Reliability determination method, device and equipment of automatic driving system, and medium

Abstract

Embodiments of the present application disclose a reliability determination method and device for an automatic driving system, an equipment and a medium. The method comprises the following steps: constructing a vehicle model, a sensor model, a control model of the automatic driving system and a scene model, obtaining a scene simulation result of the automatic driving system, constructing a response surface of a scene variable and the scene simulation result based on the scene simulation result and each scene variable corresponding to the scene model, and determining a failure probability corresponding to the automatic driving system based on the response surface, a probability density function of each scene variable, a range of each scene variable and a preset failure condition. The method realizes quantification of a reliability result of the automatic driving system, provides strong support for development and improvement of the automatic driving system, and solves the technical problem that all test scenes cannot be enumerated in the prior art.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of autonomous driving technology, and in particular to a method, device, equipment, and medium for determining the reliability of an autonomous driving system. Background Art

[0002] With the development of intelligent vehicles, autonomous driving is an inevitable trend. Autonomous vehicles must undergo rigorous functional safety testing before they can be put on the road. Autonomous driving simulation testing uses mathematical modeling to digitally recreate autonomous driving application scenarios, creating system models that are as close to the real world as possible. This allows for the verification of autonomous driving systems and algorithms through software simulation testing without the need for a real vehicle. Simulation testing offers advantages such as high scenario coverage, a safe testing process, and high testing efficiency.

[0003] However, simulation testing is difficult to exhaustively test all scenarios, and existing technologies are unable to quantify and evaluate the reliability of autonomous driving systems. Summary of the Invention

[0004] Embodiments of the present invention provide a reliability determination method, apparatus, device, and medium for an autonomous driving system to quantify the reliability results of the autonomous driving system and predict the failure probability of the autonomous driving system in all test scenarios under a certain type of scenario.

[0005] In a first aspect, an embodiment of the present invention provides a method for determining the reliability of an autonomous driving system, the method comprising:

[0006] Based on the pre-built vehicle model, sensor model, control model of the autonomous driving system, and scenario model, obtaining a scenario simulation result corresponding to the autonomous driving system;

[0007] Determining a response surface between the scenario variables and the scenario simulation results based on the scenario simulation results and the scenario variables corresponding to the scenario model;

[0008] The failure probability corresponding to the autonomous driving system is determined based on the probability density function corresponding to each of the scenario variables, the range of each of the scenario variables, the response surface, and the preset failure condition.

[0009] Optionally, the method further includes:

[0010] Acquire preset scene variables, wherein the scene variables include the type of the scene variables and the range of the scene variables;

[0011] Get pre-built simulation scenario templates;

[0012] Based on the type of the scene variable and the range of the scene variable, a first sampling process is performed on each of the scene variables, and a scene model is determined based on a result of the first sampling process and the simulation scene template.

[0013] Optionally, the method further includes:

[0014] Building scene static elements, wherein the scene static elements include at least one of road information, lane information, and environmental information;

[0015] Building scene dynamic elements, wherein the scene dynamic elements include at least one of traffic characteristic information, own vehicle information, target vehicle information, and other traffic participant information;

[0016] Acquire preset environmental conditions, simulation duration, simulation trigger conditions and simulation termination conditions, and establish a simulation scene template based on the scene static elements, the scene dynamic elements, the preset environmental conditions, the simulation duration, the simulation trigger conditions and the simulation termination conditions.

[0017] Optionally, obtaining a scenario simulation result corresponding to the autonomous driving system based on a pre-built vehicle model, sensor model, control model of the autonomous driving system, and scenario model includes:

[0018] Acquire, based on the control model, vehicle motion information at a current moment sent by the vehicle model and target information at a current moment sent by the sensor model;

[0019] Determining, by the control model, motion control information at a next moment based on the vehicle motion information at the current moment, the target information at the current moment, and the scene model, and sending the motion control information to the vehicle model;

[0020] Based on the vehicle motion information at each moment determined by the vehicle model, the motion control information at each moment determined by the control model, and the target information at each moment determined by the sensor model, a scene simulation result corresponding to the automatic driving system is obtained.

[0021] Optionally, determining the failure probability corresponding to the autonomous driving system based on the probability density function corresponding to each of the scenario variables, the range of each of the scenario variables, the response surface, and a preset failure condition includes:

[0022] Determining a sampling output result based on a probability density function corresponding to each of the scenario variables, a range of each of the scenario variables, and the response surface;

[0023] The failure probability corresponding to the automatic driving system is determined based on the sampling output result, the probability density function and the preset failure condition.

[0024] Optionally, determining the sampling output result based on the probability density function corresponding to each of the scenario variables, the range of each of the scenario variables, and the response surface includes:

[0025] Obtaining a probability density function corresponding to each of the scenario variables;

[0026] Based on the probability density function and the range of the scenario variables, performing a second sampling process on each of the scenario variables to obtain a sampling variable result of each of the scenario variables;

[0027] A sampling output result is determined based on the sampling variable results and the response surface.

[0028] Optionally, the method further includes:

[0029] Based on the scenario simulation results, calculating sensitivity information corresponding to each scenario variable;

[0030] Based on the sensitivity information corresponding to each of the scene variables, scene variables that do not meet the preset sensitivity requirements are eliminated from the scene variables.

[0031] In a second aspect, an embodiment of the present invention further provides a device for determining reliability of an autonomous driving system, the device comprising:

[0032] A simulation module, configured to obtain a scenario simulation result corresponding to the autonomous driving system based on a pre-built vehicle model, sensor model, control model of the autonomous driving system, and scenario model;

[0033] a response surface construction module, configured to determine a response surface between the scenario variables and the scenario simulation results based on the scenario simulation results and the scenario variables corresponding to the scenario model;

[0034] A failure probability determination module is used to determine the failure probability corresponding to the autonomous driving system based on the probability density function corresponding to each of the scenario variables, the range of each of the scenario variables, the response surface, and a preset failure condition.

[0035] In a third aspect, an embodiment of the present invention further provides an electronic device, comprising:

[0036] one or more processors;

[0037] a storage device for storing one or more programs,

[0038] When the one or more programs are executed by the one or more processors, the one or more processors implement the reliability determination method of the autonomous driving system provided by any embodiment of the present invention.

[0039] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, a reliability determination method of an autonomous driving system as provided in any embodiment of the present invention is implemented.

[0040] The embodiments of the above invention have the following advantages or beneficial effects:

[0041] By using a pre-built vehicle model, sensor model, control model of the autonomous driving system, and scenario model, the scenario simulation results of the autonomous driving system are obtained. Based on the scenario simulation results and the scenario variables corresponding to the scenario model, a response surface between the scenario variables and the scenario simulation results is constructed. Further, based on the response surface, the probability density function of each scenario variable, the range of each scenario variable, and preset failure conditions, the corresponding failure probability of the autonomous driving system is determined. This method quantifies the reliability results of the autonomous driving system by determining the failure probability, providing strong support for the development and improvement of the autonomous driving system. Moreover, this method predicts the failure probability of the autonomous driving system in all test scenarios under a certain type of scenario through the probability density function of the scenario variables, solving the technical problem that the existing technology cannot exhaustively enumerate all test scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings introduced here only illustrate some of the embodiments to be described by the present invention, and are not exhaustive. A person skilled in the art can derive other drawings based on these drawings without inventive effort.

[0043] Figure 1A A schematic flow chart of a method for determining the reliability of an autonomous driving system provided in the first embodiment of the present invention;

[0044] Figure 1B A schematic diagram of a response surface provided in Example 1 of the present invention;

[0045] Figure 2 This is a flow chart of a method for determining the reliability of an autonomous driving system provided in the second embodiment of the present invention;

[0046] Figure 3 This is a flow chart of a method for determining the reliability of an autonomous driving system provided in a third embodiment of the present invention;

[0047] Figure 4 This is a flow chart of a method for determining the reliability of an autonomous driving system provided in a fourth embodiment of the present invention;

[0048] Figure 5A structural schematic diagram of a reliability determination device of an automatic driving system provided in Embodiment Five of the present application;

[0049] Figure 6 A structural schematic diagram of an electronic device provided in Embodiment Six of the present application. DETAILED DESCRIPTION

[0050] The present application will be further described below in conjunction with the accompanying drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the present application, but not to limit the present application. In addition, it should be noted that, for the convenience of description, only the parts related to the present application are shown in the drawings, but not all the structures.

[0051] Embodiment One

[0052] Figure 1A A flowchart of a reliability determination method of an automatic driving system provided in Embodiment One of the present application. The present embodiment can be applicable to the case of analyzing the reliability of an automatic driving system, and is especially applicable to the case of calculating the failure probability of an automatic driving system according to a pre-constructed vehicle model, a sensor model, a control model of the automatic driving system, and a scene model. The method can be executed by a reliability determination device of the automatic driving system, which can be realized by hardware and / or software. The method specifically includes the following steps:

[0053] S110, obtaining the scene simulation result corresponding to the automatic driving system based on the pre-constructed vehicle model, sensor model, control model of the automatic driving system, and scene model.

[0054] The vehicle model can be a dynamics model simulating the motion characteristics of a vehicle in actual operation. The vehicle model can be constructed according to pre-set vehicle body parameters, aerodynamic parameters, and transmission parameters. For example, the parameters can include the overall height of the vehicle, the overall width of the vehicle, the windward area, the air density, the distance from the mass center of the sprung mass to the front axle, the height from the mass center of the sprung mass to the inside, the wheelbase of the vehicle, etc.

[0055] The sensor model can be built based on the sensor information actually installed on the vehicle. For example, the sensor model can include a laser radar model, a camera model, and a millimeter wave radar model, etc. In the construction process of the sensor model, the type of the sensor model can be selected, such as the physical level, the signal level, and the true value level, and the installation position of the sensor, the configuration of the basic parameters, etc. can be selected. Optionally, the vehicle model and the sensor model can be pre-constructed in an automatic driving simulation software.

[0056] In this embodiment, the autonomous driving system may be, for example, a control system such as an Advanced / Automatic Emergency Braking (AEB), a hill-start assist system, an anti-lock braking system, a traffic jam following system, a highway designated driver system, or an automatic parking system. Specifically, the control model of the autonomous driving system may be a model used to control the vehicle according to the autonomous driving algorithm in the autonomous driving system. The control model of the autonomous driving system can be integrated with other models in the Simulink environment.

[0057] The scenario model can be a model constructed based on the various simulation scenarios to be tested for the autonomous driving system. The scenario model can be pre-constructed in the autonomous driving simulation software. For example, taking the autonomous driving system as an automatic emergency braking system, the scenario model can be a car-to-car rear moving (CCRm) model for a vehicle rear-end collision test with a constant speed in front of a straight road, a car-to-car rear braking (CCRb) model for a vehicle rear-end collision test with a decelerated speed in front of a straight road, etc. The scenario model can be constructed based on different values ​​of each scenario variable, which includes multiple simulation test scenarios. For example, for the CCRm model, the scenario variables can be variables such as the vehicle speed, the target vehicle speed, and the bias rate. By assigning multiple values ​​to variables such as the vehicle speed, the target vehicle speed, and the bias rate, a scenario model including multiple simulation test scenarios can be obtained.

[0058] Specifically, after establishing the vehicle model, sensor model, autonomous driving system control model, and scenario model, the models can interact with each other to simulate the test scenario and obtain scenario simulation results. For example, the vehicle model can input information such as the vehicle's speed, acceleration, steering wheel angle, throttle, and brake pedal opening into the control model. The sensor model can input information such as the relative distance, relative speed, and relative angle between the target vehicle or target object and the vehicle into the control model. The scenario model can input relevant information about the target vehicle or target object into the sensor model. The vehicle model can input the vehicle's position information into the scenario model. The control model can send the vehicle's braking information to the vehicle model. The data output by each model can be used to extract the scenario simulation results of the autonomous driving system.

[0059] Among them, the scenario simulation results can include various data related to the test scenario. The data contained in the scenario simulation results under different test scenarios (scenario models) are different. For example, for the CCRm model, the scenario simulation results can include data such as time to collision (TTC, which can be calculated by the distance between the two vehicles and the relative speed of the two vehicles), whether a collision occurs, and the minimum relative distance between the two vehicles.

[0060] Exemplarily, the method of obtaining a scene simulation result corresponding to the autonomous driving system based on a pre-built vehicle model, sensor model, control model of the autonomous driving system, and scene model includes: obtaining the vehicle motion information at the current moment sent by the vehicle model and the target information at the current moment sent by the sensor model based on the control model; determining, through the control model, the motion control information at the next moment to be sent to the vehicle model based on the vehicle motion information at the current moment, the target information at the current moment, and the scene model; obtaining the scene simulation result corresponding to the autonomous driving system based on the vehicle motion information at each moment determined by the vehicle model, the motion control information at each moment determined by the control model, and the target information at each moment determined by the sensor model.

[0061] Target information can include the target vehicle or object's position, speed, acceleration, relative distance from the vehicle, relative speed, and relative angle. Vehicle motion information can include the vehicle's speed, acceleration, steering wheel angle, and throttle and brake pedal openings. Specifically, the control model can use the autonomous driving algorithm to calculate the vehicle motion information for the next moment based on the current moment's vehicle motion information, target information, and scenario model. This information is then sent to the vehicle model as braking information, causing the vehicle model to move according to this braking information at the next moment. The control model can be connected to the vehicle model, sensor model, and scenario model to form a closed loop. Based on the information output by each model at each moment, the scenario simulation results of the autonomous driving system can be obtained.

[0062] S120 : Determine a response surface between the scenario variables and the scenario simulation results based on the scenario simulation results and each scenario variable corresponding to the scenario model.

[0063] The scenario variables corresponding to the scenario model may be variables required to construct the scenario model. For example, for the CCRm model, the scenario variables may be the vehicle's speed, acceleration, initial position, direction, final position, target vehicle speed, acceleration, initial position, direction, final position, and other traffic participant states.

[0064] Specifically, this embodiment can construct a response surface between scenario variables and scenario simulation results output by the simulation system using mathematical optimization analysis software. A response surface can be constructed using any two scenario variables and one scenario simulation result. To a certain extent, the scenario variables are equivalent to input X, and the scenario simulation results output by the simulation system are equivalent to Y. The response surface can be a function that describes the XY relationship in three-dimensional space.

[0065] For example,Figure 1B Figure 1 shows a schematic diagram of a response surface. Taking the CCRm model as an example, the scenario variables are the target vehicle speed and the host vehicle speed, and the scenario simulation result is the minimum relative distance between the two vehicles. This means that the target vehicle speed and the host vehicle speed are used as inputs, and the minimum relative distance between the two vehicles is used as the output to construct a response surface. It should be noted that in this embodiment, the number of response surfaces can be one or more.

[0066] In one embodiment, considering that there may be multiple scenario variables, some have a significant impact on the scenario simulation results, while others have a minimal impact. For example, the bias rate has a very small impact on the scenario simulation results. Scenario variables with no or minimal impact can be ignored in subsequent response surface construction and failure probability analysis to simplify calculations and improve the efficiency of reliability analysis of the autonomous driving system. Specifically, the method further includes: calculating sensitivity information corresponding to each scenario variable based on the scenario simulation results; and, based on the sensitivity information corresponding to each scenario variable, eliminating scenario variables from each scenario variable that do not meet preset sensitivity requirements.

[0067] Among them, the sensitivity information can describe the impact of the uncertainty of the scene variables in the autonomous driving system on the scene simulation results. Exemplarily, the calculation of the sensitivity information corresponding to each scene variable can be calculated based on the variance method. In this way, the sensitivity of each scene variable can be analyzed, so as to determine the importance of each scene variable to the scene simulation results and quantify it, so that only the scene variables that meet the preset sensitivity requirements can be used for subsequent response surface construction or failure probability analysis, which simplifies the calculation and improves the efficiency of failure probability analysis. It should be noted that the step of calculating the sensitivity information corresponding to each of the scene variables can be performed before constructing the response surface, so as to construct the response surface for important scene variables, reduce the amount of calculation, and improve the efficiency of failure probability analysis.

[0068] S130. Determine a failure probability corresponding to the autonomous driving system based on a probability density function corresponding to each of the scenario variables, a range of each of the scenario variables, the response surface, and a preset failure condition.

[0069] The probability density function of the scenario variable can be derived based on big data analysis or IoV data extraction and analysis. For example, based on big data analysis, it can be determined that vehicle speeds on a road section generally follow a normal distribution. Therefore, the probability density function of vehicle speed can be a normal distribution function. The range of the scenario variable can be the value range of the scenario variable, for example, 0-100. It should be noted that scenario variables can have various types, such as continuous, discrete, constant, or function, and different types of scenario variables can have different value ranges.

[0070] The preset failure condition may be a pre-set failure condition of the autonomous driving system. For example, for a CCRm scenario of an AEB system, the preset failure condition may be a collision between two vehicles, that is, a minimum relative distance between the two vehicles is less than 0.

[0071] Specifically, the failure probability of the autonomous driving system under a given scenario model can be calculated using probability density functions (PDFs) of each scenario variable, its range, constructed response surfaces, and pre-set failure conditions within mathematical optimization analysis software using probabilistic methods. For example, the values ​​of the scenario variables can be sampled within their range, and the simulation results corresponding to each sampled value of the scenario variable can be determined based on the response surface. Furthermore, the failure probability of the autonomous driving system can be calculated based on the simulation results corresponding to each sampled value, the pre-set failure conditions, and the probability density function.

[0072] In this embodiment, a co-simulation can be performed using autonomous driving simulation software and mathematical optimization analysis software. The autonomous driving system's control model, vehicle model, sensor model, and scenario model are integrated within the autonomous driving simulation software to establish a simulation scenario for testing the autonomous driving control algorithm within the system. The mathematical optimization analysis software then constructs a response surface, defines the probability density function of the scenario variables, and calculates the failure probability. This co-simulation method, based on autonomous driving simulation software and mathematical optimization analysis software, enables automated testing and accelerates testing via the cloud, improving efficiency and reducing costs. Furthermore, this method offers high coverage and safety. Specifically, test scenarios can be flexibly configured, allowing for simulation testing and subsequent computational analysis of extreme and dangerous operating conditions, effectively covering blind spots in autonomous driving function testing. Furthermore, this method can quantify analysis results, employing a probability-based approach to analyze the reliability of the autonomous driving system. The reliability analysis results are interpretable and quantifiable, providing strong support for the development and improvement of autonomous driving algorithms.

[0073] The technical scheme of the embodiment, by means of the vehicle model, the sensor model, the control model of the automatic driving system and the scene model constructed in advance, the scene simulation result of the automatic driving system is obtained, the response surface of the scene variable and the scene simulation result is constructed based on the scene simulation result and each scene variable corresponding to the scene model, and further based on the response surface, the probability density function of each scene variable, the range of each scene variable and the preset failure condition, the failure probability corresponding to the automatic driving system is determined, the method realizes the quantification of the reliability result of the automatic driving system by determining the failure probability, and provides strong support for the development and improvement of the automatic driving system; and the method realizes the prediction of the failure probability of the automatic driving system under all test scenes in a certain type of scene through the probability density function of the scene variable, and solves the technical problem that all test scenes cannot be exhausted in the prior art.

[0074] Embodiment two

[0075] Figure 2 The flowchart of the reliability determination method of the automatic driving system provided by the second embodiment of the application, based on the above-mentioned embodiments, the method further comprises: obtaining the preset scene variables, wherein the scene variables include the type of the scene variables and the range of the scene variables; obtaining the simulation scene template constructed in advance; based on the type of the scene variables and the range of the scene variables, performing first sampling processing on each scene variable, and determining the scene model based on the result of the first sampling processing and the simulation scene template. The explanations of the same or corresponding terms as in the above-mentioned embodiments are not repeated here. See Figure 2 The reliability determination method of the automatic driving system provided by the second embodiment comprises the following steps:

[0076] S210, obtaining the preset scene variables, wherein the scene variables include the type of the scene variables and the range of the scene variables.

[0077] The preset scene variables can be defined according to the type of the scene to be tested. For example, taking the AEB system as an example, if the scene to be tested is CCRm, the scene variables can be defined to include the vehicle speed, the target vehicle speed, the offset rate and the like, wherein the offset rate is the proportion of the overlapping part of the host vehicle and the target vehicle to the host vehicle, the reference line of the overlap definition is the center line of the host vehicle, and in the case of 100% overlap, the center lines of the host vehicle and the target vehicle are aligned.

[0078] Of course, when defining each scenario variable, you can also define the type and range of the scenario variable. The type of the scenario variable can be continuous, discrete, constant, or function, and the range of the scenario variable can be a range of values. For example, the speed of the host vehicle or the target vehicle can be a continuous variable, while the type of the road attribute (such as the road adhesion coefficient) can be a discrete variable.

[0079] S220: Obtain a pre-built simulation scene template, perform a first sampling process on each of the scene variables based on the type of the scene variable and the range of the scene variable, and determine a scene model based on the result of the first sampling process and the simulation scene template.

[0080] Specifically, this embodiment can perform a first sampling process on the scene variable according to the type and range of the scene variable to obtain different values ​​of the scene variable. In this embodiment, random sampling can be performed according to the type and range of the scene variable. The sampling strategy and number can be determined according to the scene model. Generally, a simple random sampling strategy can be adopted, and the sampling number is 10. 2 ~10 3 The quantity can meet the needs.

[0081] In this way, multiple values ​​of each scenario variable are obtained as the result of the first sampling process. Furthermore, the result of the first sampling process, that is, the multiple values ​​of each scenario variable, is substituted into the pre-built simulation scenario template to obtain each simulation test scenario in the scenario model. Specifically, after each scenario variable takes a value, the values ​​of each scenario variable together constitute a sample. Substituting the sample into the simulation scenario template is equivalent to establishing a simulation test scenario; each simulation test scenario constitutes a scenario model. For example, a1, b1, and c1 constitute a sample; a2, b2, and c2 constitute a sample, and so on.

[0082] Exemplarily, the simulation scene template can be constructed in the following manner, that is, optionally, the method also includes: building scene static elements, wherein the scene static elements include at least one of road information, lane information, and environmental information; building scene dynamic elements, wherein the scene dynamic elements include at least one of traffic characteristic information, vehicle information, target vehicle information, and other traffic participant information; obtaining preset environmental conditions, simulation duration, simulation trigger conditions, and simulation termination conditions, and establishing a simulation scene template based on the scene static elements, the scene dynamic elements, the preset environmental conditions, the simulation duration, the simulation trigger conditions, and the simulation termination conditions.

[0083] That is, the construction of a simulation scene template includes the construction of static scene elements, the construction of dynamic scene elements, the definition of preset environmental conditions, and the setting of simulation duration, simulation trigger conditions, and simulation termination conditions. The static scene elements may include at least one of road information, lane information, and environmental information. Road information may include road geometry (such as road length and width), road surface material, and the number of lanes. Lane information may include lane length, lane width, and lane markings. Environmental information may include traffic lights, road markings, traffic signs, roadblocks, fences, and road structures (such as bridges and tunnels). Traffic characteristic information may include information such as vehicle density, vehicle speed, and the distribution of people and vehicles. Vehicle information may include information such as the vehicle's geometric model, vehicle motion parameters, and vehicle motion route. Target vehicle information may include information such as the target vehicle's geometric model, target vehicle motion parameters, and target vehicle motion route. Other traffic participant information may include motion information and geometric model information of pedestrians, animals, and other participating objects. Preset environmental conditions may be setting conditions for simulation environment elements, including but not limited to settings for data such as weather type and light visibility. The simulation trigger condition can be a condition that triggers the start of a simulation test, such as the vehicle's braking. The simulation termination condition can be a condition that ends the simulation test, such as when the vehicle's speed drops to zero. Of course, to ensure sufficient scenario simulation results are collected, the simulation termination condition can be set to the vehicle's speed dropping to zero and then a set delay time. This method achieves the construction of a simulation scenario template, and each simulation test scenario in the scenario model can be derived based on this simulation scenario template.

[0084] S230. Based on the pre-built vehicle model, sensor model, control model of the autonomous driving system, and scenario model, obtain a scenario simulation result corresponding to the autonomous driving system.

[0085] S240 : Determine a response surface between the scenario variables and the scenario simulation results based on the scenario simulation results and the scenario variables corresponding to the scenario model.

[0086] S250. Determine a failure probability corresponding to the autonomous driving system based on a probability density function corresponding to each of the scenario variables, a range of each of the scenario variables, the response surface, and a preset failure condition.

[0087] The technical scheme of the embodiment defines various scene variables, constructs a simulation scene template, and performs first sampling processing on the various scene variables according to the types of the scene variables and the ranges of the scene variables, and then determines a scene model according to the first ozone processing result and the simulation scene template, thereby realizing the establishment of various simulation test scenes. Moreover, through flexible configuration of the simulation test scenes, simulation test and subsequent calculation and analysis processes of extreme and dangerous working conditions can be realized, and the automatic driving function test blind area can be effectively covered.

[0088] Embodiment three

[0089] Figure 3 A flowchart of a reliability determination method of an automatic driving system provided by the third embodiment of the present application, based on the above-mentioned embodiments, the determination of the failure probability corresponding to the automatic driving system based on the probability density functions corresponding to the various scene variables, the ranges of the various scene variables, the response surface, and the preset failure condition, comprises: determining the sampling output results of the various scene variables based on the probability density functions corresponding to the various scene variables, the ranges of the various scene variables, and the response surface; and determining the failure probability corresponding to the automatic driving system based on the sampling output results, the probability density functions, and the preset failure condition. The explanations of the same or corresponding terms in the above-mentioned embodiments are not repeated here. See Figure 3 The reliability determination method of the automatic driving system provided by the present embodiment comprises the following steps:

[0090] S310, based on the pre-constructed vehicle model, sensor model, control model of the automatic driving system, and scene model, obtaining the scene simulation result corresponding to the automatic driving system.

[0091] S320, based on the scene simulation result and the various scene variables corresponding to the scene model, determining the response surface of the scene variables and the scene simulation result.

[0092] S330, determining the sampling output result based on the probability density functions corresponding to the various scene variables, the ranges of the various scene variables, and the response surface.

[0093] The sampling output result can be the scene simulation result corresponding to the sampling value of the scene variable in the response surface. For example, the determination of the sampling output result based on the probability density functions corresponding to the various scene variables, the ranges of the various scene variables, and the response surface can be: obtaining the probability density functions corresponding to the various scene variables; performing second sampling processing on the various scene variables based on the probability density functions and the ranges of the scene variables to obtain the sampling variable results of the various scene variables; and determining the sampling output result based on the sampling variable results and the response surface.

[0094] Specifically, a second sampling process can be performed on the scenario variables to obtain different values ​​for each scenario variable. Furthermore, the scenario simulation results corresponding to the sampled variable results can be quickly obtained through the response surface, i.e., the sampling output results. It should be noted that the sampling output results can include the simulated scenario simulation results and the results of the fitting when constructing the response surface.

[0095] Optionally, a second sampling process can be performed on the scene variables based on preset constraints. These constraints can be pre-set conditions that constrain the values ​​of the sampled scene variables; for example, the vehicle's speed is greater than the speed of the preceding vehicle. Constraints can avoid performing reliability analysis on values ​​of scene variables with less influence, thereby reducing the computational complexity of the reliability analysis and further improving analysis efficiency.

[0096] Among them, a suitable second sampling processing method can be selected according to the number of scenario variables, the form of failure conditions and the number of samples. The second sampling processing method includes but is not limited to Monte Carlo sampling, Latin hypercube sampling, importance sampling and adaptive sampling.

[0097] For example, the Latin hypercube sampling method can be used to perform a second sampling process on each scenario variable. The Latin hypercube sampling method uses the principle of stratification to randomly sample in the design space, which can not only ensure that the sampling points are not clustered and have good spatial coverage, but also ensure high sampling efficiency. The process of using the Latin hypercube sampling method to perform a second sampling process on each scenario variable can be: (1) according to the N sample points to be sampled, the space of each scenario variable is divided into N parts; (2) a random sampling with equal probability is performed in each subspace domain of the scenario variable to obtain a total of N data; (3) the N data of each scenario variable are randomly matched into N sample points (where each data is used only once during matching).

[0098] S340. Determine the failure probability corresponding to the autonomous driving system based on the sampling output result, the probability density function, and a preset failure condition.

[0099] Specifically, the preset failure conditions can be used to determine whether the scenario simulation result corresponding to each scenario variable value in the sampling output result is a failure result, and the distribution probability of the scenario variable corresponding to each failure result in the probability density function can be obtained. The failure probability of the autonomous driving system can be calculated based on the distribution probability of the scenario variable corresponding to each failure result.

[0100] The technical solution of this embodiment determines sampling output results using the probability density functions corresponding to each scenario variable, the range of each scenario variable, and the response surface. Furthermore, the failure probability corresponding to the autonomous driving system is determined using the sampling output results, the probability density function, and preset failure conditions. This quantifies the reliability analysis results of the autonomous driving system and predicts the failure probability of the autonomous driving system under all test scenarios. For example, if a scenario variable has a range of [0, 100] and is continuous within this range, then the values ​​of the scenario variable can be infinite, and therefore the number of simulation test scenarios can also be infinite. However, existing technologies cannot test all simulation test scenarios. Using the method provided by this embodiment, a sample can be used to replace the whole through sampling. The failure probability is calculated using the sampling output results, the probability density function, and the preset failure conditions. This allows for the reliability calculation and analysis of the autonomous driving system using an efficient sampling method, resolving the technical problem of existing technologies that cannot exhaustively test all test scenarios, thereby improving the accuracy of the reliability analysis results of the autonomous driving system.

[0101] Example 4

[0102] Figure 4 This is a flow chart of a method for determining the reliability of an autonomous driving system provided in the fourth embodiment of the present invention. This embodiment is applicable to situations where the reliability of an autonomous driving system is analyzed, and in particular, to situations where the failure probability of an autonomous driving system is calculated based on a pre-built vehicle model, sensor model, control model of the autonomous driving system, and scenario model. Figure 4 As shown, the reliability determination method of the autonomous driving system provided in this embodiment includes the following steps:

[0103] S410. Establish a data exchange channel between the autonomous driving simulation software and the mathematical optimization analysis software.

[0104] Currently, autonomous driving simulation software and mathematical optimization analysis software generally provide a wealth of interfaces, which can open up data interaction channels between the simulation software and the mathematical optimization analysis software directly (intercommunication between the simulation software and the mathematical optimization analysis software interfaces) or indirectly (with the help of tools such as Python programs).

[0105] S420. Build a vehicle model and sensor model in the autonomous driving simulation software, and integrate the control model.

[0106] Specifically, the control model can be built in Simulink or implemented using C++ code and integrated with other models (such as vehicle models and sensor models). Autonomous driving simulation software can also include vehicle and sensor models. The vehicle model inputs information such as vehicle speed, acceleration, steering wheel angle, and throttle and brake pedal openings into the control model; the sensor model inputs information such as the relative distance, relative speed, and relative angle between the target object and the vehicle into the control model. This approach connects the control model (autonomous driving control algorithm) and the vehicle model to form a closed loop.

[0107] S430. Build a scenario model in the autonomous driving simulation software.

[0108] Specifically, the construction of the scene model includes the following steps:

[0109] Step 1: Based on the autonomous driving system being tested, identify the test scenario type and define the scenario variables (including the type and range of the scenario variables).

[0110] Step 2: Build a simulation scene template;

[0111] The establishment of simulation scene templates includes: (1) defining static elements, including establishing scene road models, setting road length, lane width, road material properties, lane lines, and configuring environmental information; (2) defining dynamic elements, including setting the status of the vehicle, target vehicle, and other traffic participants, such as vehicle speed, initial position, and driving direction; (3) setting environmental conditions (lighting, rain, snow, fog, wind, etc.); (4) setting simulation duration, simulation trigger conditions, and simulation termination conditions.

[0112] Step 3: Perform the first sampling of the scenario variables, determine each simulation test scenario based on the sampling results and the simulation scenario template, and construct a scenario model based on each simulation test scenario.

[0113] S440: Perform joint simulation based on the autonomous driving simulation software and the mathematical optimization analysis software, and feed back the scene simulation results to the mathematical optimization analysis software through the data channel via the autonomous driving simulation software.

[0114] Specifically, the vehicle model, sensor model, control model of the autonomous driving system, and scenario model are simulated through the autonomous driving simulation software to obtain scenario simulation results. The autonomous driving simulation software transmits the scenario simulation results to the mathematical optimization analysis software through the data channel.

[0115] S450: construct a response surface between scenario variables and scenario simulation results through mathematical optimization analysis software, and determine the sensitivity information of each scenario variable.

[0116] Specifically, a response surface between the scenario variables and the scenario simulation results can be constructed based on the scenario variables and the scenario simulation results through an internal algorithm in the mathematical optimization analysis software.

[0117] S460: Define the distribution type of the scene variable and the probability density function of the scene variable.

[0118] S470: Define preset constraint conditions and preset failure conditions.

[0119] Specifically, mathematical optimization analysis software can be used to define preset constraints and failure conditions. For example, in the CCRm scenario for testing an AEB system, the preset constraint condition could be defined as the vehicle's speed exceeding the preceding vehicle's speed, and the preset failure condition could be defined as a collision between the two vehicles, meaning the minimum relative distance between them is less than 0.

[0120] S480: Perform a second sampling of each scenario variable and determine the failure probability based on the probability density function, the preset failure condition, and the response surface.

[0121] Specifically, each scenario variable can be sampled a second time based on the preset constraints, the range of the scenario variables, and the probability density function. Furthermore, the simulation results corresponding to the sampled values ​​of the scenario variables can be determined based on the response surface. Based on the preset failure conditions, it is judged whether the simulation results are failure results. Then, based on the distribution probability of the scenario variable values ​​corresponding to the failure results in the probability density function, the failure probability can be calculated.

[0122] The technical solution of this embodiment establishes a simulation scenario template based on the tested autonomous driving control algorithm. Based on this scenario, the scenario variables and their ranges are defined in mathematical optimization analysis software. An appropriate random sampling method is selected based on the type of scenario variables, and the scenario variables are sampled and combined to form a specific simulation test scenario. The generated simulation test scenario is automatically simulated in the simulation software, and the simulation results are automatically transmitted to the mathematical optimization analysis software. In the mathematical optimization analysis software, a high-quality response surface is constructed between the scenario variables and the scenario simulation results. The sensitivity of the scenario variables is quantified using variance analysis methods to clarify the importance of the scenario variables. Using efficient sampling methods and algorithms based on the probability distribution function of the scenario variables, the reliability of the autonomous driving system is calculated and quantified by defining system failure conditions and calculating their failure probabilities.

[0123] Example 5

[0124] Figure 5This is a structural diagram of a reliability determination device for an autonomous driving system provided in Example 5 of the present invention. This embodiment is applicable to situations where the reliability of an autonomous driving system is analyzed, and in particular, is applicable to situations where the failure probability of an autonomous driving system is calculated based on a pre-built vehicle model, sensor model, control model of the autonomous driving system, and scenario model. The device specifically includes: a simulation module 510, a response surface construction module 520, and a failure probability determination module 530.

[0125] A simulation module 510 is configured to obtain a scenario simulation result corresponding to the autonomous driving system based on a pre-built vehicle model, sensor model, control model of the autonomous driving system, and scenario model;

[0126] A response surface construction module 520 is configured to determine a response surface between the scenario variables and the scenario simulation results based on the scenario simulation results and the scenario variables corresponding to the scenario model;

[0127] The failure probability determination module 530 is used to determine the failure probability corresponding to the autonomous driving system based on the probability density function corresponding to each of the scenario variables, the range of each of the scenario variables, the response surface, and the preset failure condition.

[0128] Optionally, the device also includes a scene model construction module, which is used to obtain preset scene variables; obtain a pre-built simulation scene template; perform a first sampling process on each of the scene variables based on the type of the scene variables and the range of the scene variables, and determine the scene model based on the result of the first sampling process and the simulation scene template, wherein the scene variables include the type of the scene variables and the range of the scene variables.

[0129] Optionally, the device also includes a scene template construction module, which is used to build scene static elements, wherein the scene static elements include at least one of road information, lane information, and environmental information; build scene dynamic elements, wherein the scene dynamic elements include at least one of traffic characteristic information, vehicle information, target vehicle information, and other traffic participant information; obtain preset environmental conditions, simulation duration, simulation trigger conditions, and simulation termination conditions, and establish a simulation scene template based on the scene static elements, the scene dynamic elements, the preset environmental conditions, the simulation duration, the simulation trigger conditions, and the simulation termination conditions.

[0130] Optionally, the simulation module 510 is specifically configured to:

[0131] obtain vehicle motion information at a current time sent by the vehicle model and target information at the current time sent by the sensor model based on the control model; determine, by the control model, motion control information at a next time sent to the vehicle model based on the vehicle motion information at the current time, the target information at the current time, and the scenario model; and obtain a scenario simulation result corresponding to the autonomous driving system based on vehicle motion information at each time determined by the vehicle model, motion control information at each time determined by the control model, and target information at each time determined by the sensor model.

[0132] Optionally, the failure probability determination module 530 comprises a sampling unit and a probability calculation unit, wherein,

[0133] The sampling unit is configured to determine a sampling output result based on the probability density function corresponding to each scenario variable, the range of each scenario variable, and the response surface.

[0134] The probability calculation unit is configured to determine the failure probability corresponding to the autonomous driving system based on the sampling output result, the probability density function, and a preset failure condition.

[0135] Optionally, the sampling unit is specifically configured to:

[0136] obtain the probability density function corresponding to each scenario variable; perform second sampling processing on each scenario variable based on the probability density function and the range of the scenario variable to obtain a sampling variable result of each scenario variable; and determine a sampling output result based on the sampling variable result and the response surface.

[0137] Optionally, the apparatus further comprises a sensitivity analysis module configured to calculate sensitivity information corresponding to each scenario variable based on the scenario simulation result; and eliminate, from each scenario variable, a scenario variable that does not meet a preset sensitivity requirement based on the sensitivity information corresponding to each scenario variable.

[0138] In this embodiment, a pre-built vehicle model, sensor model, control model of the autonomous driving system, and scenario model are obtained through a simulation module, and the scenario simulation results of the autonomous driving system are determined. Based on the scenario simulation results and the scenario variables corresponding to the scenario model, a response surface between the scenario variables and the scenario simulation results is constructed through a response surface construction module. Further, based on the response surface, the probability density function of each scenario variable, the range of each scenario variable, and preset failure conditions, a failure probability corresponding to the autonomous driving system is determined through a failure probability determination module. This method quantifies the reliability results of the autonomous driving system by determining the failure probability, providing strong support for the development and improvement of the autonomous driving system. Moreover, this method predicts the failure probability of the autonomous driving system in all test scenarios through the probability density function of the scenario variables, solving the technical problem that the existing technology cannot exhaustively enumerate all test scenarios.

[0139] The reliability determination device of the autonomous driving system provided in an embodiment of the present invention can execute the reliability determination method of the autonomous driving system provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0140] It is worth noting that the various units and modules included in the above system are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the embodiments of the present invention.

[0141] Example 6

[0142] Figure 6 This is a structural diagram of an electronic device provided in Example 6 of the present invention. Figure 6 A block diagram of an exemplary electronic device 12 suitable for implementing embodiments of the present invention is shown. Figure 6 The electronic device 12 shown is merely an example and should not limit the functionality and scope of use of the embodiments of the present invention. The device 12 is typically an electronic device that performs reliability analysis of an autonomous driving system.

[0143] like Figure 6 As shown, the electronic device 12 is implemented as a general-purpose computing device. Components of the electronic device 12 may include, but are not limited to, one or more processors or processing units 16, a memory 28, and a bus 18 connecting the various components (including the memory 28 and the processing unit 16).

[0144] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration bus, a processor or local bus using any of a variety of bus architectures. By way of example, these architectures include Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MCA) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

[0145] Electronic device 12 typically includes a variety of computer readable media. These media can be any available media that is locally and / or remotely accessible by computer 12, including volatile and non-volatile media, removable and non-removable media.

[0146] Memory 28 can include computer readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. Electronic device 12 can further include other removable / non-removable, volatile / non-volatile computer storage media. By way of example only, storage device 34 can be used for reading from and writing to a non-removable, non-volatile magnetic media (not shown and typically called a "hard drive"). Figure 6 Although not specifically shown, computer 12 can further comprise other removable / non-removable, volatile / non-volatile computer storage media including, for example, a magnetic disk drive. Figure 6As shown in FIG. 1, the electronic device 12 includes one or more processor(s) 16, one or more memory device(s) 28, one or more communication interfaces 30, one or more input / output (I / O) interface(s) 22, and one or more bus(es) 18 for facilitating communication between the aforementioned components and / or external devices. The one or more communication interfaces 30 enable the electronic device 12 to communicate with one or more devices over one or more network(s) 32, such as one or more wired and / or wireless networks. In these and other implementations, the one or more communication interfaces 30 include, but are not limited to, one or more communication buses 34, one or more transmitters 36, one or more receivers 38, one or more antennas 40, and / or one or more other communication interfaces 42, such as a Universal Serial Bus (USB) interface, a Bluetooth® interface, a Near Field Communication (NFC) interface, and / or a ZigBee® interface, to name a few.

[0147] The electronic device 12 can also communicate with one or more external devices 14 such as a keyboard, a mouse, a camera, etc., and a display by means of one or more I / O interfaces 22. The electronic device 12 can also communicate with one or more devices that enable a user to interact with the electronic device 12 by means of one or more I / O interfaces 22, and / or one or more devices that enable the electronic device 12 to communicate with one or more other computing devices. Such communication can be facilitated by means of one or more communication interfaces 30. In these and other implementations, the one or more communication interfaces 30 enable the electronic device 12 to communicate with one or more devices over one or more wired and / or wireless networks. As shown, the one or more communication interfaces 30 communicate with the other components of the electronic device 12 by means of one or more communication buses 34. It should be appreciated that although the one or more communication interfaces 30 are shown as a single component in FIG. 1, the one or more communication interfaces 30 can include any number and / or combination of known communication interfaces, such as the one or more communication buses 34, the one or more transmitters 36, the one or more receivers 38, the one or more antennas 40, and / or the one or more other communication interfaces 42, to name a few.

[0148] The processor 16 performs functions of various applications and data processing by running programs stored in the memory 28 and can include, for example, a processor 16 configured to implement the methods of the present application described above. As shown in FIG. 1, the electronic device 12 also includes one or more memory device(s) 28, which can include a volatile memory, or a non-volatile storage, or a combination thereof. The one or more memory device(s) 28 can include, for example, a RAM, a ROM, an EEPROM, an Erasable Programmable Read Only Memory (EPROM), an Erasable Programmable ROM (EPROM), a flash memory, and / or a solid state drive (SSD), to name a few. The one or more memory device(s) 28 can also include a database, a database management system, and / or another similar database management program. The one or more memory device(s) 28 can store data, programs, and / or other information used by the processor 16.

[0149] Based on the pre-built vehicle model, sensor model, control model of the autonomous driving system, and scenario model, obtaining a scenario simulation result corresponding to the autonomous driving system;

[0150] Determining a response surface between the scenario variables and the scenario simulation results based on the scenario simulation results and the scenario variables corresponding to the scenario model;

[0151] The failure probability corresponding to the autonomous driving system is determined based on the probability density function corresponding to each of the scenario variables, the range of each of the scenario variables, the response surface, and the preset failure condition.

[0152] Of course, those skilled in the art will understand that the processor can also implement the technical solution of the reliability determination method of the autonomous driving system provided in any embodiment of the present invention.

[0153] Example 7

[0154] Embodiment 7 of the present invention further provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the steps of the method for determining the reliability of an autonomous driving system provided in any embodiment of the present invention are implemented. The method includes:

[0155] Based on the pre-built vehicle model, sensor model, control model of the autonomous driving system, and scenario model, obtaining a scenario simulation result corresponding to the autonomous driving system;

[0156] Determining a response surface between the scenario variables and the scenario simulation results based on the scenario simulation results and the scenario variables corresponding to the scenario model;

[0157] The failure probability corresponding to the autonomous driving system is determined based on the probability density function corresponding to each of the scenario variables, the range of each of the scenario variables, the response surface, and the preset failure condition.

[0158] The computer storage medium of the embodiment of the present invention may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device.

[0159] A computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0160] Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0161] The computer program code for performing the operations of the embodiments of the present invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0162] Note that the above are only preferred embodiments of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments, and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments and may include many other equivalent embodiments without departing from the concept of the present invention. The scope of the present invention is determined by the scope of the appended claims.

Claims

1. A method for determining the reliability of an autonomous driving system, characterized in that: The method comprises: Based on the pre-built vehicle model, sensor model, control model of the autonomous driving system, and scenario model, obtaining a scenario simulation result corresponding to the autonomous driving system; Determining a response surface between the scenario variables and the scenario simulation results based on the scenario simulation results and the scenario variables corresponding to the scenario model; Determining a sampling output result based on the probability density function corresponding to each of the scenario variables, the range of each of the scenario variables, and the response surface; wherein the sampling output result includes a fitting result when constructing the response surface; Determining a failure probability corresponding to the autonomous driving system based on the sampling output result, the probability density function, and a preset failure condition; The determining of the failure probability corresponding to the autonomous driving system based on the sampling output result, the probability density function, and a preset failure condition includes: By presetting failure conditions, it is determined whether the scenario simulation results corresponding to the values ​​of each scenario variable in the sampling output results are failure results, and the distribution probability of the scenario variables corresponding to each failure result in the probability density function is obtained. The failure probability of the autonomous driving system is calculated based on the distribution probability of the scenario variables corresponding to each failure result.

2. The method according to claim 1, characterized in that The method further comprises: Acquire preset scene variables, wherein the scene variables include the type of the scene variables and the range of the scene variables; Get pre-built simulation scenario templates; Based on the type of the scene variable and the range of the scene variable, a first sampling process is performed on each of the scene variables, and a scene model is determined based on a result of the first sampling process and the simulation scene template.

3. The method according to claim 2, characterized in that The method further comprises: Building scene static elements, wherein the scene static elements include at least one of road information, lane information, and environmental information; Building scene dynamic elements, wherein the scene dynamic elements include at least one of traffic characteristic information, own vehicle information, target vehicle information, and other traffic participant information; Acquire preset environmental conditions, simulation duration, simulation trigger conditions and simulation termination conditions, and establish a simulation scene template based on the scene static elements, the scene dynamic elements, the preset environmental conditions, the simulation duration, the simulation trigger conditions and the simulation termination conditions.

4. The method according to claim 1, wherein The obtaining of a scenario simulation result corresponding to the autonomous driving system based on a pre-built vehicle model, a sensor model, a control model of the autonomous driving system, and a scenario model includes: Acquire, based on the control model, vehicle motion information at a current moment sent by the vehicle model and target information at a current moment sent by the sensor model; Determining, by the control model, motion control information at a next moment based on the vehicle motion information at the current moment, the target information at the current moment, and the scene model, and sending the motion control information to the vehicle model; Based on the vehicle motion information at each moment determined by the vehicle model, the motion control information at each moment determined by the control model, and the target information at each moment determined by the sensor model, a scene simulation result corresponding to the automatic driving system is obtained.

5. The method according to claim 1, wherein The determining of the sampling output result based on the probability density function corresponding to each of the scenario variables, the range of each of the scenario variables, and the response surface includes: Obtaining a probability density function corresponding to each of the scenario variables; Based on the probability density function and the range of the scenario variables, performing a second sampling process on each of the scenario variables to obtain a sampling variable result of each of the scenario variables; A sampling output result is determined based on the sampling variable results and the response surface.

6. The method according to claim 1, characterized in that The method further comprises: Based on the scenario simulation results, calculating sensitivity information corresponding to each scenario variable; Based on the sensitivity information corresponding to each of the scene variables, scene variables that do not meet the preset sensitivity requirements are eliminated from the scene variables.

7. A reliability determination device for an automatic driving system, characterized in that: The device comprises: A simulation module, configured to obtain a scenario simulation result corresponding to the autonomous driving system based on a pre-built vehicle model, sensor model, control model of the autonomous driving system, and scenario model; a response surface construction module, configured to determine a response surface between the scenario variables and the scenario simulation results based on the scenario simulation results and the scenario variables corresponding to the scenario model; Failure probability determination module, including: a sampling unit, configured to determine a sampling output result based on a probability density function corresponding to each of the scenario variables, a range of each of the scenario variables, and the response surface; wherein the sampling output result includes a fitting result when constructing the response surface; a probability calculation unit, configured to determine a failure probability corresponding to the autonomous driving system based on the sampling output result, the probability density function, and a preset failure condition; The probability calculation unit is specifically used to determine whether the scenario simulation results corresponding to the values ​​of each scenario variable in the sampling output results are failure results through preset failure conditions, obtain the distribution probability of the scenario variables corresponding to each failure result in the probability density function, and calculate the failure probability of the autonomous driving system based on the distribution probability of the scenario variables corresponding to each failure result.

8. An electronic device, characterized in that: The electronic device comprises: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the reliability determination method of the automatic driving system as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the reliability determination method of the automatic driving system as described in any one of claims 1 to 6 is implemented.

Images