Method for constructing autonomous security guarantee architecture based on spatio-temporal whole-process information interaction

Abstract

The present invention provides a method for constructing an autonomous safety guarantee architecture based on spatio-temporal whole-process traffic information interaction. It includes: constructing a database, generating components of the safety guarantee system by fusing large language models in multiple languages; calibrating the spatial and functional attributes of the components using large language models; calibrating the relationships between components using large language model reasoning; generating the interaction time-sequence information flow of the components according to the relationships between the components; constructing a logical architecture covering functional modules and information flow according to the functional attributes and information flow; constructing a physical architecture covering spatial distribution and information flow according to the spatial attributes and information flow. The present invention solves the problems of numerous components related to safety guarantee, complex relationships, and difficulties in extracting information flow and constructing the architecture brought about by the rapid development of autonomous road traffic systems; provides architectural support for the construction of an autonomous road safety guarantee system for active intervention, and improves the autonomous road safety guarantee level.

Description

Technical Field

[0001] The present invention relates to the technical field of road traffic safety assurance, and in particular to a method for constructing an autonomous safety assurance architecture based on spatio-temporal whole-process information interaction. Background Art

[0002] The autonomous road traffic system is the leading direction of the new round of global intelligent transportation technology development. It is committed to integrating technologies such as vehicle-road cooperation, autonomous driving, big data analysis, satellite positioning, and 5G communication, so as to reduce manual intervention in the intelligent transportation system and increase its autonomous ability. With the continuous improvement of system complexity, how to ensure the safe and reliable operation of this intelligent system has become a key problem to be solved urgently.

[0003] At present, although the autonomous road traffic system has achieved many research results at the technical level, its safety assurance system is still in its infancy, especially lacking a safety assurance architecture for autonomous decision-making intervention. There are the following difficulties in constructing an autonomous safety assurance architecture: the ITS framework and the components of the autonomous traffic system architecture are too complex, and at the same time contain many elements and relationships that have no close connection with autonomous safety assurance; the information technology develops rapidly, and the functional attributes of the components in the existing architecture are difficult to adapt to the technology development; the relationships between components are complex, and it is difficult to generate the information flow of the safety assurance system for autonomous decision-making intervention.

[0004] Therefore, it is necessary to comprehensively utilize the current massive Internet resources and industry and R & D materials, and use advanced artificial intelligence technologies such as language large models to fully grasp the current development of autonomous road traffic technologies, extract the components of the autonomous road traffic safety assurance architecture for active decision-making intervention, construct an autonomous safety assurance architecture for spatio-temporal whole-process traffic information interaction, provide autonomous perception-decision-making-intervention functions, provide architecture support for the construction of an autonomous road safety assurance system for active intervention, and improve the level of autonomous road safety assurance. Summary of the Invention

[0005] An embodiment of the present invention provides a method for constructing an autonomous safety assurance architecture based on spatio-temporal whole-process information interaction to effectively improve the level of autonomous road safety assurance.

[0006] To achieve the above object, the present invention adopts the following technical solutions.

[0007] A method for constructing an autonomous safety assurance architecture based on spatio-temporal whole-process traffic information interaction includes:

[0008] Construct a combined database of a local database and Internet resources, and perform natural language processing on the database using multiple language large models to generate components of an autonomous safety assurance system;

[0009] Calibrate the spatial and functional attributes of the components of the security assurance system using a large model, where the spatial attributes include the spatial location of the current last-level component, and the functional attributes include the function of the current last-level component in the road traffic system;

[0010] Use large model reasoning to calibrate the component relationships of the components of the security assurance system;

[0011] Generate component interaction timing information flows based on the component relationships of the security assurance system;

[0012] Construct a logical architecture covering functional modules and information flows based on the functional attributes of the components of the security assurance system and the component interaction timing information flows;

[0013] Construct a physical architecture covering spatial distribution and information flows based on the spatial attributes of the components of the security assurance system and the component interaction timing information flows.

[0014] Preferably, build a combined database of local databases and Internet resources, perform natural language processing on the database using large models in multiple languages, and generate autonomous security assurance system components, including:

[0015] Step 1: Build a local database and calibrate the reference degree level of the data, generate a thesaurus for constructing questions, and define the first-level components as including {perception system, decision-making system, intervention system, auxiliary system, object of action};

[0016] Step 2: Select large models in multiple languages for deployment, unify the data interface standards of the large models in multiple languages, and build a fusion framework of large models in multiple languages according to the processes of intention understanding, document parsing, document indexing and vector embedding, knowledge retrieval and re-ranking, and text answering;

[0017] Step 3: For each component at the current level, generate a question bank according to the thesaurus, use the large model fusion framework to generate inference results, extract keywords, and generate the next-level alternative components corresponding to each current-level component;

[0018] Step 4: For the next-level alternatives, generate a synonym bank for the keywords based on the reference degree of the data from which the keywords are sourced, and select the keyword from the data with the highest reference degree in the same group of synonyms as the component name;

[0019] Step 5: Use the expert scoring method to score the importance of the components, and retain the components with an average score exceeding the threshold to form the formal next-level components;

[0020] Step 6: If the fourth-level components have been generated or the next-level components are empty, stop, otherwise go to Step 3, and use all the generated components at all levels to form an autonomous security assurance system.

[0021] Preferably, the components of the autonomous security assurance system include multiple levels, and the high-level components include multiple low-level components; among them, there are 5 first-level components, which are divided according to the basic logical order of perception - decision - intervention, the objects they act on, and the auxiliary systems they rely on, namely the perception system, the decision system, the intervention system, the auxiliary system, and the object of action; the second-level components are systems or modules generated by refining the functions of the first-level components; the third-level components and the fourth-level components are individual entity units generated by further refining each system or module in the second-level components.

[0022] Preferably, using large model reasoning to calibrate the relationships between the components of the security assurance system includes:

[0023] (1) Extract the materials used by the components to form a core database;

[0024] (2) Determine the specific relationship types according to the active / passive / interactive relationships, and the relationship types include: perception - being perceived, intervention - being intervened, sending information - receiving information - sending information / receiving information, and dependence - being dependent;

[0025] (3) According to the relationship types, generate a question bank for constructing the relationships between components for each component;

[0026] (4) Use the large model to generate inference results, perform relationship matching, and determine the alternative relationships between components;

[0027] (5) Starting from the lowest-level components, if all sibling components at the same level have a passive relationship with the same component A, then promote the passive relationship between the sibling components at the same level and component A to the parent component of the sibling components, and complete the calibration of the relationships between all components of the security assurance system using the relationships between all components.

[0028] Preferably, generating the component interaction time-series information flow according to the relationships between the components of the security assurance system includes:

[0029] (1) Map the component interaction time series according to the basic order of perception - decision - intervention and the relationships between components;

[0030] (2) Generate the information flow between components using the component interaction time series;

[0031] (3) Evaluate the rationality of the information flow through expert scoring, and remove the information flow whose rationality does not meet the threshold;

[0032] (4) Calibrate different intervention levels for the reasonable information flow.

[0033] Preferably, construct a logical architecture covering functional modules and information flows based on the functional attributes of the components of the security assurance system and the sequential information flows of component interactions, including: (1) Construct a basic functional logic framework based on the first- and second-level components of the security assurance system, including four basic frameworks of perception, decision-making, intervention, and communication, as well as an internal framework;

[0034] (2) Map the component functions to different positions in the functional logic framework according to the functional attributes of the third- and fourth-level components of the security assurance system;

[0035] (3) Construct the information flows within the logical functional architecture according to the sequential information flows of component interactions to generate a functional logic architecture suitable for different intervention levels.

[0036] Preferably, construct a physical architecture covering spatial distribution and information flows based on the spatial attributes of the components of the security assurance system and the sequential information flows of component interactions, including:

[0037] (1) Construct a basic physical framework according to the spatial attribute categories of the components of the security assurance system, and this basic physical framework includes cloud, road, roadside, and other spaces;

[0038] (2) Map the third- and fourth-level components to different positions in the basic physical framework according to the spatial relationships of the lowest-level components of the security assurance system;

[0039] (3) Construct the information flows within the physical architecture according to the sequential information flows of component interactions to generate a physical architecture covering different intervention levels.

[0040] It can be seen from the technical solutions provided by the embodiments of the present invention above that the present invention solves the problems of numerous and complex components related to security assurance, difficult extraction of information flows, and difficult construction of architectures brought about by the rapid development of autonomous road traffic systems; it provides architectural support for constructing an autonomous road safety assurance system for active intervention and improves the autonomous road safety assurance level.

[0041] Additional aspects and advantages of the present invention will be given in part in the following description, and these will become obvious from the following description or can be understood through the practice of the present invention. Description of the Drawings

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0043] Figure 1Schematic flowchart of a method for constructing an autonomous safety guarantee architecture based on spatio-temporal whole-process traffic information interaction provided by an embodiment of the present invention;

[0044] Figure 2 Schematic flowchart of a method for generating components of an autonomous safety guarantee system provided by an embodiment of the present invention;

[0045] Figure 3 Example diagram of components of an autonomous safety guarantee system provided by an embodiment of the present invention;

[0046] Figure 4 Schematic flowchart of a method for calibrating component relationships provided by an embodiment of the present invention;

[0047] Figure 5 Example diagram of information flow of an autonomous safety guarantee system provided by an embodiment of the present invention;

[0048] Figure 6 Example diagram of a logical architecture of an autonomous safety guarantee provided by an embodiment of the present invention;

[0049] Figure 7 Example diagram of a physical architecture of an autonomous safety guarantee provided by an embodiment of the present invention. Detailed implementation manners

[0050] The following details the implementation manners of the present invention. Examples of the implementation manners are shown in the drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The implementation manners described below with reference to the drawings are exemplary and are only used to explain the present invention and should not be construed as a limitation of the present invention.

[0051] Those skilled in the art of the present technology can understand that, unless specifically stated otherwise, the singular forms "a", "an", "the" and "said" used herein may also include the plural forms. It should be further understood that the term "comprising" used in the specification of the present invention means the presence of the described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or their groups. It should be understood that when we say an element is "connected" or "coupled" to another element, it can be directly connected or coupled to other elements, or there may also be intermediate elements. In addition, the "connection" or "coupling" used herein may include wireless connection or coupling. The phrase "and / or" used herein includes any and all combinations of one or more of the associated listed items.

[0052] Those skilled in the art can understand that, unless otherwise defined, all terms used herein (including technical terms and scientific terms) have the same meaning as the general understanding of those of ordinary skill in the art to which the present invention belongs. It should also be understood that terms such as those defined in a general dictionary should be understood to have a meaning consistent with the meaning in the context of the prior art, and will not be interpreted with an idealized or overly formal meaning unless defined as such here.

[0053] For the convenience of understanding the embodiments of the present invention, the following will further explain with several specific embodiments in conjunction with the accompanying drawings, and each embodiment does not constitute a limitation on the embodiments of the present invention.

[0054] The processing flow of a method for constructing an autonomous safety guarantee architecture based on spatio-temporal whole-process traffic information interaction provided by the embodiments of the present invention is as Figure 1 shown, including the following processing steps:

[0055] Step S1: Construct a local database and an Internet resource combined database, and perform natural language processing on the database using a multi-language large model to generate components of the autonomous safety guarantee system.

[0056] Step S2: Use the large model to calibrate the spatial and functional attributes of the components of the safety guarantee system, where the spatial attribute includes the spatial position of the current last-level component, and the functional attribute includes the function of the current last-level component in the road traffic system.

[0057] Step S3: Use large model reasoning to calibrate the component relationships.

[0058] Step S4: Generate the component interaction timing information flow according to the component relationships.

[0059] Step S5: Construct a logical architecture covering functional modules and information flows according to the functional attributes and information flows.

[0060] Step S6: Construct a physical architecture covering spatial distribution and information flows according to the spatial attributes and information flows.

[0061] The schematic diagram of the process for generating components of the autonomous safety guarantee system provided by the embodiments of the present invention is as Figure 2 shown, including the following processing steps;

[0062] Step 1: Use relevant Chinese and English periodicals, dissertations, national standards, local standards, and group norms in the fields of autonomous transportation, vehicle-road cooperation, intelligent driving, etc. to construct a local database, calibrate the reference level of the materials, and generate a thesaurus of synonyms required for constructing problems. For example, {component, system, object, module, unit} is a group of synonyms used to enrich the description of the component generation process, and {intervention, control, manipulation} is a group of synonyms used to expand the search scope and understanding dimension, and define that the first-level components include {perception system, decision-making system, intervention system, auxiliary system, object of action};

[0063] Step 2: Select language models such as ChatGLM3-6B, Qwen2-7B, Llama3, and Llama3-Chinese for deployment and unify their data interface standards. Build a fusion framework of multiple language models according to the processes of intention understanding, document parsing, document indexing and vector embedding, knowledge retrieval and re-ranking, and text answering. By combining the generation ability of the large model and the retrieval mechanism of the external knowledge base, improve the effect of natural language processing tasks;

[0064] Step 3: For each component at the current level, generate a question bank according to the thesaurus of synonyms, use the large model fusion framework to generate inference results, extract keywords, and generate the next-level alternative components corresponding to each current-level component;

[0065] Step 4: For the next-level alternative, generate a thesaurus of synonyms for the keywords based on the reference degree of the materials from which the keywords are sourced, and select the keyword from the material with the highest reference degree in the same group of synonyms as the component name;

[0066] Step 5: Use the expert scoring method to score the importance of the components, and retain the components with an average score exceeding the threshold to form the official next-level components;

[0067] Step 6: If the fourth-level components have been generated or the next-level components are empty, stop; otherwise, go to Step 3, and use all the generated components at all levels to form an autonomous security guarantee system.

[0068] Figure 3 This is an example diagram of the components of an autonomous security guarantee system provided by an embodiment of the present invention. Refer to Figure 3 , the components of the autonomous security guarantee system can be divided into multiple levels, and the high-level components include multiple low-level components; among them, there are 5 first-level components, which are divided according to the basic logical order of perception - decision-making - intervention, the object on which it acts, and the auxiliary system on which it depends, namely the perception system, the decision-making system, the intervention system, the auxiliary system, and the object of action; there are 15 second-level components, which are systems or modules generated by refining the functions of the first-level components; the third-level components and the fourth-level components are individual entity units generated by further refining each system or module in the second-level components.

[0069] For example, according to the functions of the perception system (primary component), the autonomous safety assurance system can be divided into three modules (secondary components), namely, the perception module, the perception positioning module, and the perception communication module; the perception module can be further refined into a vehicle-end perception module and a road-end perception module (tertiary components) according to its spatial attributes; in the vehicle-end perception module, there are three entity units (quaternary components), namely, an in-vehicle camera, an in-vehicle lidar, and an in-vehicle millimeter-wave radar, which work together to enable the vehicle to perform various perceptions in all directions and with high precision during road traffic driving.

[0070] Figure 4 The following is a schematic flowchart of the component relationship calibration provided for the embodiments of the present invention. Refer to Figure 4 , and the steps for calibrating the component relationships of the present invention are as follows:

[0071] Step 1: Extract the materials used by the components to form a core database;

[0072] Step 2: Determine the specific relationship types according to the active / passive / interactive relationships, including: perception - being perceived, intervention - being intervened, sending information - receiving information - sending information / receiving information, dependence - being dependent;

[0073] Step 3: Generate a question bank for constructing the relationships between components for each component according to the relationship types;

[0074] Step 4: Use the large model to generate inference results, perform relationship matching, and determine the alternative relationships between components;

[0075] Step 5: If there are passive relationships between the sibling components at the same level and the same component A, then promote the passive relationships between the sibling components at the same level and component A to the parent component of the sibling components, and complete the calibration of the component relationships using the relationships between all components of the safety assurance system.

[0076] Figure 5 The following is an example diagram of the information flow of the autonomous safety assurance system provided for the embodiments of the present invention; refer to Figure 5, taking the traffic event of non-motor vehicles running red lights at intersections as an example, the information flow process of the autonomous safety guarantee system can be described as follows: At the beginning, the first-level intervention vehicle A and the second-level intervention vehicle B sense the behavior of non-motor vehicles running red lights through the motion detection function of their respective vehicle-end sensing modules, and the roadside affiliated infrastructure D senses the behavior of non-motor vehicles running red lights and the location of the third-level intervention vehicle C through the motion detection and location detection functions of the roadside sensing module; Vehicles A, B, and facility D upload the sensed information to the cloud decision-making communication module of the cloud platform through their respective sensing communication units; The cloud decision-making module of the cloud platform makes decision zoning and research and judgment decisions on the current traffic situation according to the sensing information from the cloud communication module, and generates warning information and intervention parameters; Subsequently, it is received and sent to the cloud intervention generation module by the cloud intervention communication module; The cloud intervention generation module generates intervention instructions, and sends the intervention instructions to the vehicle-end intervention communication modules of vehicles A, B, and C through the cloud intervention communication module, and each vehicle-end intervention execution module receives and executes the intervention instructions. Among them, for the intervention instructions for risk warning of vehicle A and vehicle B, the driver is reminded to brake through the driver interaction unit, and the intervention instruction to brake is sent to its vehicle motion control unit for vehicle C; The driver of vehicle A operates the vehicle motion control unit of this vehicle to decelerate and brake to avoid, maintain a safe vehicle distance, and vehicle C brakes. After the driver interaction unit of vehicle B reminds the driver and detects no intention to brake, the cloud platform sends an intervention instruction again, executes the intervention instruction for driving advice on vehicle B and keeps monitoring, the driver interaction unit of vehicle B reminds the driver again, and when it detects that the TTC (time to collision) is less than 6s, executes the intervention instruction for vehicle control takeover on vehicle B, and vehicle B brakes emergently. After the non-motor vehicle passes, the system cancels the warning for vehicle A and the vehicle control takeover for vehicles B and C, and ends the information flow.

[0077] Figure 6 This is an example diagram of the autonomous safety guarantee logic architecture provided by the embodiment of the present invention. Refer to Figure 6 , the autonomous safety guarantee logic architecture includes four functional layers: sensing, decision-making, intervention, and communication. According to the functional attribute coding, information flow, and autonomy level of the components, the functions of the components are mapped to different positions in the functional logic framework. The sensing functional layer classifies the component sensing functions into four categories: motion sensing, distance sensing, position sensing, and meteorological monitoring to achieve a full-range perception of various traffic elements and environments in the road traffic system; The decision-making functional layer classifies the component decision-making functions into decision zoning, hierarchical decision-making, and layered decision-making according to the generation order of decisions to describe the detailed process of decision generation; The intervention functional layer divides the component intervention functions into two categories: intervention generation and intervention execution according to the generation and execution process of the intervention, and at the same time also reflects the different intervention methods corresponding to different autonomy levels; In addition, during the implementation process of the sensing, decision-making, and intervention functions, the communication functional layer provides important functional support for data transmission and information interaction throughout the process.

[0078] Figure 7 This is an example diagram of an autonomous safety - guarantee physical architecture provided by an embodiment of the present invention. Referring to Figure 7 , the physical architecture includes the physical components and spatial attributes of a perception system, a decision - making system, an intervention system, an auxiliary system, and an object of action. According to the spatial - attribute coding, information flow, and autonomy level of the components, each component is mapped to different positions in the physical framework. The spatial attributes include spatial position and sub - spatial position. Among them, the spatial position can be divided into vehicle - end, roadside, and cloud - end. Specifically, each module and unit for vehicle - end perception, decision - making, and intervention is located at the vehicle - end of each level of intervention vehicles; the roadside - attached infrastructure, and the spatial positions of each module and unit for roadside perception, decision - making, and intervention and objects of action such as intervention vehicles are all at the roadside, and can be further refined into two sub - spatial positions: road and roadside; each module and unit for decision - making and intervention on the cloud platform is located at the cloud - end; other spaces are mainly the positions where the positioning system, communication system, and geographic information system in the auxiliary system are located, including the earth orbit where the satellite positioning system is located, the interior of the vehicle carrier where the geographic information system and inertial navigation system are located, the cloud - end where the geographic information system, cellular mobile communication positioning system, and communication system are located, the base stations of the cellular mobile communication positioning system and communication system, etc. Different spatial components interact with each other through the communication system and the communication units of each module in each system.

[0079] In summary, the autonomous road safety - guarantee system for spatio - temporal whole - process traffic information interaction of the present invention conducts architecture design, with the main line of autonomous perception - decision - making - intervention, having strong pertinence; comprehensively utilizing the current vast amount of Internet resources, industry, and R & D materials, making full use of advanced artificial intelligence technologies such as language large - model to fully grasp the current technological development of autonomous road systems, extracting the road traffic autonomous safety - guarantee architecture components for active decision - making and intervention, mining the functions and relationships of the components, ensuring the advancement and coverage of component functions, as well as the consistency of the architecture, components, and relationships, and improving the operability of the architecture.

[0080] The present invention constructs an autonomous safety - guarantee architecture for spatio - temporal whole - process traffic information interaction, provides autonomous perception - decision - intervention components, functions, and information flow, solves the problems of numerous and complex safety - guarantee - related components, difficult information - flow extraction, and architecture construction due to the rapid development of autonomous road traffic systems; provides architecture support for constructing an autonomous road safety - guarantee system for active intervention, and improves the level of autonomous road safety - guarantee.

[0081] Those of ordinary skill in the art can understand that the drawings are only schematic diagrams of an embodiment, and the modules or processes in the drawings are not necessarily essential for implementing the present invention.

[0082] As can be seen from the description of the above embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of the present invention.

[0083] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. The key point of each embodiment is to illustrate the differences from other embodiments. In particular, for the device or system embodiments, since they are basically similar to the method embodiments, the description is relatively simple. For the relevant parts, reference can be made to the partial description of the method embodiments. The device and system embodiments described above are only illustrative. The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0084] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A method for constructing an autonomous safety assurance architecture based on the interaction of traffic information in the entire time and space, characterized in that: include: Construct a combined database of local and Internet resources, and use a large multi-language model to perform natural language processing on the database to generate autonomous security system components; Using the large model to calibrate the spatial and functional attributes of the safety assurance system components, wherein the spatial attributes include the spatial location of the current last-level component, and the functional attributes include the function of the current last-level component in the road traffic system; Using large model reasoning, calibrating component relationships of components of the safety assurance system; generating a component interaction time sequence information flow according to the component relationship of the security assurance system; Constructing a logical architecture covering functional modules and information flows according to the functional attributes of the components of the security assurance system and the information flows of the interaction sequence of the components; According to the spatial properties of the components of the security system and the temporal information flow of the interaction of the components, a physical architecture covering the spatial distribution and information flow is constructed; The construction of a local database and an Internet resource combination database, using a multi-language large model to perform natural language processing on the database, and generating autonomous security system components, includes: Step 1: Build a local database and calibrate the reference level of the data, generate the synonym database required to construct the problem, and define the first-level components including {perception system, decision-making system, intervention system, auxiliary system, and object of action}; Step 2: Select multiple language models for deployment, unify the data interface standards of the language models, and build a fusion framework for multiple language models according to the process of intent understanding, document parsing, document indexing and vector embedding, knowledge retrieval and re-ranking, and text answering; Step 3: For each component at the current level, generate a question library based on the synonym library, use the large model fusion framework to generate reasoning results, extract keywords, and generate the next level candidate components corresponding to each current level component; Step 4: Generate a synonym library for the keywords based on the reference degree of the keyword source for the next level of candidates, and select the keyword from the highest reference degree in the same group of synonyms as the component name; Step 5: Use the expert scoring method to score the importance of the components, and retain the components with average scores exceeding the threshold to form the formal next-level components; Step 6: If the fourth-level components have been generated or the next-level components are empty, stop, otherwise go to step 3 and use all the generated level components to form an autonomous security assurance system; The components of the autonomous safety assurance system include multiple levels, and the high-level components contain multiple low-level components; there are 5 first-level components, which are divided according to the basic logical order of perception-decision-intervention and the objects they act on and the auxiliary systems they rely on, namely perception system, decision-making system, intervention system, auxiliary system and object of action; the second-level group is a system or module generated by the functional refinement of the first-level components; the third-level and fourth-level components are individual entity units generated by further refinement of each system or module in the second-level components.

2. The method according to claim 1, characterized in that The method of using large model reasoning to calibrate the component relationships of the components of the safety assurance system includes: (1) Extract the data used for the components to form a core database; (2) Determine the specific relationship type according to the active / passive / interactive relationship, which includes: perception-being perceived, intervention-being intervened, sending information-receiving information-sending information / receiving information, and dependence-being dependent; (3) Based on the relationship type, generate a question library for each component required to construct the relationship between components; (4) Use the large model to generate inference results, perform relationship matching, and determine alternative relationships between components; (5) Starting from the lowest level component, if the brother components at the same level all have a passive relationship with the same component A, the passive relationship between the brother components at the same level and component A is promoted to the parent component of the brother components, and the component relationship calibration is completed using the relationship between all components of the security system.

3. The method according to claim 1, characterized in that The generating of component interaction time sequence information flow according to the component relationship of the security assurance system includes: (1) Obtain the component interaction sequence based on the basic sequence of perception-decision-intervention and the relationship mapping between components; (2) Generate information flow between components using the component interaction sequence; (3) Evaluate the rationality of information flows through expert scoring and remove information flows that do not meet the rationality threshold; (4) Calibrate different intervention levels for reasonable information flows.

4. The method according to claim 1, characterized in that: The said constructing a logical architecture covering functional modules and information flows according to the functional attributes of the components of the security assurance system and the information flows of the interaction sequence of the components includes: (1) constructing a basic functional logical framework according to the primary and secondary components of the security assurance system, including four basic frameworks of perception, decision-making, intervention and communication and an internal framework; (2) mapping component functions to different positions in the functional logic framework according to the functional attributes of the third-level and fourth-level components of the security assurance system; (3) Based on the component interaction time sequence information flow, the information flow within the logical functional architecture is constructed to generate a functional logical architecture that is suitable for different intervention levels.

5. The method according to claim 1, characterized in that The physical architecture covering spatial distribution and information flow is constructed according to the spatial attributes of the components of the security system and the component interaction time sequence information flow, including: (1) constructing a basic physical framework according to the spatial attribute categories of the components of the safety assurance system, the basic physical framework including clouds, roads, roadsides and other spaces; (2) mapping the third and fourth level components to different positions of the basic physical framework according to the spatial relationship of the lowest components of the safety assurance system; (3) Based on the component interaction time sequence information flow, the information flow within the physical architecture is constructed to generate a physical architecture covering different intervention levels.

Images