Test method and device for autonomous vehicle

Abstract

The embodiment of the application discloses a kind of test method and device for automatic driving vehicle, it is related to automatic driving technical field.The method includes: obtaining the bad example data of the automatic driving vehicle to be tested;According to the bad example data, scene enumeration is carried out, and the scene obtained by enumeration includes the automatic driving vehicle to be tested, or includes the combination of the automatic driving vehicle to be tested and at least one obstacle contained in the bad example data;The scene obtained by enumeration is simulated respectively, and the abnormal scene of the same abnormal type as the bad example data is determined;Using the abnormal scene, the responsibility entity of the bad example data is determined from the automatic driving vehicle and the obstacle contained in the bad example data.The application can improve the accuracy and efficiency of bad example analysis, reduce labor cost.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic driving, in particular to a test method and device for an automatic driving vehicle. BACKGROUND

[0002] An automatic driving vehicle relies on perception sensors, artificial intelligence, global positioning systems and other collaborative cooperation to enable the vehicle to safely and automatically drive. A large number of tests are required for the landing of an automatic driving vehicle, and a large number of tests are also required for the iteration of algorithms. Therefore, simulation testing can greatly accelerate the algorithm iteration efficiency and product optimization speed. A large number of bad cases (badcase) are generated during the road testing of an automatic driving vehicle, such as sudden braking during a turn. Then, the bad cases need to be manually analyzed to label the responsible entity. However, this method relies on human experience, has low accuracy, high labor cost and low efficiency, and is limited by the problem of human efficiency, which limits the test coverage. SUMMARY

[0003] Therefore, the present application provides a test method and device for an automatic driving vehicle to improve the accuracy and efficiency of bad case analysis and reduce labor costs.

[0004] The present application provides the following solutions:

[0005] In a first aspect, a test method for an automatic driving vehicle is provided, and the method comprises:

[0006] Obtaining bad case data of a tested automatic driving vehicle;

[0007] Performing scene enumeration according to the bad case data, wherein the enumerated scenes include the tested automatic driving vehicle or a combination of the tested automatic driving vehicle and at least one obstacle included in the bad case data;

[0008] Respectively performing simulation testing on the enumerated scenes to determine an abnormal scene of the same abnormal type as the bad case data;

[0009] Determining a responsible entity of the bad case data from the tested automatic driving vehicle and the obstacle included in the bad case data by using the abnormal scene.

[0010] According to an implementable manner in the embodiments of the present application, the scene enumeration according to the bad case data comprises:

[0011] determine entities in the first region and / or the second region according to the bad case data, the entities including the tested autonomous vehicle and obstacles, the first region being a region within a preset first distance range from a position of the tested autonomous vehicle when an anomaly corresponding to the bad case data occurs, and the second region being a driving region determined according to a navigation path of the tested autonomous vehicle;

[0012] combine the entities to enumerate scenarios, wherein at least the tested autonomous vehicle is included in each combination.

[0013] According to an implementable manner in the embodiments of the present application, the simulation test on the enumerated scenarios respectively includes:

[0014] the simulation test on the enumerated scenarios is sequentially performed according to an order from less to more of the number of entities included in the scenarios, wherein the entities include the tested autonomous vehicle or obstacles;

[0015] if the same anomaly type as the bad case data occurs in the scenario currently subjected to the simulation test, the scenario currently subjected to the simulation test is determined as an abnormal scenario, and the simulation test on the enumerated scenarios is ended; otherwise, the next scenario is subjected to the simulation test.

[0016] According to an implementable manner in the embodiments of the present application, determining the responsible entity of the bad case data from the autonomous vehicle and the obstacles included in the bad case data by using the abnormal scenario includes:

[0017] if only the tested autonomous vehicle is included in the abnormal scenario, the tested autonomous vehicle is determined as the responsible entity of the bad case data;

[0018] if a combination of the tested autonomous vehicle and at least one obstacle is included in the abnormal scenario, the obstacle included in the abnormal scenario is determined as the responsible entity.

[0019] According to an implementable manner in the embodiments of the present application, the method further includes:

[0020] performing generalization processing on motion information of the responsible entity in the abnormal scenario to obtain a generalized scenario;

[0021] output information of the generalized scenario.

[0022] According to an implementable manner in the embodiments of the present application, before the information of the generalized scenario is output, the method further includes:

[0023] performing simulation test on the generalized scenario to screen a generalized scenario that has the same anomaly type as the bad case data;

[0024] The outputting the information of the generalized scenario includes: outputting the information of the generalized scenario screened.

[0025] According to a possible implementation of the embodiments of the present application, the abnormal type corresponding to the bad example data includes: sudden braking, collision, or driving risk exceeding a preset risk level.

[0026] In a second aspect, a testing device for an automatic driving vehicle is provided, and the device includes:

[0027] a bad example acquisition unit configured to acquire bad example data of a tested automatic driving vehicle;

[0028] a scenario enumeration unit configured to perform scenario enumeration according to the bad example data, and the enumerated scenarios include the tested automatic driving vehicle or a combination of the tested automatic driving vehicle and at least one obstacle included in the bad example data;

[0029] a simulation testing unit configured to perform simulation testing on the enumerated scenarios respectively, and determine an abnormal scenario with the same abnormal type as the bad example data;

[0030] a responsibility presumption unit configured to determine a responsible entity of the bad example data from the tested automatic driving vehicle and the obstacle included in the bad example data by using the abnormal scenario.

[0031] According to a third aspect, a computer readable storage medium is provided, and the medium stores a computer program, and the program is executed by a processor to implement the steps of the method in any one of the first aspect.

[0032] According to a fourth aspect, an electronic device is provided, and the device includes:

[0033] one or more processors; and

[0034] a memory associated with the one or more processors, and the memory is configured to store program instructions, and the program instructions are executed by the one or more processors to implement the steps of the method in any one of the first aspect.

[0035] According to the embodiments of the present application, the following technical effects are disclosed:

[0036] 1) According to the embodiments of the present application, the scenario enumeration is performed according to the bad example data, and the simulation testing is performed on each of the enumerated scenarios to determine an abnormal scenario with the same abnormal type as the bad example data, and then the responsible entity of the bad example data is determined by using the abnormal scenario. The simulation testing based on the scenario enumeration can realize the automatic analysis of the bad example, and does not need to rely on the manual experience for analysis and annotation, thereby reducing the labor cost, improving the efficiency, being easy to scale, and having higher accuracy.

[0037] 2) The application combines obstacles and the tested autonomous vehicle from the area that has a certain impact on the safety of autonomous driving, so as to more efficiently enumerate scenarios and further improve test efficiency.

[0038] 3) The application generalizes the motion information of the responsible entity in the abnormal scene to obtain a generalized scene, and can further simulate and test the generalized scene, screen the generalized scene with the same abnormal type as the bad example data, and output the information of the generalized scene, thereby expanding the coverage of the abnormality.

[0039] Of course, implementing any product of the present application does not necessarily require all the advantages described above. BRIEF DESCRIPTION OF DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor on the basis of these drawings.

[0041] Figure 1 It is a system architecture diagram suitable for the embodiments of the present application;

[0042] Figure 2 It is a test method for an autonomous vehicle provided by the embodiments of the present application;

[0043] Figure 3 It is a schematic diagram of area division provided by the embodiments of the present application;

[0044] Figures 4a to 4c It is a schematic diagram of three scene simulation tests provided by the embodiments of the present application, respectively;

[0045] Figure 5 It is a schematic block diagram of a test device provided by the embodiments of the present application;

[0046] Figure 6 It is a schematic block diagram of an electronic device provided by the embodiments of the present application. DETAILED DESCRIPTION

[0047] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art belong to the scope of protection of the present application.

[0048] The terminology used in the description of the application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in the description of the application and the appended claims, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

[0049] It should be understood that the term "and / or" as used herein merely describes associated objects in association, which means that there can be three relationships, for example, A and / or B, which means that there are three cases of A alone, A and B together, and B alone. In addition, the character " / " herein generally represents an "or" relationship between the front and rear associated objects.

[0050] Depending on the context, the word "if" as used herein can be interpreted to mean "when" or "while" or "in response to determining" or "in response to detecting." Similarly, the phrase "if it is determined" or "if (a stated condition or event) is detected" can be interpreted to mean "when it is determined" or "in response to determining" or "when (a stated condition or event) is detected" or "in response to detecting (a stated condition or event)".

[0051] Many high-value bad examples are generated during real road testing (referred to as "road testing") of an autonomous vehicle. If a traditional method is used for manual analysis, labeling, etc., it will result in high labor costs, uncontrollable quality (relying on manual experience, which is difficult to control), low efficiency, and inability to scale. Therefore, how to efficiently and scalably process the bad examples of the autonomous vehicle road test is a problem that needs to be solved. The present application proposes a completely new idea.

[0052] In order to facilitate the understanding of the embodiments of the present application, first, the system architecture based on which the embodiments of the present application are based is simply described. Figure 1 An exemplary system architecture to which the embodiments of the present application can be applied is shown in FIG. 1. Figure 1 As shown in FIG. 1, the system mainly includes a data storage device and a testing device.

[0053] The data storage device is used to store bad example data of the tested autonomous vehicle. The bad example data is obtained during road testing of the tested autonomous vehicle, and refers to some scene data in which the autonomous vehicle has an abnormality. The bad example data can be stored in a database or in the form of a data file in the data storage device.

[0054] The data storage device can be any device with data storage function. For example, it can be a device for collecting and storing data during road testing of the autonomous vehicle, and for example, it can be a device for centrally storing data obtained during road testing of the autonomous vehicle.

[0055] Figure 1 The test device in the system is mainly used to test the autonomous vehicle by using the test method provided by the embodiments of the present application to obtain the responsible entity of the bad case. The responsible entity is the entity that causes the bad case. In the test process, the bad case is used for scene enumeration, and the enumerated scene is constructed and tested in a simulation environment.

[0056] The autonomous vehicle involved in the present application is a general description, which can be an unmanned vehicle or an assisted driving vehicle.

[0057] The autonomous vehicle mainly includes three main modules: a perception module, a decision module, and an execution module. The perception module perceives road information, real-time traffic information, obstacle information, and vehicle state information by using data collected by the perception sensor in the autonomous vehicle. The decision module determines the control information of the autonomous vehicle by using the information perceived by the perception module. The execution module is used to execute the control information determined by the decision module, thereby realizing the driving control of the autonomous vehicle. The perception sensor in the autonomous vehicle can include an image sensor, a radar, an infrared sensor, an ultrasonic sensor, etc. The image sensor can include a camera, a video camera, etc. The radar can include a laser radar, a millimeter wave radar, etc.

[0058] The test device can be any device with computing capability, such as a notebook computer, a PC (Personal Computer), etc. Even the test device can be set on the server side to complete the test of the perception fusion system of the autonomous vehicle.

[0059] It should be understood that Figure 1 The number of data storage devices and test devices in the system architecture is only illustrative. According to the needs of implementation, there can be any number of data storage devices and test devices.

[0060] Figure 2 The test method for the autonomous vehicle provided by the embodiments of the present application can be executed by the test device in the system architecture. Figure 1 As shown in the method, the method can include the following steps: Figure 2

[0061] Step 202: Obtain the bad case data of the tested autonomous vehicle.

[0062] Step 204: Enumerate scenes according to the bad case data, and the enumerated scenes include the tested autonomous vehicle or a combination of the tested autonomous vehicle and at least one obstacle included in the bad case data.

[0063] ​Step 206: Perform simulation test on each of the enumerated scenarios to determine an abnormal scenario of the same abnormal type as the badcase data.

[0064] Step 208: Determine the responsible entity of the badcase data from the tested autonomous vehicle and the obstacles contained in the badcase data by using the abnormal scenario.

[0065] As can be seen from the above process, the present application performs scene enumeration based on the badcase data, performs simulation test on each of the enumerated scenarios to determine an abnormal scenario of the same abnormal type as the badcase data, and then determines the responsible entity of the badcase data by using the abnormal scenario. This simulation test method based on scene enumeration can realize automatic analysis of the badcase data, without relying on manual experience for analysis and annotation, thereby reducing labor costs, improving efficiency, being easy to scale, and having higher accuracy.

[0066] The above method provided by the embodiments of the present application will be described in detail below. The step 202, i.e., "obtaining badcase data of the tested autonomous vehicle", will be described in detail below in combination with embodiments.

[0067] The badcase data in this step is generated in the process of road testing of the tested autonomous vehicle. That is, the tested autonomous vehicle is allowed to travel on a designated road to test whether the autonomous vehicle can travel normally in the actual road environment. If an abnormality occurs during the travel, the abnormal data is the badcase data. The badcase data contains relevant data of the tested autonomous vehicle and environmental data when the abnormality occurs. These badcase data can be collected by the autonomous vehicle itself, such as the data of the autonomous vehicle itself, road data and obstacle data collected by the autonomous vehicle when the abnormality occurs. It can also be collected by sensors installed on the road and traffic facilities, such as motion data of each vehicle, pedestrian, etc. collected by cameras installed on traffic light racks, speed sensors installed on the roadside, etc.

[0068] The abnormality of the autonomous vehicle during the road test corresponding to the badcase can include but is not limited to: sudden braking, collision, travel risk exceeding a preset risk level, etc.

[0069] For example, during the road test, the autonomous vehicle suddenly brakes, and usually the driving safety and user experience are considered to avoid sudden braking as much as possible. Therefore, if sudden braking occurs, it can be considered that the autonomous vehicle is abnormal, and the self-data and environmental data at this time constitute the badcase data. The deceleration of the autonomous vehicle is greater than or equal to a preset threshold value, which can be considered as sudden braking, and the threshold value can be set according to experience value, test value, etc.

[0070] For example, in the process of road testing, the autonomous vehicle may collide, which is usually prohibited in actual driving, so if a collision occurs, it can be considered that the autonomous vehicle has an abnormality, at which time the self-data and environment data are obtained, which constitute the bad example data of this time.

[0071] For example, in the process of road testing, the autonomous vehicle is usually equipped with a safety officer who sits in the autonomous vehicle during road testing to deal with some special situations. The safety officer can evaluate the driving risk according to the actual situation, and if the risk exceeds the preset risk level, the acquisition of bad example data can be triggered. Alternatively, after the road test is completed, the risk evaluation model can evaluate the driving data of the autonomous vehicle, and consider that a bad example has occurred if the risk evaluation exceeds the preset risk level.

[0072] These bad example data can be stored in the autonomous vehicle, uploaded to the server by the autonomous vehicle, or stored on other devices with data storage functions. Accordingly, the test device in the embodiments of the present application can obtain the bad example data of the tested autonomous vehicle from the tested autonomous vehicle, the server, or other devices with data storage functions.

[0073] The above step 204, i.e., "enumerating scenarios according to bad example data", will be described in detail below in conjunction with embodiments.

[0074] In the process of scenario enumeration, first, the obstacle list is obtained according to the bad example data. Since the bad example data is the data and environment data of the autonomous vehicle obtained when the abnormality occurs, the obstacle list is actually the obstacle list of the autonomous vehicle when the abnormality corresponding to the bad example data occurs, that is, it is determined which obstacles exist around the autonomous vehicle when the abnormality occurs. These obstacles may cause the abnormality to occur, but it is not clear which one or which ones caused the abnormality to occur, which needs to be investigated. In the embodiments of the present application, the way of investigation is not manual analysis, but the combination of the autonomous vehicle and the obstacles to enumerate different scenarios, and then simulate and test each scenario to analyze.

[0075] The obstacles involved in the embodiments of the present application can include vehicles, pedestrians, traffic facilities, and other objects that may have safety hazards to vehicles, such as trees and animals.

[0076] Because autonomous vehicles have strong data collection capabilities, the range of data they collect is typically very large. Performing full combination enumeration and simulation testing would result in a massive computational burden. Obstacles with relatively low safety impact on autonomous vehicles do not require combination enumeration and simulation testing. Therefore, as one feasible approach, entities within a first and / or second region can be identified based on bad case data. For example, obstacles within the first and / or second regions can form an obstacle list, including the autonomous vehicle under test and the obstacles. These entities are then combined to enumerate various scenarios, where each combination includes at least the autonomous vehicle under test. For example, the autonomous vehicle under test could be included as one scenario, and the autonomous vehicle under test could be combined with at least one obstacle from the obstacle list.

[0077] The first region mentioned above refers to the area within a preset first distance range from the position of the tested autonomous vehicle when the anomaly corresponding to the bad sample data occurs. In other words, the area adjacent to the autonomous vehicle can be used as the first region. For example, a preset radius value can be used, or the radius value can be determined based on the speed information of the autonomous vehicle; then, the area within the aforementioned radius value centered on the position of the autonomous vehicle is determined from the perception sensor data as the first region.

[0078] by Figure 3 For example, when an anomaly corresponding to bad case data occurs, a region within a radius centered on the location of the tested autonomous vehicle 'a' is selected as the first region. The radius can be a preset empirical or experimental value, or it can be determined based on the speed of the autonomous vehicle at that time. Obstacles within this first region then pose a significant safety risk to autonomous vehicle 'a'.

[0079] The second area is the driving area determined based on the navigation path of the autonomous vehicle being tested. Since autonomous vehicles have a defined driving path, usually following a navigation route, the second area can be understood as the area that may have an impact during the autonomous vehicle's driving process.

[0080] For example, the navigation path of an autonomous vehicle can be obtained, and the driving area can be defined as the lane where the autonomous vehicle is located and the area corresponding to the nearest N lanes on the navigation path, where N is a positive integer.

[0081] In determining the last N lanes, the lane where the autonomous vehicle is located, the driving intention at the time, etc. can be determined. For example, when the vehicle is straight and in the middle lane, the current lane and the left and right lanes on the navigation path corresponding to the second region can be determined. For another example, when the vehicle is turning right and in the rightmost lane, the current lane and the left lane on the navigation path corresponding to the second region can be determined. For another example, when the vehicle is straight and in the leftmost lane, the current lane and the right lane on the navigation path corresponding to the second region can be determined. For another example, when the vehicle is turning around and in the leftmost lane, the current lane, the left lane and the right lane on the navigation path corresponding to the second region can be determined.

[0082] Still taking the intersection in Figure 3 For example, the navigation path of the autonomous vehicle a is right turn at the intersection, and the current lane and the left lane of the autonomous vehicle a on the navigation path corresponding to the second region can be determined.

[0083] After obtaining the obstacle information in the first region and the second region, the obstacles form an obstacle list, and it is assumed that the obstacle list includes obstacles b-f. After scene enumeration, the following scenarios can be obtained:

[0084] Scenario 1: a

[0085] Scenarios 2-6: a+b, a+c, a+d, a+e, a+f

[0086] Scenarios 7-16: a+b+c, a+b+d, a+b+e, a+b+f, a+c+d, a+c+e, a+c+f, a+d+e, a+d+f, a+e+f

[0087] It should be noted that the above scene enumeration can be combined in order from few to many entities (tested autonomous vehicle and obstacles), and each combination includes at least the tested autonomous vehicle. The maximum number of entities in the scene can be set in advance, for example, set to 3, that is, the maximum number of entities in the scene is 3 when the scene is enumerated.

[0088] The above step 206, i.e. "performing simulation testing on the enumerated scenarios to determine the abnormal scenarios with the same abnormal type as the bad example data", will be described in detail below in conjunction with an embodiment.

[0089] The simulation test in this step refers to restoring the scene, that is, building a simulation environment to restore the position and motion state of the entity contained in the scene at the time when the abnormality occurred in the road test, to verify whether the same abnormality will still occur. The simulation test is respectively performed on each of the enumerated scenes, and if a scene appears the same type of abnormality as the bad example data, it is determined that the scene is an abnormal scene.

[0090] To improve the efficiency of the simulation test, the simulation test is sequentially performed on the enumerated scenes in order from few to many in terms of the number of entities contained in the scenes, and if the current simulation test scene appears the same type of abnormality as the bad example data, it is determined that the current simulation test scene is an abnormal scene, and the simulation test on the enumerated scenes is ended. Otherwise, the simulation test on the next scene is continued.

[0091] Taking the enumerated scenes in the above example as an example, the simulation test is first performed on scene 1, which contains only the tested autonomous vehicle a. As shown in FIG. 2A, a simulation environment is built to restore the position, motion state (such as speed, attitude, direction), etc. of the tested autonomous vehicle a at the time, to verify whether the same emergency braking (assuming that emergency braking occurred at this time during road testing) will occur. Figure 4a

[0092] The simulation test is then performed on scenes 2-6. Taking scene 2 as an example, scene 2 contains the tested autonomous vehicle a and the pedestrian b. As shown in FIG. 2B, a simulation environment is built to restore the position, motion state (such as speed, attitude, direction), etc. of the tested autonomous vehicle a and the pedestrian b at the time, to verify whether the same emergency braking (assuming that emergency braking occurred at this time during road testing) will occur. Figure 4b

[0093] If the emergency braking occurs in scene 2 during the simulation test of scenes 2-6, it is determined that scene 2 is an abnormal scene, and the simulation test on the enumerated scenes is stopped. Otherwise, the simulation test on the subsequent scenes is continued.

[0094] In the simulation test on scenes 7-16, scene 7 is taken as an example, which contains the tested autonomous vehicle a, the pedestrian b, and the vehicle c. As shown in FIG. 2C, a simulation environment is built to restore the position, motion state (such as speed, attitude, direction), etc. of the tested autonomous vehicle a, the pedestrian b, and the vehicle c at the time, to verify whether the same emergency braking (assuming that emergency braking occurred at this time during road testing) will occur. Figure 4c

[0095] The above step 208, i.e., "determining the responsible entity of the bad example data from the autonomous vehicle and the obstacle contained in the bad example data", is described in detail below in conjunction with an embodiment.

[0096] ​​​If the determined abnormal scenario only includes the tested autonomous vehicle, that is, the abnormality will also occur when the tested autonomous vehicle is restored to the scene at that time, it means that the tested autonomous vehicle itself is the problem, and it is taken as the responsible entity.

[0097] If the determined abnormal scenario includes the tested autonomous vehicle and at least one obstacle combination, the obstacle contained in the abnormal scenario is taken as the responsible entity. For example, scenario 2 is an abnormal scenario, scenario 2 contains a+b, because scenario 1 has been simulated and tested before and is not an abnormal scenario, so b is taken as the responsible entity. For another example, if scenario 7 is an abnormal scenario, scenario 7 contains a+b+c, scenarios 1, 2 and 3 have been simulated and tested before and are not abnormal scenarios, which means that b+c jointly caused the abnormality, so b+c is taken as the responsible entity.

[0098] After determining the responsible entity of the bad example data, in order to expand the coverage of the abnormality, the motion information (including position, motion state, etc.) of the responsible entity in the abnormal scenario can be further generalized to obtain a generalized scenario, and then the information of the generalized scenario is output.

[0099] In the generalization process, the motion information of the responsible entity can be generalized to more instances in the form of sampling, thereby forming new scenarios, which are called generalized scenarios. Assuming that scenario 2 is an abnormal scenario and b is the responsible entity, some new value combinations of the position, speed, direction, etc. of b can be sampled to obtain new instances as generalized scenarios.

[0100] The obtained generalized scenarios can be simulated and tested again to screen the generalized scenarios that appear the same abnormal type as the bad example data. For example, after generalizing the position, speed, direction, etc. of b, multiple generalized scenarios are obtained, and the generalized scenarios are simulated and tested, only the generalized scenarios that also appear the automatic driving vehicle a emergency braking are screened out, and the information of the generalized scenarios is output.

[0101] The information of the above-mentioned responsible entity and generalized scenario can be used to provide the R&D and operation personnel to improve the algorithm in the autonomous vehicle, and can also be used to form test cases to test the autonomous vehicle.

[0102] The above describes specific embodiments of the present specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different than the order in the embodiments and still achieve the desired result. In addition, the processes depicted in the figures do not necessarily require the particular order shown, or sequential order, to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous or necessary.

[0103]

[0103] According to another aspect, embodiments provide a testing device. Figure 5 A schematic block diagram of a testing device according to an embodiment is shown. As shown, the device 500 can include a bad case obtaining unit 501, a scenario enumeration unit 502, a simulation testing unit 503, and a responsibility inferring unit 504, and can further include a generalization processing unit 505. The main functions of the constituent units are as follows: Figure 5 The bad case obtaining unit 501 is configured to obtain bad case data of a tested autonomous vehicle.

[0104] The scenario enumeration unit 502 is configured to enumerate scenarios according to the bad case data, the enumerated scenarios including the tested autonomous vehicle, or including a combination of the tested autonomous vehicle and at least one obstacle contained in the bad case data.

[0105] The simulation testing unit 503 is configured to perform simulation testing on the enumerated scenarios respectively, and determine an abnormal scenario that has the same abnormal type as the bad case data.

[0106] The responsibility inferring unit 504 is configured to determine a responsible entity of the bad case data from the tested autonomous vehicle and the obstacle contained in the bad case data, by using the abnormal scenario.

[0107] As one of the implementable manners, the bad case obtaining unit 501 can obtain the above-mentioned bad case data from a data storage device. The data storage device can be the tested autonomous vehicle, a server, or other devices with data storage functions.

[0108] As one of the implementable manners, the scenario enumeration unit 502 can be specifically configured to: determine an obstacle in a first region and / or a second region according to the bad case data, the first region being a region within a preset first distance range from the position of the tested autonomous vehicle when the abnormality corresponding to the bad case data occurs, and the second region being a driving region determined according to the navigation path of the tested autonomous vehicle; take the tested autonomous vehicle as one of the scenarios, and combine the tested autonomous vehicle with the determined at least one obstacle respectively, to enumerate the scenarios.

[0109]

[0110] ​As one of the implementable manners, the simulation test unit 503 can be specifically configured to: sequentially perform simulation test on the enumerated scenarios in order from few to many in terms of the number of entities contained in the scenarios, wherein the entities include the tested autonomous vehicle or obstacles; if the scenario of the current simulation test has the same abnormal type as the bad example data, determine that the scenario of the current simulation test is an abnormal scenario, and end the simulation test on the enumerated scenarios; otherwise, continue to perform simulation test on the next scenario.

[0111] The simulation test unit 503 can build a simulation environment to restore the position and motion state of the entity contained in the scenario at the time when the abnormality occurs in the road test, to verify whether the same abnormality will still occur. The simulation test is performed on each of the enumerated scenarios, and if a scenario has the same type of abnormality as the bad example data, it means that the scenario is an abnormal scenario.

[0112] As one of the implementable manners, the responsibility presumption unit 504 can be specifically configured to: if the abnormal scenario only includes the tested autonomous vehicle, determine that the tested autonomous vehicle is the responsible entity of the bad example data; if the abnormal scenario includes a combination of the tested autonomous vehicle and at least one obstacle, the obstacle contained in the abnormal scenario is taken as the responsible entity.

[0113] Further, the generalization processing unit 505 is configured to perform generalization processing on the motion information of the responsible entity in the abnormal scenario to obtain a generalized scenario; and output information of the generalized scenario.

[0114] As one of the implementable manners, the generalization processing unit 505 can perform simulation test on the generalized scenario, and screen a generalized scenario having the same abnormal type as the bad example data; and the output information of the generalized scenario includes: output information of the screened generalized scenario.

[0115] The abnormal type corresponding to the bad example data includes: sudden braking, collision, or driving risk exceeding a preset risk level.

[0116] Each of the embodiments in the specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the difference from other embodiments. In particular, for the device embodiments, since they are basically similar to the method embodiments, they are described more simply, and the relevant parts can be referred to the part of the method embodiments. The device embodiments described above are only schematic, and the units described as separate components can or can not be physically separate, and the components displayed as units can or can not be physical units, i.e. they can be located in one place, or distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment. Those skilled in the art can understand and implement it without creative labor.

[0117] It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation portal for user to choose authorization or refusal.

[0118] In addition, the embodiment of the present application also provides a computer readable storage medium, which stores a computer program, and the program is executed by a processor to realize the steps of the method in any one of the preceding method embodiments.

[0119] An electronic device includes:

[0120] one or more processors; and

[0121] a memory associated with the one or more processors, the memory configured to store program instructions that, when executed by the one or more processors, perform the steps of the method of any one of the preceding method embodiments.

[0122] The present application also provides a computer program product, which includes a computer program, and the computer program realizes the steps of the method in any one of the preceding method embodiments when executed by a processor.

[0123] wherein, Figure 6An exemplary architecture of the electronic device can include a processor 610, a video display adapter 611, a disk drive 612, an input / output interface 613, a network interface 614, and a memory 620. The processor 610, the video display adapter 611, the disk drive 612, the input / output interface 613, the network interface 614, and the memory 620 can be communicatively connected through a communication bus 630.

[0124] The processor 610 can be implemented in the form of a general-purpose CPU, a microprocessor, an application specific integrated circuit (ASIC), or one or more integrated circuits, and is configured to execute related programs to implement the technical solutions provided in the present application.

[0125] The memory 620 can be implemented in the form of a ROM (Read Only Memory), a RAM (Random Access Memory), a static storage device, a dynamic storage device, or the like. The memory 620 can store an operating system 621 for controlling the operation of the electronic device 600, a basic input / output system (BIOS) 622 for controlling the low-level operation of the electronic device 600. In addition, a web browser 623, a data storage management system 624, and a test device 625, and the like can also be stored. The test device 625 can be an application program for implementing the foregoing steps in the embodiments of the present application. In summary, when the technical solutions provided in the present application are implemented by software or firmware, the related program codes are stored in the memory 620 and executed by the processor 610.

[0126] The input / output interface 613 is configured to connect an input / output module to realize information input and output. The input / output module can be configured as a component in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. The input device can include a keyboard, a mouse, a touch screen, a microphone, various sensors, and the like, and the output device can include a display, a speaker, a vibrator, an indicator, and the like.

[0127] The network interface 614 is configured to connect a communication module (not shown in the figure) to realize the communication interaction between the device and other devices. The communication module can realize communication through a wired manner (such as USB, network cable, etc.) or a wireless manner (such as mobile network, WIFI, Bluetooth, etc.).

[0128] Bus 630 includes a path for transferring information between the various components (e.g., processor 610, video display adapter 611, disk drive 612, input / output interface 613, network interface 614, and memory 620).

[0129] It should be noted that although the above device only shows the processor 610, video display adapter 611, disk drive 612, input / output interface 613, network interface 614, memory 620, bus 630, etc., but in the process of implementation, the device can also include other components necessary for normal operation. In addition, those skilled in the art can understand that the above device can also only contain the components necessary to implement the scheme of the present application, and does not have to contain all the components shown in the figure.

[0130] From the above description of the embodiments, those skilled in the art can clearly understand that the present application can be implemented by means of software and the necessary general hardware platform. Based on such understanding, the technical solutions of the present application can be embodied in the form of a computer program product, which can be stored in a storage medium such as ROM / RAM, disk, optical disk, etc., including a number of instructions for making a computer device (which can be a personal computer, server, or network device, etc.) execute the methods described in various embodiments or some parts of the embodiments of the present application.

[0131] The above provides a detailed introduction to the technical solutions of the present application. The specific examples are applied to the principle and implementation of the present application. The above description of the embodiments is only used to help understand the method and its core idea of the present application; at the same time, for those skilled in the art, according to the idea of the present application, the specific implementation and application range will be changed. In conclusion, the content of the specification should not be understood as a limitation of the present application.

Claims

1. A test method for an autonomous vehicle, characterized by, The method comprises: obtaining bad example data of a tested autonomous vehicle; the bad example data comprises relevant data of the tested autonomous vehicle and environmental data when an abnormality occurs in a test process; determining a plurality of entities in a first region and / or a second region according to the bad example data, the entities comprising the tested autonomous vehicle and an obstacle, the first region being a region within a preset first distance range from a position of the tested autonomous vehicle when an abnormality corresponding to the bad example data occurs, the second region being a driving region determined according to a navigation path of the tested autonomous vehicle; combining the determined plurality of entities according to preset maximum number of entities contained in a single scene to enumerate a plurality of scenes, the enumerated scenes comprising the tested autonomous vehicle or a combination of the tested autonomous vehicle and at least one obstacle contained in the bad example data; sequentially performing simulation tests on the enumerated scenes in order of a number of entities contained in the scenes from few to many to determine an abnormal scene of the same abnormal type as the bad example data; wherein, when performing simulation test on a current simulation test scene, a simulation environment is built to restore only the entities and their positions and motion states contained in the current simulation test scene to test whether an abnormality of the same type as the bad example data occurs in the current simulation test scene, and if so, the current simulation test scene is the abnormal scene; determining a responsible entity of the bad example data according to the entities contained in the abnormal scene and the entities contained in a scene that is confirmed as a non-abnormal scene through simulation test before the abnormal scene.

2. The method of claim 1, wherein, The sequentially performing simulation tests on the enumerated scenes comprises: sequentially performing simulation tests on the enumerated scenes, wherein the entities comprise the tested autonomous vehicle or the obstacle; if the current simulation test scene appears to be of the same abnormal type as the bad example data, determining that the current simulation test scene is the abnormal scene and ending the simulation tests on the enumerated scenes; otherwise, continuing the simulation test on a next scene.

3. The method of claim 1, wherein, Determining the responsible entity of the bad example data from the autonomous vehicle and the obstacle contained in the bad example data using the abnormal scene comprises: if the abnormal scene only comprises the tested autonomous vehicle, determining that the tested autonomous vehicle is the responsible entity of the bad example data; if the abnormal scene comprises a combination of the tested autonomous vehicle and at least one obstacle, determining the obstacle contained in the abnormal scene as the responsible entity.

4. The method of claim 1, wherein, The method further comprises: performing generalization processing on motion information of the responsible entity in the abnormal scene to obtain a generalized scene; outputting information of the generalized scene.

5. The method of claim 4, wherein, Before outputting the information of the generalized scene, the method further comprises: performing simulation test on the generalized scene to screen a generalized scene of the same abnormal type as the bad example data; outputting the information of the generalized scene comprises outputting information of the screened generalized scene.

6. The method according to any one of claims 1 to 5, characterized in that, The abnormal type corresponding to the bad example data includes sudden braking, collision, or driving risk exceeding a preset risk level.

7. A testing device for an autonomous vehicle, characterized by The device comprises: a bad example acquisition unit configured to acquire bad example data of a tested autonomous vehicle; the bad example data includes relevant data of the tested autonomous vehicle when an abnormality occurs during testing and environmental data; a scene enumeration unit configured to determine a plurality of entities in a first region and / or a second region according to the bad example data, the entities including the tested autonomous vehicle and obstacles, the first region being a region within a preset first distance range from the position of the tested autonomous vehicle when the abnormality corresponding to the bad example data occurs, and the second region being a driving region determined according to the navigation path of the tested autonomous vehicle; the plurality of entities are combined according to preset maximum number of entities included in a single scene to enumerate a plurality of scenes, the enumerated scenes including the tested autonomous vehicle or a combination of the tested autonomous vehicle and at least one obstacle included in the bad example data; a simulation test unit configured to sequentially simulate and test the enumerated scenes in order of the number of entities included in the scenes from few to many to determine an abnormal scene of the same abnormal type as the bad example data; when simulating and testing a current simulation test scene, a simulation environment is built to restore only the entities and their positions and motion states included in the current simulation test scene to test whether an abnormality of the same type as the bad example data occurs in the current simulation test scene, and if so, the current simulation test scene is the abnormal scene; a responsibility presumption unit configured to determine a responsible entity of the bad example data according to the entities included in the abnormal scene and the entities included in a scene that is simulated and tested before the abnormal scene and confirmed to be a non-abnormal scene.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to implement the steps of the method of any one of claims 1 to 6.

9. An electronic device, comprising: comprises: one or more processors; and a memory associated with the one or more processors, the memory being configured to store program instructions that, when executed by the one or more processors, perform the steps of the method of any one of claims 1 to 6.

Images