Testing method, device, storage medium and electronic device for automatic driving system

Through the quality and data analysis of video clips, the vehicle is simulated to obtain test data of the autonomous driving system, solving the problems of low test accuracy, low data effectiveness and high cost, and improving the testing efficiency and safety of the autonomous driving system.

CN114740759BActive Publication Date: 2024-11-08CHINA FAW CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210405866.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-04-18
Publication Date
2024-11-08
Estimated Expiration
2042-04-18

AI Technical Summary

Technical Problem

The problems of low testing accuracy, low test data effectiveness and high testing cost of autonomous driving systems.

Method used

By obtaining video clips, performing quality analysis and data analysis, obtaining target test parameters, simulate testing vehicles to obtain test data for developing and optimizing autonomous driving systems.

Benefits of technology

It significantly reduces the testing cost of the autonomous driving system, improves the testing accuracy and data effectiveness, improves the safety of the system, ensures the rapid development of autonomous driving algorithms, and provides better protection for drivers.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114740759B_ABST
    Figure CN114740759B_ABST
Patent Text Reader

Abstract

The present invention discloses a test method, device, storage medium and electronic device for an autonomous driving system. The method comprises: obtaining a first number of video clips, wherein the display content of the video clips is used to describe the driving conditions between a first vehicle and a second vehicle, and the display range of the video clips is a road section within a preset range around the first vehicle; performing quality analysis on the first number of video clips to obtain a second number of video clips, wherein the second number is less than or equal to the first number; performing data analysis on the second number of video clips to obtain target test parameters; performing a simulation test on the first vehicle according to the target test parameters to obtain test data, wherein the test data is used to develop and optimize the autonomous driving system. The present invention solves the technical problems of low test accuracy, low test data validity and high test cost of the autonomous driving system in the related art.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving, and in particular to a testing method, device, storage medium and electronic device for an autonomous driving system. Background Art

[0002] With the continuous development of smart cars, the research on the development and testing of autonomous driving systems is also deepening. At present, when testing autonomous driving systems, testers need to drive the test vehicles on real roads for testing, and they also need to apply for driving licenses and test roads before testing, which leads to high testing costs and time-consuming and labor-intensive testing. At the same time, the validity of the test data is low, which makes the test accuracy low. Summary of the invention

[0003] Embodiments of the present invention provide a testing method, device, storage medium and electronic device for an autonomous driving system to at least solve the technical problems of low testing accuracy, low test data validity and high testing cost of the autonomous driving system in related technologies.

[0004] According to one embodiment of the present invention, a method for testing an autonomous driving system is provided, comprising:

[0005] Acquire a first number of video clips, wherein the displayed content of the video clips is used to describe the driving conditions between a first vehicle and a second vehicle, the display range of the video clips is a road section within a preset range around the first vehicle, the first vehicle is a vehicle being tested in the video clips, and the second vehicle is a vehicle in front of the first vehicle; perform quality analysis on the first number of video clips to obtain a second number of video clips, wherein the second number is less than or equal to the first number; perform data analysis on the second number of video clips to obtain target test parameters; perform a simulation test on the first vehicle according to the target test parameters to obtain test data, wherein the test data is used to develop and optimize the autonomous driving system.

[0006] Optionally, the first number of video clips include video clips of multiple types of drivers driving the first vehicle.

[0007] Optionally, performing quality analysis on the first number of video clips to obtain the second number of video clips includes: performing quality analysis on the first number of video clips based on the total number of vehicles displayed in the video clips and the position of the first vehicle to obtain a quality level of each video clip in the first number of video clips; and selecting video clips whose quality levels meet preset standards to obtain the second number of video clips.

[0008] Optionally, data analysis is performed on the second number of video clips to obtain target test parameters, including: obtaining a correlation coefficient matrix based on the scene data in the second number of video clips, wherein the correlation coefficient matrix is ​​composed of correlation coefficients between different scene data; obtaining eigenvalues ​​of different scene data and eigenvectors corresponding to the eigenvalues ​​based on the correlation coefficient matrix; obtaining contribution rates of multiple scene data principal components based on the eigenvalues, wherein the contribution rate is used to reflect the importance of the scene data in the video clips; obtaining a factor loading matrix based on selected principal components and eigenvectors, wherein the selected principal component is a principal component whose cumulative contribution rate is greater than a cumulative contribution rate threshold, and the cumulative contribution rate is obtained by accumulating the contribution rate of each principal component from large to small, and when the accumulated result is greater than the cumulative contribution rate threshold, the principal component to be accumulated is the selected principal component, and the factor loading matrix is ​​composed of corresponding coefficients; constructing an expression based on the factor loading matrix and the scene data, and determining the scene data corresponding to the item with the largest absolute value of the corresponding coefficient in the expression as the target test parameter.

[0009] Optionally, the scene data includes the speed of the first vehicle, the acceleration of the first vehicle, the yaw rate of the first vehicle, the distance between the first vehicle and the second vehicle, the speed of the second vehicle, the acceleration of the second vehicle, the distance between the first vehicle and the third vehicle, the speed of the third vehicle, and the acceleration of the third vehicle.

[0010] Optionally, testing is performed according to target test parameters to obtain test data, including: calculating the average value of the target test parameters in a second number of video clips, and constructing a numerical traversal table based on the average value, wherein the numerical traversal table includes test scenarios and simulation parameters; according to the numerical traversal table, controlling an in-loop test device to test the first vehicle to obtain test data.

[0011] Optionally, the test scenario is a road scenario simulated when the in-loop test device performs testing, and the test scenario corresponds to one or more simulation parameters.

[0012] According to one embodiment of the present invention, there is also provided a testing device for an automatic driving system, comprising:

[0013] A video shooting module is used to obtain a first number of video clips, wherein the video clips are used to reflect the driving conditions between the first vehicle and the second vehicle, and the video clips include road conditions within a preset range around the first vehicle, the first vehicle is the vehicle being tested in the video clips, and the second vehicle is the vehicle in front of the first vehicle; a quality analysis module is used to perform quality analysis on the first number of video clips to obtain a second number of video clips, wherein the second number is less than or equal to the first number; a data analysis module is used to perform data analysis on the second number of video clips to obtain target test parameters; an in-loop test module is used to perform testing according to the target test parameters to obtain test data, wherein the test data is used to develop and optimize the autonomous driving system.

[0014] Optionally, the first number of video clips include video clips of multiple types of drivers driving the first vehicle.

[0015] Optionally, the quality analysis module is also used to perform quality analysis on a first number of video clips based on the total number of vehicles displayed in the video clips and the position of the first vehicle to obtain a quality level of each video clip in the first number of video clips; and select video clips whose quality levels meet preset standards to obtain a second number of video clips.

[0016] Optionally, the data analysis module is further used to obtain a correlation coefficient matrix based on the scene data in the second number of video clips, wherein the correlation coefficient matrix is ​​composed of correlation coefficients between different scene data; according to the correlation coefficient matrix, obtain eigenvalues ​​of different scene data and eigenvectors corresponding to the eigenvalues; according to the eigenvalues, obtain contribution rates of multiple scene data principal components, wherein the contribution rate is used to reflect the importance of the scene data in the video clip; according to the selected principal components and eigenvectors, obtain a factor loading matrix, wherein the selected principal components are principal components whose cumulative contribution rates are greater than a cumulative contribution rate threshold, and the cumulative contribution rate is the accumulation of the contribution rates of each principal component from large to small, and when the accumulated result is greater than the cumulative contribution rate threshold, the principal component to be accumulated is the selected principal component, and the factor loading matrix is ​​composed of corresponding coefficients; construct an expression based on the factor loading matrix and the scene data, and determine the scene data corresponding to the item with the largest absolute value of the corresponding coefficient in the expression as the target test parameter.

[0017] Optionally, the scene data includes the speed of the first vehicle, the acceleration of the first vehicle, the yaw rate of the first vehicle, the distance between the first vehicle and the second vehicle, the speed of the second vehicle, the acceleration of the second vehicle, the distance between the first vehicle and the third vehicle, the speed of the third vehicle, and the acceleration of the third vehicle.

[0018] Optionally, the in-loop test module is also used to calculate the average value of the target test parameter in a second number of video clips, and construct a numerical traversal table based on the average value, wherein the numerical traversal table includes test scenarios and simulation parameters; according to the numerical traversal table, control the in-loop test device to test the first vehicle to obtain test data.

[0019] Optionally, the test scenario is a road scenario simulated when the in-loop test device performs testing, and the test scenario corresponds to one or more simulation parameters.

[0020] According to one embodiment of the present invention, a non-volatile storage medium is further provided, in which a computer program is stored, wherein the computer program is configured to execute any of the above-mentioned methods for testing the autonomous driving system when running on a computer or a processor.

[0021] According to one embodiment of the present invention, there is also provided an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute any of the above-mentioned testing methods for the autonomous driving system.

[0022] In an embodiment of the present invention, a first number of video clips are obtained, quality analysis is performed on the first number of video clips, a second number of video clips are obtained, and data analysis is performed on the second number of video clips to obtain target test parameters. A simulation test is performed on the first vehicle according to the target test parameters to obtain test data for developing and optimizing the automatic driving system. This can significantly reduce the test cost of the automatic driving system, and a large amount of effective test data can be obtained in a short time, thereby improving the accuracy of the automatic driving system test. At the same time, it also helps to improve the safety of the automatic driving system, ensure the rapid development of the automatic driving algorithm, and provide better protection for the driver. It is of great significance to improve the applicability of the automatic driving system, ensure automobile safety, and reduce traffic accidents, thereby solving the technical problems of low test accuracy, low test data validity, and high test cost of the automatic driving system test in related technologies. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0024] Figure 1 is a flow chart of a method for testing an automatic driving system according to one embodiment of the present invention;

[0025] Figure 2 is a schematic diagram of a video clip according to one embodiment of the present invention;

[0026] Figure 3 is a schematic diagram of the classification of the position of the first vehicle according to one embodiment of the present invention;

[0027] Figure 4 is a schematic structural diagram of an in-loop test device according to one embodiment of the present invention;

[0028] Figure 5 is a data flow diagram according to one embodiment of the present invention;

[0029] Figure 6 4 is a structural block diagram of a testing device for an automatic driving system according to one embodiment of the present invention. DETAILED DESCRIPTION

[0030] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0031] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0032] According to one embodiment of the present invention, an embodiment of a testing method for an autonomous driving system is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in an order different from that shown here.

[0033] The method embodiment can be executed in an electronic device including a memory and a processor in a vehicle, a similar control device, or an automatic driving system. Taking the operation on the electronic device of the vehicle as an example, the electronic device of the vehicle may include one or more processors and a memory for storing data. Optionally, the electronic device of the above-mentioned vehicle may also include a communication device and a display device for communication functions. It can be understood by those of ordinary skill in the art that the above-mentioned structural description is only illustrative and does not limit the structure of the electronic device of the above-mentioned vehicle. For example, the electronic device of the vehicle may also include more or fewer components than the above-mentioned structural description, or have a configuration different from the above-mentioned structural description.

[0034] The processor may include one or more processing units. For example, the processor may include a central processing unit (CPU), a graphics processing unit (GPU), a digital signal processing (DSP) chip, a microcontroller unit (MCU), a field-programmable gate array (FPGA), a neural network processor (NPU), a tensor processing unit (TPU), an artificial intelligence (AI) type processor, and the like. Among them, different processing units may be independent components or integrated into one or more processors. In some instances, the electronic device may also include one or more processors.

[0035] The memory can be used to store computer programs, such as storing a computer program corresponding to the test method of the autonomous driving system in the embodiment of the present invention. The processor implements the test method of the autonomous driving system by running the computer program stored in the memory. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory may further include a memory remotely located relative to the processor, and these remote memories may be connected to the electronic device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0036] The communication device is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the communication device includes a network adapter (network interface controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the communication device can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0037] The display device may be, for example, a touch screen type liquid crystal display (LCD) and a touch display (also referred to as a "touch screen" or "touch display screen"). The liquid crystal display may enable a user to interact with a user interface of a mobile terminal. In some embodiments, the mobile terminal has a graphical user interface (GUI), and a user may interact with the GUI by finger contact and / or gestures on a touch-sensitive surface, wherein the human-computer interaction functions here may optionally include the following interactions: creating web pages, drawing, word processing, making electronic documents, games, video conferencing, instant messaging, sending and receiving emails, call interfaces, playing digital videos, playing digital music and / or web browsing, etc. The executable instructions for executing the above human-computer interaction functions are configured / stored in a computer program product or a readable storage medium executable by one or more processors.

[0038] In this embodiment, a method for testing an automatic driving system of an electronic device running on the above vehicle is provided. Figure 1 is a flow chart of a method for testing an automatic driving system according to one embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0039] Step S101: Obtain a first number of video clips.

[0040] The displayed content of the video clip is used to describe the driving situation between the first vehicle and the second vehicle, the first vehicle is the vehicle being tested in the video clip, and the second vehicle is the vehicle in front of the first vehicle. It can be understood that the video clip is shot on a real road, and the video clip includes the driving action of the first vehicle, the driving action of the second vehicle, and the interaction between the first vehicle and the second vehicle. For example, the interaction may be that the first vehicle changes its own driving speed according to the driving speed of the second vehicle.

[0041] The display range of the video clip is the road section within the preset range around the first vehicle. The preset range includes a range of 200m in front and behind the first vehicle in the longitudinal direction, and the longitudinal direction is the direction in which the first vehicle is traveling. The preset range includes the first lane where the first vehicle is located, the second lane on the left side of the first lane, the third lane on the right side of the first lane, and multiple reference objects of the same size arranged at equal intervals on both sides of the road in the transverse direction. The transverse direction is a direction perpendicular to the longitudinal direction. Among them, shooting the lanes around the first lane can more realistically reflect the road conditions, and the purpose of setting the reference objects is to calculate the distance between vehicles when processing the video clip.

[0042] like Figure 2 As shown, it is a schematic diagram of a video clip according to one embodiment of the present invention. Figure 2 The video clips shown include a first vehicle, a second vehicle, a plurality of other vehicles, and a plurality of reference objects of the same size and arranged at equal intervals on both sides of the road. The drone follows the first vehicle above the first vehicle to shoot, and the drone is used to shoot a first number of video clips. A special color sticker may be attached to the roof of the first vehicle to facilitate the screening and extraction of the first vehicle when processing the video clips shot by the drone.

[0043] Step S102: Perform quality analysis on the first number of video clips to obtain a second number of video clips.

[0044] The second number is less than or equal to the first number. Since there may be some video clips with low validity among the video clips shot by the drone, the authenticity and accuracy of the data obtained from these video clips with low validity are also low. Therefore, in order to ensure the authenticity and accuracy of the data, that is, to ensure the validity of the video clips, the first number of video clips will be subjected to quality analysis, and video clips that meet the quality requirements will be selected, thereby obtaining the second number of video clips.

[0045] Step S103: performing data analysis on the second number of video clips to obtain target test parameters.

[0046] Data analysis is performed on a second number of video clips that can ensure the authenticity and accuracy of the data, and the obtained target test parameters can be used to simulate the most realistic road scene.

[0047] Step S104: Perform a simulation test on the first vehicle according to the target test parameters to obtain test data.

[0048] The test data is used to develop and optimize the autonomous driving system. After the first vehicle is simulated and tested according to the target test parameters that are closest to the real road scene, the test data obtained can be used to develop and optimize the autonomous driving system, thereby improving the applicability of the autonomous driving system, ensuring vehicle safety, and reducing traffic accidents.

[0049] Through the above steps, a first number of video clips are obtained, and the quality analysis of the first number of video clips is performed to obtain a second number of video clips. Then, data analysis is performed on the second number of video clips to obtain target test parameters. The first vehicle is simulated and tested according to the target test parameters to obtain test data for developing and optimizing the automatic driving system. This can significantly reduce the testing cost of the automatic driving system, and can obtain a large amount of effective test data in a short time, improve the accuracy of the automatic driving system test, and also help improve the safety of the automatic driving system, ensure the rapid development of the automatic driving algorithm, and provide better protection for the driver. It is of great significance to improve the applicability of the automatic driving system, ensure the safety of the car, and reduce traffic accidents, thereby solving the technical problems of low accuracy of automatic driving system testing, low validity of test data, and high testing cost in related technologies.

[0050] Optionally, the first number of video clips include video clips of multiple types of drivers driving the first vehicle.

[0051] In order to make the acquired video clips and driving data widely representative, multiple types of drivers include novice drivers and non-novice drivers (skilled drivers), gender includes male drivers and female drivers, and age covers drivers of all age groups that meet driving conditions. The total number of drivers is not less than 50, which can be understood as the first number is not less than 50.

[0052] Optionally, in step S102, performing quality analysis on the first number of video clips to obtain the second number of video clips may include the following execution steps:

[0053] Step S102a: performing quality analysis on the first number of video segments according to the total number of vehicles displayed in the video segments and the position of the first vehicle to obtain a quality level of each video segment in the first number of video segments.

[0054] A quality analysis is performed on the first number of video clips from two dimensions: the total number of vehicles and the position of the first vehicle. The total number of vehicles is the total number of vehicles contained in the video clip, and the total number of vehicles is divided into three levels, namely S1, M1 and L1. For example, the S1 level means that the total number of vehicles contained in the video clip is less than 3, the M1 level means that the total number of vehicles contained in the video clip is less than or equal to 8 and greater than or equal to 3, and the L1 level means that the total number of vehicles contained in the video clip is greater than 3.

[0055] Since the drone may not be directly above the first vehicle when shooting, it may not be able to follow the first vehicle in time. Therefore, the validity of the video clip needs to consider the position of the first vehicle. The position of the first vehicle is the relative position of the first vehicle in the video clip and the longitudinal center of the video clip. The position of the first vehicle is divided into three levels, namely S2, M2 and L2. Figure 3 As shown, it shows a schematic diagram of the classification of the position of the first vehicle, according to Department, Department, Chuhe The dividing line at the top divides the drone shooting range in the vertical direction, that is, divides the video clips. S2 level means that the position of the first vehicle is within the drone shooting range. and Within the dividing line, M2 level means the first vehicle is located within the drone’s shooting range. and Within the demarcation line or and Within the dividing line, L2 level means that the position of the first vehicle is within the drone’s shooting range. Outside the dividing line or Outside the dividing line.

[0056] Specifically, the total number of vehicles displayed in the video clip and the position of the first vehicle are input into the fuzzy controller to perform quality analysis on the video clip to obtain the quality level of the video clip. The mode of designing the fuzzy controller is a dual-input single-output mode. The first input variable of the fuzzy controller is the total number of vehicles in the video clip, and the corresponding fuzzy set is divided into the three levels of S1 level, M1 level and L1 level mentioned above. The second input variable of the fuzzy controller is the position of the first vehicle, and the corresponding fuzzy set is divided into the three levels of S2 level, M2 level and L2 level mentioned above. The output variable of the fuzzy controller is the quality of the video clip. Exemplarily, the membership function of the fuzzy controller selects a steeper membership function, such as a triangular membership function, to ensure that the output variable is sufficiently sensitive to the change of the input variable, so that the corresponding fuzzy set is divided into five levels: S level (excellent), A level (good), B level (average), C level (poor) and D level (extremely poor). That is, the video quality level output by the fuzzy controller is S level (excellent), A level (good), B level (average), C level (poor) and D level (extremely poor). The fuzzy rules between the input and output of the fuzzy controller are shown in the following table:

[0057] Table 1

[0058]

[0059] Step S102b: Select video clips whose quality levels meet the preset standard to obtain a second number of video clips.

[0060] Based on the fuzzy controller constructed in step 102a, the total number of vehicles in each video segment of the first number of video segments and the position of the first vehicle are input into the fuzzy controller as the first input variable and the second input variable, and the fuzzy controller outputs the quality level of the corresponding video segment, and selects the video segment whose quality level meets the preset standard. The video segment that meets the preset standard is the video segment whose quality level is B or above output by the fuzzy controller, and the video segment whose quality level is B or above is the second number of video segments.

[0061] Optionally, in step S103, performing data analysis on the second number of video clips to obtain target test parameters may include the following execution steps:

[0062] Step S103a: Obtain a correlation coefficient matrix according to the scene data in the second number of video clips.

[0063] The scene data includes the speed of the first vehicle, the acceleration of the first vehicle, the yaw rate of the first vehicle, the distance between the first vehicle and the second vehicle, the speed of the second vehicle, the acceleration of the second vehicle, the distance between the first vehicle and the third vehicle, the speed of the third vehicle and the acceleration of the third vehicle, which are recorded as column vectors in the form of O1, O2, O3, O4, O5, O6, O7, O8 and O9 respectively, and the third vehicle is the vehicle behind the first vehicle. For example, the video clip includes 9 sampling points (hereinafter, each video clip includes 9 sampling points as an example, that is, the number of sampling points k=9), O 11 is the speed of the first vehicle at the first sampling point, O 21 is the speed of the first vehicle at the second sampling point, and so on. The column vector is formed by the speed of the first vehicle at each sampling point. The same is true for O2, O3, O4, O5, O6, O7, O8 and O9, which will not be elaborated here.

[0064] Since the second number of video clips contain more content, it is necessary to extract the dimensions of the scene data, select the main scene data, and focus on testing it in the subsequent automatic driving system test. ij (i,j=1,2,…,9) are different scene data O i With O j The correlation coefficient between the different scene data is calculated according to the scene data in the second number of video clips. i With O j The correlation coefficient between Among them, O ki Represents the scene data O at the kth sampling point in the video clip i , Represents the scene data O in the second number of video clips i And according to the scene data in the second number of video clips, calculate the scene data correlation coefficient matrix Among them, the correlation coefficient matrix is ​​composed of the correlation coefficients between different scene data.

[0065] Step S103b: Obtain eigenvalues ​​of different scene data and eigenvectors corresponding to the eigenvalues ​​according to the correlation coefficient matrix.

[0066] By solving the characteristic equation (λI-R)u=0, the characteristic value λ is obtained i (i=1, 2, ..., 9) and each eigenvalue λ i The corresponding eigenvector U = (u1, u2…, u9), where I is the unit matrix. Eigenvalue λ i After solving, the eigenvalue λ iArrange from large to small, for example, λ1≥λ2≥…≥λ9≥0 (the following text uses λ1≥λ2≥…≥λ9≥0 as an example for explanation).

[0067] Step S103c: Obtain contribution rates of principal components of multiple scene data according to the eigenvalues.

[0068] Among them, the contribution rate is used to reflect the importance of scene data in the video clip. i The contribution rate of Represents the contribution rate of each scene data in all the second number of video clips. Calculate the principal component Z of each scene data i The contribution rate of the scene data principal component Z i The contribution rates are arranged from large to small, for example, Z1≥Z2≥…≥Z9≥0 (Z1≥Z2≥…≥Z9≥0 is taken as an example in the following text), and Z1 with the largest contribution rate corresponds to λ1 with the largest eigenvalue, Z2 corresponds to λ2, and so on.

[0069] Step S103d: Obtain a factor loading matrix based on the selected principal components and eigenvectors.

[0070] Among them, the selected principal component is the principal component whose cumulative contribution rate is greater than the cumulative contribution rate threshold. The cumulative contribution rate is obtained by accumulating the contribution rates of each principal component from large to small. When the accumulated result is greater than the cumulative contribution rate threshold, the accumulated principal component is the selected principal component.

[0071] For example, there are 6 scene data principal components A, B, C, D, E, and F, among which the contribution rate of A is 25%, the contribution rate of B is 20%, the contribution rate of C is 15%, the contribution rate of D is 18%, the contribution rate of E is 12%, and the contribution rate of F is 10%. The contribution rates of the scene data principal components are arranged from large to small, that is, 25%>20%>18%>15%>12%>10% (abbreviated as A>B>D>C>E>F). According to the order of contribution rate from large to small, the contribution rate of A and the contribution rate of B are accumulated first, and the accumulated result is 25%+20%=45% (abbreviated as A+B=45%), and then the contribution rate of A, the contribution rate of B, and the contribution rate of C are accumulated, and the accumulated result is 25%+20%+18%=63% (abbreviated as A+B+C=63%), and so on. Taking the cumulative contribution rate threshold of 80% as an example, when the contribution rate of A is added to the contribution rate of E, the cumulative result is 25% + 20% + 18% + 15% + 12% = 90% (abbreviated as A + B + D + C + E = 90%), that is, the cumulative contribution rate of 90% is greater than the contribution rate threshold of 80%. Therefore, the selected principal components (the principal components whose cumulative contribution rate is greater than the contribution rate threshold) are A, B, D, C and E.

[0072] Calculate the cumulative contribution rate The selected principal components are obtained (hereinafter, the selected principal components are taken as Z1, Z2, Z3 and Z4 as an example), and the factor loading matrix is ​​calculated based on the selected principal components:

[0073]

[0074] Among them, the factor loading matrix A is composed of the corresponding coefficients (i.e. a 11 …a 91 、a 12 …a 92 、a 13 …a 93 、a 14 …a 94 ) constitutes, ω j is the standardized feature vector, that is

[0075] Step S103e: construct an expression according to the factor loading matrix and the scenario data, and determine the scenario data corresponding to the item with the largest absolute value of the corresponding coefficient in the expression as the target test parameter.

[0076] The expressions constructed based on the factor loading matrix and scenario data are as follows (there are 4 principal components selected, so 4 expressions are constructed):

[0077] M1=a 11 O1+a 21 O2+…+a 91 O9

[0078] M2=a 12 O1+a 22 O2+…+a 92 O9

[0079] M3=a 13 O1+a 23 O2+…+a 93 O9

[0080] M4=a 14 O1+a 24 O2+…+a 94 O9

[0081] Since the greater the absolute value of the corresponding coefficient, the more important the scene data corresponding to the item is in the constructed expression, the scene data corresponding to the item with the largest absolute value of the corresponding coefficient in the constructed expressions M1, M2, M3 and M4 are selected respectively. 91 、a 12 、a 23 、a 54Taking the maximum absolute value as an example, the corresponding scene data are O9, O1, O2, and O5, that is, the target test parameters are determined to be O9, O1, O2, and O5.

[0082] Optionally, in step S104, performing a test according to the target test parameters to obtain test data may include the following execution steps:

[0083] Step S104a, calculating the average value of the target test parameter in the second number of video clips, and constructing a numerical traversal table according to the average value;

[0084] Step S104b: According to the numerical traversal table, the in-loop test device is controlled to test the first vehicle to obtain test data.

[0085] Calculate the average values ​​of the target test parameters O9, O1, O2, and O5 in the second number of video clips to obtain The constructed numerical traversal table is shown in Table 2 below. The numerical traversal table includes test scenarios and simulation parameters. The test scenario is a road scenario simulated when the in-loop test device is tested. The simulation parameters are parameters in the simulated road scenario. The test scenario corresponds to one or more simulation parameters. k1 and k2 are variable coefficients. The numerical traversal table is set in the industrial computer of the in-loop test device (i.e., in the PreScan software) to form a corresponding test scenario, so that the first vehicle can be tested and test data for developing and optimizing the autonomous driving system can be obtained.

[0086] Table 2

[0087]

[0088] like Figure 4 As shown, it is a schematic diagram of the structure of the in-loop test device. A motor base plate 10 and an electric cylinder base plate 15 are fixed on the ground. A motor 9 and an electric cylinder 14 are respectively arranged on the top surfaces of the motor base plate 10 and the electric cylinder base plate 15. A power shaft 8 is arranged at the top power output end of the motor 9. A second bevel gear 7 is fixed to the top end of the power shaft 8. The rear portion of the second bevel gear 7 is in contact and meshing contact with the lower portion of the first bevel gear 6. The first bevel gear 6 is fixed to the front side of the left moving wheel 4. The left outer periphery of the left moving wheel 4 is in contact with the left inner periphery of the moving belt 3. The right inner periphery of the moving belt 3 is in contact with the right outer periphery of the right moving wheel 5. A central shaft 11 is fixed to the front side of the right moving wheel 5. The outer periphery of the central shaft 11 is in contact with the inner periphery of the bearing sleeve 12. The lower outer periphery of the bearing sleeve 12 is fixedly connected to the top end of the action rod 13. The upper outer periphery of the moving belt 3 is in contact with the bottom wheel of the test vehicle (i.e., the first vehicle). A display bracket 1 is fixed to the top end of the right side of the test vehicle. A display 2 is fixed to the right end of the display bracket 1. The display 2 is used to display a simulated road scene.

[0089] like Figure 5 The figure shows a data flow diagram of the test method of the automatic driving system. After the numerical traversal table is input into the industrial computer, the industrial computer will transmit the road information to the electric cylinder controller and the motor controller according to the data in the numerical traversal table. The electric cylinder controller and the motor controller control the electric cylinder 14 and the motor 9 to start working. The movement of the motor 9 will drive the motion belt 3 to start moving, thereby simulating the relative movement of the vehicle and the road during actual driving. The movement of the electric cylinder 14 will change the horizontal angle of the motion belt 3, thereby simulating the slope of the road during actual driving. At the same time, the industrial computer will also transmit the scene data (including simulation parameters and other scene data) to the automatic driving controller according to the data in the numerical traversal table. The automatic driving controller will transmit the automatic driving control instructions to the test vehicle to control the test vehicle to perform automatic driving (to ensure safety and better obtain test data, there will usually be a driver in the test vehicle). The industrial computer will generate a simulation video according to the data in the numerical traversal table and transmit it to the display 2. The display 2 displays the virtual road scene in real time, so that the driver sitting in the test vehicle can get a better virtual driving experience. The vehicle status data of the test vehicle during the test, that is, the obtained test data, will also be fed back to the industrial computer for the development and optimization of the automatic driving system.

[0090] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0091] In this embodiment, a test device for an autonomous driving system is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and will not be repeated hereafter. As used below, the term "module" may be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0092] Figure 6 is a structural block diagram of a test device for an automatic driving system according to one embodiment of the present invention. Figure 6As shown, a test device 600 for an autonomous driving system is used as an example, and the device includes: a video shooting module 601, used to obtain a first number of video clips, wherein the video clips are used to reflect the driving conditions between the first vehicle and the second vehicle, and the video clips include road conditions within a preset range around the first vehicle, the first vehicle is the vehicle tested in the video clip, and the second vehicle is the vehicle in front of the first vehicle; a quality analysis module 602, used to perform quality analysis on the first number of video clips to obtain a second number of video clips, wherein the second number is less than or equal to the first number; a data analysis module 603, used to perform data analysis on the second number of video clips to obtain target test parameters; an in-the-loop test module 604, used to perform testing according to the target test parameters to obtain test data, wherein the test data is used to develop and optimize the autonomous driving system.

[0093] Optionally, the first number of video clips include video clips of multiple types of drivers driving the first vehicle.

[0094] Optionally, the quality analysis module 602 is also used to perform quality analysis on a first number of video clips based on the total number of vehicles displayed in the video clips and the position of the first vehicle to obtain a quality level of each video clip in the first number of video clips; and select video clips whose quality levels meet preset standards to obtain a second number of video clips.

[0095] Optionally, the data analysis module 603 is further used to obtain a correlation coefficient matrix based on the scene data in the second number of video clips, wherein the correlation coefficient matrix is ​​composed of correlation coefficients between different scene data; according to the correlation coefficient matrix, obtain eigenvalues ​​of different scene data and eigenvectors corresponding to the eigenvalues; according to the eigenvalues, obtain contribution rates of multiple scene data principal components, wherein the contribution rate is used to reflect the importance of the scene data in the video clip; according to the selected principal components and eigenvectors, obtain a factor loading matrix, wherein the selected principal component is a principal component whose cumulative contribution rate is greater than a cumulative contribution rate threshold, and the cumulative contribution rate is accumulated from large to small. When the accumulated result is greater than the cumulative contribution rate threshold, the principal component to be accumulated is the selected principal component, and the factor loading matrix is ​​composed of corresponding coefficients; an expression is constructed according to the factor loading matrix and the scene data, and the scene data corresponding to the item with the largest absolute value of the corresponding coefficient in the expression is determined as the target test parameter.

[0096] Optionally, the scene data includes the speed of the first vehicle, the acceleration of the first vehicle, the yaw rate of the first vehicle, the distance between the first vehicle and the second vehicle, the speed of the second vehicle, the acceleration of the second vehicle, the distance between the first vehicle and the third vehicle, the speed of the third vehicle, and the acceleration of the third vehicle.

[0097] Optionally, the in-loop test module 604 is also used to calculate the average value of the target test parameter in the second number of video clips, and construct a numerical traversal table based on the average value, wherein the numerical traversal table includes test scenarios and simulation parameters; according to the numerical traversal table, control the in-loop test device to test the first vehicle to obtain test data.

[0098] Optionally, the test scenario is a road scenario simulated when the in-loop test device performs testing, and the test scenario corresponds to one or more simulation parameters.

[0099] It should be noted that the above modules can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.

[0100] An embodiment of the present invention further provides a non-volatile storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when running on a computer or a processor.

[0101] Optionally, in this embodiment, the non-volatile storage medium may be configured to store a computer program for performing the following steps:

[0102] Step S1, obtaining a first number of video clips;

[0103] Step S2, performing quality analysis on the first number of video clips to obtain a second number of video clips;

[0104] Step S3, performing data analysis on the second number of video clips to obtain target test parameters;

[0105] Step S4, performing a simulation test on the first vehicle according to the target test parameters to obtain test data.

[0106] Optionally, in this embodiment, the above-mentioned non-volatile storage medium may include but is not limited to: a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and other media that can store computer programs.

[0107] An embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0108] Optionally, in this embodiment, the processor in the electronic device may be configured to run a computer program to perform the following steps:

[0109] Step S1, obtaining a first number of video clips;

[0110] Step S2, performing quality analysis on the first number of video clips to obtain a second number of video clips;

[0111] Step S3, performing data analysis on the second number of video clips to obtain target test parameters;

[0112] Step S4, performing a simulation test on the first vehicle according to the target test parameters to obtain test data.

[0113] Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments and optional implementation modes, and this embodiment will not be described in detail here.

[0114] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0115] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0116] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0117] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0118] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0119] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM), random access memory (RAM), mobile hard disk, magnetic disk or optical disk, etc., various media that can store program codes.

[0120] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for testing an automatic driving system, characterized in that: The method comprises: Acquire a first number of video clips, wherein display contents of the video clips are used to describe driving conditions between a first vehicle and a second vehicle, the display range of the video clips is a road section within a preset range around the first vehicle, the first vehicle is a vehicle being tested in the video clips, and the second vehicle is a vehicle in front of the first vehicle; Performing quality analysis on the first number of video clips to obtain a second number of video clips, wherein the second number is less than or equal to the first number; performing data analysis on the second number of video clips to obtain target test parameters; Performing a simulation test on the first vehicle according to the target test parameters to obtain test data, wherein the test data is used to develop and optimize the automatic driving system; Among them, the quality analysis of the first number of video clips to obtain the second number of video clips includes: performing quality analysis on the first number of video clips according to the total number of vehicles displayed in the video clips and the positions of the first vehicles to obtain the quality level of each of the first number of video clips; selecting video clips whose quality levels meet preset standards to obtain the second number of video clips.

2. The method according to claim 1, characterized in that The first number of video clips include video clips of multiple types of drivers driving the first vehicle.

3. The method according to claim 1, characterized in that The performing data analysis on the second number of video clips to obtain target test parameters comprises: Obtaining a correlation coefficient matrix according to the scene data in the second number of video clips, wherein the correlation coefficient matrix is ​​composed of correlation coefficients between different scene data; According to the correlation coefficient matrix, eigenvalues ​​of different scene data and eigenvectors corresponding to the eigenvalues ​​are obtained; Obtaining contribution rates of a plurality of principal components of the scene data according to the eigenvalues, wherein the contribution rates are used to reflect the importance of the scene data in the video clip; According to the selected principal component and the eigenvector, a factor loading matrix is ​​obtained, wherein the selected principal component is the principal component whose cumulative contribution rate is greater than the cumulative contribution rate threshold, and the cumulative contribution rate is the cumulative contribution rate of each principal component from large to small. When the cumulative result is greater than the cumulative contribution rate threshold, the principal component to be accumulated is the selected principal component, and the factor loading matrix is ​​composed of corresponding coefficients; An expression is constructed according to the factor loading matrix and the scenario data, and the scenario data corresponding to the item with the largest absolute value of the corresponding coefficient in the expression is determined as the target test parameter.

4. The method according to claim 3, characterized in that The scene data includes the speed of the first vehicle, the acceleration of the first vehicle, the yaw rate of the first vehicle, the distance between the first vehicle and the second vehicle, the speed of the second vehicle, the acceleration of the second vehicle, the distance between the first vehicle and a third vehicle, the speed of the third vehicle, and the acceleration of the third vehicle.

5. The method according to claim 1, characterized in that: The testing according to the target test parameters to obtain the test data includes: Calculating an average value of the target test parameter in the second number of video clips, and constructing a numerical traversal table according to the average value, wherein the numerical traversal table includes test scenarios and simulation parameters; According to the numerical traversal table, an in-loop test device is controlled to test the first vehicle to obtain test data.

6. The method according to claim 5, characterized in that The test scenario is a road scenario simulated when the in-the-loop test device performs a test, and the test scenario corresponds to one or more of the simulation parameters.

7. A test device for an automatic driving system, characterized in that: The device comprises: a video shooting module, configured to obtain a first number of video clips, wherein the video clips are used to reflect driving conditions between a first vehicle and a second vehicle, the video clips include road conditions within a preset range around the first vehicle, the first vehicle is a vehicle being tested in the video clips, and the second vehicle is a vehicle in front of the first vehicle; a quality analysis module, configured to perform quality analysis on the first number of video clips to obtain a second number of video clips, wherein the second number is less than or equal to the first number; a data analysis module, configured to perform data analysis on the second number of video clips to obtain target test parameters; an in-loop test module, configured to perform a test according to the target test parameters to obtain test data, wherein the test data is used to develop and optimize the autonomous driving system; Among them, the quality analysis module is also used to: perform quality analysis on the first number of video clips according to the total number of vehicles displayed in the video clips and the position of the first vehicle to obtain the quality level of each of the first number of video clips; select video clips whose quality levels meet preset standards to obtain the second number of video clips.

8. A non-volatile storage medium, characterized in that: The non-volatile storage medium stores a computer program, wherein the computer program is configured to execute the test method for the autonomous driving system described in any one of claims 1 to 6 when running on a computer or a processor.

9. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to run the computer program to execute the testing method of the autonomous driving system as described in any one of claims 1 to 6 above.