Method, device and equipment for evaluating critical risk of expected function safety and storage medium

Abstract

The invention provides an expected function safety critical risk evaluation method, apparatus and device, and a storage medium. The method comprises the steps of constructing a risk evaluation basic test environment driven by an AI model; according to a hierarchical scene element system defined in an expected function safety standard, transmitting the hierarchical scene element system as a basic scene template to an AI model so as to perform scene generation training; based on a functional safety standard, the evaluation system for the scene exposure probability is transmitted to an AI model for model training, so that high-risk scenes are screened out, and a key scene library is formed; inputting a dynamic algorithm formula into an AI risk model based on behavior performance in an intelligent driving function operation process so as to carry out risk calculation and training of transverse and longitudinal behavior performance in different scenes; and inputting the severity evaluation index into an AI risk model, taking a preset critical collision point as a judgment basis of a critical risk scene, and taking the screened hazard scene as a hazard event. The objective of the invention is to optimize a traditional expected function safety hazard event identification process and improve scene coverage and data reliability of hazard event identification.

Description

Technical Field

[0001] This application belongs to the field of expected functional safety technology for intelligent connected vehicles, and particularly relates to a method, apparatus, device and storage medium for assessing critical risks of expected functional safety. Background Technology

[0002] As the level of automotive intelligence improves, the functional safety and expected functional safety technologies of automobiles are becoming increasingly important. The industry is paying more and more attention to functional safety and expected functional safety. However, due to the large number of scenarios involved in expected functional safety, the current scenario screening is often inefficient and incomplete. At the same time, when identifying hazardous scenarios based on the screened scenario library, there is often a lack of quantitative evidence to support it. Therefore, there are pain points such as difficulty in scenario identification and difficulty in hazard identification. Summary of the Invention

[0003] In view of this, this application aims to propose a method, apparatus, device and storage medium for assessing critical risks of expected functional safety, in order to solve the problems of difficulty in scene identification and difficulty in hazard identification under different scenarios in the current expected functional safety development process.

[0004] To achieve the above objectives, the technical solution of this application is implemented as follows:

[0005] Firstly, this application provides a method for assessing critical risks to expected functional safety, including:

[0006] Construct a basic testing environment for risk assessment driven by AI models, including deploying the AI ​​model to be evaluated to the testing platform and calibrating the parameter thresholds of the model's input and output interfaces;

[0007] Based on the hierarchical scenario element system defined in the expected functional safety standard, and as a "basic scenario template", it is fed into the AI ​​model for scenario generation training;

[0008] Based on functional safety standards, the evaluation system for the probability of scene exposure is fed into the AI ​​model for model training, in order to screen out high-risk scenes and form a key scene library;

[0009] Based on the behavior of the intelligent driving function during operation, it is divided into longitudinal control behavior and lateral control behavior. The vehicle control behavior is combined with the scene, surrounding environment and participants, and the dynamic algorithm formula is input into the AI ​​risk model to calculate and train the risk of lateral and longitudinal behavior in different scenarios.

[0010] The severity assessment index is input into the AI ​​risk model, and the preset critical collision point is used as the basis for judging the critical risk scenario. The selected hazard scenarios are then regarded as hazard events.

[0011] Secondly, based on the same inventive concept, this application also provides a device for assessing critical risks to expected functional safety, comprising:

[0012] The test environment construction module is configured to build a basic test environment for AI model-driven risk assessment, which includes deploying the AI ​​model to be evaluated to the test platform and calibrating the parameter thresholds of the model's input and output interfaces.

[0013] The scene model training module is configured to use a hierarchical scene element system defined in the expected functional safety standard, and is fed into the AI ​​model as a "basic scene template" for scene generation training.

[0014] The scenario screening module is configured to be based on automotive functional safety standards, and to feed the evaluation system for the probability of scenario exposure into the AI ​​model for model training, so as to screen out high-risk scenarios and form a key scenario library.

[0015] The risk model training module is configured to be based on the behavior of the intelligent driving function during operation, which is divided into longitudinal control behavior and lateral control behavior. The vehicle control behavior is combined with the scene, surrounding environment and participants, and the dynamic algorithm formula is input into the AI ​​risk model to calculate and train the risk of lateral and longitudinal behavior in different scenarios.

[0016] The critical data screening module is configured to input severity evaluation indicators into the AI ​​risk model, use preset critical collision points as the basis for judging critical risk scenarios, and treat the screened hazard scenarios as hazard events.

[0017] Thirdly, based on the same inventive concept, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method described in the first aspect.

[0018] Fourthly, based on the same inventive concept, this application also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions for causing the computer to perform the method as described in the first aspect.

[0019] Compared with existing technologies, the expected functional safety critical risk assessment method, apparatus, equipment, and storage medium described in this application have the following beneficial effects:

[0020] The critical risk assessment method for anticipated functional safety described in this application trains an AI model by combining a basic scenario library, a scenario library selection method, a vehicle behavior dynamics calculation formula, and the severity evaluation index of J2980. This aims to optimize the traditional process of identifying hazard events in anticipated functional safety, improve the scenario coverage and data reliability of hazard event identification, and effectively address the pain points of difficulty in scenario identification and hazard identification under different scenarios in the current anticipated functional safety development process. Attached Figure Description

[0021] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0022] Figure 1 This is a flowchart of the expected functional safety critical risk assessment method described in the embodiments of this application;

[0023] Figure 2 This is a schematic diagram of the basic elements of the scene library described in the embodiments of this application;

[0024] Figure 3 This is a schematic diagram illustrating the relative collision speed between the vehicle and the target vehicle as described in an embodiment of this application;

[0025] Figure 4 This is a schematic diagram illustrating the basis for selecting the S value as described in the embodiments of this application;

[0026] Figure 5 This is a schematic diagram of the expected functional safety critical risk assessment device described in the embodiments of this application;

[0027] Figure 6 This is a schematic diagram of the hardware structure of the electronic device described in an embodiment of this application. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with specific embodiments and the accompanying drawings.

[0029] It should be noted that, unless otherwise defined, the technical or scientific terms used in the embodiments of this application should have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms "first," "second," and similar terms used in the embodiments of this application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are only used to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0030] The embodiments of this application are described in detail below with reference to the accompanying drawings.

[0031] Please see Figure 1 As shown in the figure, this embodiment provides a method for assessing critical risks to expected functional safety, which specifically includes the following steps:

[0032] Step A10: Construct a basic test environment for risk assessment driven by AI models, including deploying the AI ​​model to be evaluated to the test platform and calibrating the parameter thresholds of the model's input and output interfaces.

[0033] Specifically, in this embodiment, the actual purpose of this step is to build a test environment for the AI ​​large model. In this step, the input interface (including data from sensors such as cameras and radar) and the output interface (including controllers of the vehicle's external performance, such as braking system, power system, and steering system) will be calibrated. This embodiment focuses on the evaluation of the expected functional safety scenario library and critical risk scenarios. The input and output interfaces can be set according to the actual products of the enterprise, which will not be elaborated on here.

[0034] Step A20: Based on the hierarchical scenario element system defined in the expected functional safety standard, and as a "basic scenario template", it is sent to the AI ​​model for scenario generation training.

[0035] Specifically, in this embodiment, the hierarchical scene elements based on the standard ISO 21448 are used as a "basic scene template" and fed to the AI ​​model for scene generation training. The underlying logic is based on, for example, Figure 2The different road structure elements shown (such as lane line type, intersection form, traffic facility layout, etc.) are input into the AI ​​big model for learning. The AI's underlying logic arranges and combines the scene elements to build a structured scene library covering different environmental conditions, traffic participant states and interactive behaviors.

[0036] Step A30: Based on functional safety standards, the evaluation system for the probability of scene exposure is fed into the AI ​​model for model training in order to screen out high-risk scenes and form a key scene library.

[0037] Specifically, in this embodiment, based on the given E-value selection criteria, the exposure probability index of different scenarios is input into the AI ​​big model. Through training the AI ​​big model, the scenario library generated in step A20 is classified according to the exposure E-value, which can filter out scenarios under different exposure levels and realize the screening of key scenarios or high-frequency scenarios.

[0038] For example, on an urban road, under normal weather conditions, the vehicle is behind another vehicle, and there is a target vehicle traveling ahead in the same lane. According to ISO 26262 and VDA702 regulations, E=4 for urban road driving; E=4 for normal weather conditions; E=4 for a target vehicle traveling ahead. Therefore, based on the principle that E is greater than 0, this scenario falls into the category of critical scenarios.

[0039] Compared to the current development of scenario libraries for expected functional safety, this embodiment relies more on brainstorming by developers and the arrangement and combination of scenario elements under different scenario architecture layers. This can lead to a large number of scenario libraries, a low percentage of effective scenarios, and an inability to efficiently distinguish key scenarios. In this embodiment, scenario elements and scenario exposure evaluation systems are integrated into an AI model, which facilitates a more efficient and accurate process of creating a scenario library-level screening of key scenarios during the expected functional safety development process.

[0040] Step A40: Based on the behavior of the intelligent driving function during operation, it is divided into longitudinal control behavior and lateral control behavior. The vehicle control behavior is combined with the scene, surrounding environment and participants. The dynamic algorithm formula is input into the AI ​​risk model to calculate and train the risk of lateral and longitudinal behavior in different scenarios (calculate the actual parameters such as the relative collision speed, acceleration and deceleration between the vehicle and the target object, and use them as the data source for different scenarios).

[0041] Specifically, in this embodiment, longitudinal behavior (acceleration / deceleration control) serves as an example to demonstrate the dynamics model building process, simulating a driving environment including the driver vehicle and a target vehicle, such as... Figure 3 As shown.

[0042] With the vehicle as the following vehicle and the target vehicle as the preceding vehicle, by defining the initial speed of the preceding vehicle, the acceleration / deceleration of the preceding vehicle, the initial speed of the following vehicle, the acceleration / deceleration of the following vehicle before it reacts, the acceleration / deceleration of the following vehicle after it reacts, and the initial relative distance between the two vehicles, data such as the time of the frontal collision and the relative collision speed can be calculated. Figure 3 The specific parameters are defined as follows:

[0043] v 前 : Initial speed of the vehicle in front; a 前 : Acceleration / deceleration of the vehicle in front; v 后 : Initial speed of the following vehicle; a 后1 Acceleration / deceleration when the following vehicle does not react; a 后2 : Acceleration / deceleration of the following vehicle after its reaction; △S: Initial relative distance between the two vehicles; t1: Reaction time; t2: Action time; FTTI = t1 + t2; v 前末 : Final velocity of the preceding vehicle; v 后末 Final velocity of the following vehicle; Collision Δv: Relative velocity at the time of collision; S1: Distance traveled by the preceding vehicle before it stopped within the FTTI; S 反 S: The distance traveled by the following vehicle within its reaction time; 动 : The distance traveled by the following vehicle during the action time; S 前刹停 +△S: Distance traveled by the vehicle in front when it decelerates to a stop + initial relative distance between the two vehicles; Distance of the vehicle in front S1 + △S: Distance traveled by the vehicle in front before it stops + initial relative distance between the two vehicles; Distance of the vehicle behind S2 = S 反 +S 动 ;

[0044] Judgment criteria: When the vehicle in front has not stopped, a collision occurs if the distance S2 between the following vehicle and the vehicle in front is greater than or equal to the distance S1 + ΔS; when the vehicle in front decelerates to a stop, a collision occurs if the distance S2 between the following vehicle and the vehicle in front is greater than or equal to the distance S. 前刹停 A collision occurs when the distance is +△S.

[0045] Formula explanation:

[0046] v 前末 =v 前 +(t1+t2)×a 前 ;

[0047] v 后末 =v 后 +a 后1 ×t1+a 后2 ×t2;

[0048] Collision Δv = (v 后末 -v 前末 ) × 3.6;

[0049] S1=v 前 ×(t1+t2)+0.5×a 前×(t1+t2)×(t1+t2;

[0050] S 反 =v 后 ×t1+0.5×a 后1 ×t1×t1;

[0051] S 动 =(v 后 +a 后1 ×t1)×t2+0.5×a 后2 ×t2×t2;

[0052] Distance to the vehicle in front (when the vehicle in front has come to a complete stop): S 前刹 Stop + ΔS = (v 前 +V 前末 ) × (v 前末 -v 前 ) / 2 / a 前 +△S;

[0053] Distance to the vehicle in front (when the vehicle in front has not stopped) = S1 + ΔS;

[0054] The distance to the following vehicle is S2 = S 反 +S 动 ;

[0055] TTC:

[0056] If the speeds of the vehicles in front and behind are the same, the relative distance = the speed difference between the vehicles in front and behind (including acceleration) × TTC;

[0057] If the speeds of the vehicles in front and behind are different, the relative distance = the speed difference between the vehicles in front and behind × TTC.

[0058] Analysis example:

[0059] Assume the following parameters: distance between the two vehicles is 1.8sTTC, the following vehicle is traveling at 80km / h, the preceding vehicle is traveling at 40km / h, decelerating at 9m / s², t1=0.5s; t2=0.71s; FTTI=1.21s;

[0060] Given the given conditions, calculations show that: △S = 20m;

[0061] v 前末 =v 前 +(t1+t2)×a 前 =0.221111111m / s;

[0062] v 后末 =v 后 +a 后1 ×t1+a 后2 ×t2=22.22222222m / s;

[0063] S1=v 前 ×(t1+t2)+0.5×a 前 ×(t1+t2)×(t1+t2)=6.855994444m;

[0064] S 反 =v 后 ×t1+0.5×a 后1 ×t1×t1=11.11111111m;

[0065] S 动 =(v 后 +a 后1 ×t1)×t2+0.5×a 后2 ×t2×t2=15.77777778m;

[0066] Since the vehicle in front did not stop, S2 should be compared with S1+△S.

[0067] S2=S 反 +S 动 =26.88888889m;

[0068] S1 + ΔS = 26.85599444

[0069] S2 > S1 + △S; a collision will occur.

[0070] Collision Δv = (v 后末 -v 前末 ) × 3.6 = 79.204 km / h.

[0071] Step A50: Input the severity evaluation index into the AI ​​risk model, use the preset critical collision point as the basis for judging critical risk scenarios, and take the selected hazard scenarios as hazard events.

[0072] Specifically, in this embodiment, Figure 4 The S-value evaluation index is input into the AI ​​model, considering the critical point from S0 to S1. The critical point from S0 to S1 is used as the basis for judging critical risk scenarios. That is, the range greater than the critical point is considered to have harm, and the range less than the critical point is considered to have no harm.

[0073] Taking the assumed parameters of the embodiment as an example, the calculated relative collision speed Δv = 79.204 km / h; according to Figure 4 The scope is defined as follows: S is S3, which is greater than 0. Since the scenario involved is a critical scenario, i.e., E > 0, it can be analyzed as a hazardous event.

[0074] In this embodiment, the example scenario and parameters can be used as a hazard event to support the analysis of subsequent expected functional safety triggering conditions. A hazard event library can be generated through the above analysis method, which can effectively improve the efficiency of expected functional safety analysis.

[0075] Compared to current methods for identifying hazardous events that are expected to ensure functional safety, this embodiment employs more qualitative analysis methods, and the critical risk points of collision scenarios are often difficult to define. This embodiment, based on the integration of basic scenarios, vehicle dynamics models, and the S-value evaluation system in J2980, can realize the ODD boundary defined by the application function. By inputting initial parameter values, the collision critical point can be calculated based on key scenarios, which can more effectively screen out hazardous scenarios and support the analysis of triggering conditions for subsequent functional deficiencies and human misuse.

[0076] The method described in this embodiment forms a complete method for assessing critical risks to expected functional safety through AI risk model training and critical risk screening. After inputting initial parameter values ​​into the AI ​​model, it automatically generates a critical risk scenario library, which serves to screen hazardous scenarios and is used as the basis for judging expected functional safety hazard events, supporting the analysis of triggering conditions for subsequent functional deficiencies and human misuse.

[0077] It should be noted that the above description describes some embodiments of this application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in a different order than that shown in the above embodiments and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0078] Based on the same inventive concept, and corresponding to the methods of any of the above embodiments, the embodiments of this application also provide a device for assessing critical risks to expected functional safety.

[0079] like Figure 5 As shown, the expected functional safety critical risk assessment device includes:

[0080] Test environment construction module 11 is configured to build a basic test environment for risk assessment driven by AI model, including deploying the AI ​​model to be evaluated to the test platform and calibrating the parameter thresholds of the model input interface and output interface;

[0081] The scene model training module 12 is configured to be based on the hierarchical scene element system defined in the expected functional safety standard, and is fed to the AI ​​model as a "basic scene template" for scene generation training.

[0082] The scenario screening module 13 is configured to input the evaluation system for the probability of scenario exposure in the automotive functional safety standard into the AI ​​model for model training, so as to screen out high-risk scenarios and form a key scenario library.

[0083] The risk model training module 14 is configured to be based on the behavior of the intelligent driving function during operation, which is divided into longitudinal control behavior and lateral control behavior. The vehicle control behavior is combined with the scene, surrounding environment and participants, and the dynamic algorithm formula is input into the AI ​​risk model to calculate and train the risk of lateral and longitudinal behavior in different scenarios.

[0084] The critical data screening module 15 is configured to input severity evaluation indicators into the AI ​​risk model, use the preset critical collision point as the basis for judging critical risk scenarios, and treat the screened hazard scenarios as hazard events.

[0085] For ease of description, the above apparatus is described in terms of its functions, divided into various modules. Of course, in implementing the embodiments of this application, the functions of each module can be implemented in one or more software and / or hardware.

[0086] The apparatus of the above embodiments is used to implement the corresponding method in any of the foregoing embodiments and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0087] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, embodiments of this application also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the methods described in any of the above embodiments.

[0088] Figure 6 This embodiment illustrates a more specific hardware structure of an electronic device, which may include a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, memory 1020, input / output interface 1030, and communication interface 1040 are interconnected internally via the bus 1050.

[0089] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0090] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program code is stored in the memory 1020 and is called and executed by the processor 1010.

[0091] The input / output interface 1030 is used to connect input / output modules to realize information input and output. The input / output modules can be configured as components in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touch screens, microphones, various sensors, etc., and output devices may include displays, speakers, vibrators, indicator lights, etc.

[0092] The communication interface 1040 is used to connect a communication module (not shown in the figure) to enable communication between this device and other devices. The communication module can communicate via wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0093] Bus 1050 includes a pathway for transmitting information between various components of the device, such as processor 1010, memory 1020, input / output interface 1030, and communication interface 1040.

[0094] It should be noted that although the above-described device only shows the processor 1010, memory 1020, input / output interface 1030, communication interface 1040, and bus 1050, in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the embodiments of this specification, and not necessarily all the components shown in the figures.

[0095] The electronic devices described above are used to implement the corresponding methods in any of the foregoing embodiments and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0096] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this application also provides a non-transitory computer-readable storage medium that stores computer instructions for causing the computer to perform the methods described in any of the above embodiments.

[0097] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0098] The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to perform the method described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0099] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of this application (including the claims) is limited to these examples; within the framework of this application, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of the embodiments of this application as described above, which are not provided in the details for the sake of brevity.

[0100] Additionally, to simplify the description and discussion, and to avoid obscuring the embodiments of this application, the well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided drawings. Furthermore, the apparatus may be shown in block diagram form to avoid obscuring the embodiments of this application, and this also takes into account the fact that the details of the implementation of these block diagram apparatuses are highly dependent on the platform on which the embodiments of this application will be implemented (i.e., these details should be fully understood by those skilled in the art). While specific details (e.g., circuits) have been set forth to describe exemplary embodiments of this application, it will be apparent to those skilled in the art that the embodiments of this application can be implemented without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0101] Although this application has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0102] The embodiments of this application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the embodiments of this application should be included within the protection scope of this application.

Claims

1. A method for assessing critical risks to expected functional safety, characterized in that, include: Construct a basic testing environment for risk assessment driven by AI models, including deploying the AI ​​model to be evaluated to the testing platform and calibrating the parameter thresholds of the model's input and output interfaces; Based on the hierarchical scenario element system defined in the expected functional safety standard, and as a "basic scenario template", it is fed into the AI ​​model for scenario generation training; Based on functional safety standards, the evaluation system for the probability of scene exposure is fed into the AI ​​model for model training, in order to screen out high-risk scenes and form a key scene library; Based on the behavior of the intelligent driving function during operation, it is divided into longitudinal control behavior and lateral control behavior. The vehicle control behavior is combined with the scene, surrounding environment and participants, and the dynamic algorithm formula is input into the AI ​​risk model to calculate and train the risk of lateral and longitudinal behavior in different scenarios. The severity assessment index is input into the AI ​​risk model, and the preset critical collision point is used as the basis for judging the critical risk scenario. The selected hazard scenarios are then regarded as hazard events.

2. The method according to claim 1, characterized in that: Using road structure elements as the core dimension, AI models are used to arrange and combine multi-dimensional scene elements in a regular manner to build a structured scene library covering different environmental conditions, traffic participant states, and interactive behaviors.

3. The method according to claim 2, characterized in that: Based on the given exposure value selection criteria, the exposure probability index of different scenarios is input into the AI ​​model. Through AI model training, the generated scenario library is classified and filtered according to the exposure value.

4. The method according to claim 1, characterized in that: The severity rating metric defines the range of relative collision speeds and thus the severity rating.

5. The method according to claim 4, characterized in that: A response above the critical collision point is considered hazardous, while a response below the critical collision point is considered harmless.

6. A device for assessing critical risks to expected functional safety, characterized in that, include: The test environment construction module is configured to build a basic test environment for AI model-driven risk assessment, which includes deploying the AI ​​model to be evaluated to the test platform and calibrating the parameter thresholds of the model's input and output interfaces. The scene model training module is configured to use a hierarchical scene element system defined in the expected functional safety standard, and is fed into the AI ​​model as a "basic scene template" for scene generation training. The scenario screening module is configured to be based on automotive functional safety standards, and to feed the evaluation system for the probability of scenario exposure into the AI ​​model for model training, so as to screen out high-risk scenarios and form a key scenario library. The risk model training module is configured to be based on the behavior of the intelligent driving function during operation, which is divided into longitudinal control behavior and lateral control behavior. The vehicle control behavior is combined with the scene, surrounding environment and participants, and the dynamic algorithm formula is input into the AI ​​risk model to calculate and train the risk of lateral and longitudinal behavior in different scenarios. The critical data screening module is configured to input severity evaluation indicators into the AI ​​risk model, use preset critical collision points as the basis for judging critical risk scenarios, and treat the screened hazard scenarios as hazard events.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the method as described in any one of claims 1-5.

8. A non-transitory computer-readable storage medium, characterized in that, in, The non-transitory computer-readable storage medium stores computer instructions for causing a computer to perform the method described in any one of claims 1-5.

Images