobile to obtain a test result of an ith self-adaptive acceleration test;
the determination parameter unit is set to: determining control parameters of a control variable-based adaptive acceleration evaluation method according to the obtained test result of the ith adaptive acceleration test;
the calculation unit is configured to: calculating a test index of the ith self-adaptive acceleration test according to the determined control parameters;
the processing unit is configured to: when the convergence of the test result is judged according to the calculated test index of the ith self-adaptive acceleration test, the determined test index is determined to be a test evaluation result;
wherein i is an integer greater than 1 but less than or equal to n, n being the number of times of the self-adaptive acceleration test when the test result converges; the test scene X i ～q (i) ，q (i) For the importance sampling function of the ith self-adaptive acceleration test, an initial importance sampling function q (1) =q; the control parameter is a parameter contained in the control variable of the adaptive acceleration evaluation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310070263.9A CN116088476B (en) | 2023-01-12 | Method and device for realizing self-adaptive acceleration test evaluation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310070263.9A CN116088476B (en) | 2023-01-12 | Method and device for realizing self-adaptive acceleration test evaluation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116088476A true CN116088476A (en) | 2023-05-09 |
| CN116088476B CN116088476B (en) | 2024-09-24 |
Family
ID=
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20220004769A1 (en) * | 2020-07-01 | 2022-01-06 | Chang'an University | Method and device for constructing autonomous driving test scenes, terminal and readable storage media |
| CN114544191A (en) * | 2021-12-28 | 2022-05-27 | 文远苏行(江苏)科技有限公司 | Automatic driving test scene quantitative evaluation method and related equipment |
| CN114815605A (en) * | 2022-04-07 | 2022-07-29 | 中国科学院深圳先进技术研究院 | Automatic driving test case generation method and device, electronic equipment and storage medium |
| CN115060291A (en) * | 2022-07-14 | 2022-09-16 | 芜湖雄狮汽车科技有限公司 | Test evaluation method and device for adaptive cruise function, vehicle and storage medium |
| CN115179959A (en) * | 2022-07-18 | 2022-10-14 | 福州大学 | Intelligent driving vehicle behavior prediction method based on self-adaptive updating threshold of driving road |
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20220004769A1 (en) * | 2020-07-01 | 2022-01-06 | Chang'an University | Method and device for constructing autonomous driving test scenes, terminal and readable storage media |
| CN114544191A (en) * | 2021-12-28 | 2022-05-27 | 文远苏行(江苏)科技有限公司 | Automatic driving test scene quantitative evaluation method and related equipment |
| CN114815605A (en) * | 2022-04-07 | 2022-07-29 | 中国科学院深圳先进技术研究院 | Automatic driving test case generation method and device, electronic equipment and storage medium |
| CN115060291A (en) * | 2022-07-14 | 2022-09-16 | 芜湖雄狮汽车科技有限公司 | Test evaluation method and device for adaptive cruise function, vehicle and storage medium |
| CN115179959A (en) * | 2022-07-18 | 2022-10-14 | 福州大学 | Intelligent driving vehicle behavior prediction method based on self-adaptive updating threshold of driving road |
Non-Patent Citations (1)
| Title |
|---|
| 马飒飒;费扬洁;赵守伟;: "安全关键软件防危性评测方法研究", 计算机工程与设计, no. 23, 16 December 2008 (2008-12-16) * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112634181A (en) | Method and apparatus for detecting ground point cloud points | |
| US20240249170A1 (en) | Data source correlation techniques for machine learning and convolutional neural models | |
| CN117233615A (en) | Battery charging process abnormality detection method and device based on comparison learning network | |
| CN115080391A (en) | Method and device for determining automatic driving key scene | |
| CN116088476B (en) | Method and device for realizing self-adaptive acceleration test evaluation | |
| CN116009516B (en) | Method and device for realizing self-adaptive acceleration test | |
| CN106569981A (en) | Statistic parameter determination method and system applicable to large-scale data set | |
| CN116088476A (en) | Method and device for realizing self-adaptive acceleration test evaluation | |
| CN116009516A (en) | Method and device for realizing self-adaptive acceleration test | |
| CN116467599A (en) | Training method for realizing model generated by test scene | |
| CN115170304B (en) | Method and device for extracting risk feature description | |
| CN115510998A (en) | Transaction abnormal value detection method and device | |
| CN115457274A (en) | Vehicle-mounted view angle shielding target detection method and device based on deep learning | |
| CN112508080A (en) | Vehicle model identification method, device, equipment and medium based on experience playback | |
| CN118032374A (en) | Test evaluation method, computer storage medium and terminal | |
| CN118297143A (en) | Self-adaptive test strategy adjustment method and device | |
| Hou et al. | Belief space partitioning for symbolic motion planning | |
| CN117979229B (en) | Advanced aiming precision optimization method, device, equipment and storage medium | |
| CN113395031B (en) | Complex system fault diagnosis method and system based on multistage model | |
| CN115273456A (en) | Method and system for judging illegal driving of two-wheeled electric vehicle and storage medium | |
| Zhang et al. | Variational mode decomposition enabled temporal convolutional network model for state of charge estimation | |
| CN118245763A (en) | Table type intensive reinforcement learning method, computer storage medium and terminal | |
| CN112698394B (en) | Method and device for determining size of design surface element of two-dimensional observation system | |
| CN118226755A (en) | Method, device, computer storage medium and terminal for self-adaptive test evaluation | |
| US20240135805A1 (en) | Apparatus for Predicting Traffic Flow on New Road and Method Thereof |
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 |