Vehicle part interaction logic verification file generation method and interaction logic verification method

Abstract

The invention provides a vehicle part interaction logic verification file generation method and an interaction logic verification method, and relates to the technical field of formal verification. The method comprises the following steps: acquiring multi-modal data of vehicle parts; analyzing the multi-modal data of the vehicle parts to obtain vehicle model information; according to the vehicle model information, respectively generating a vehicle part state transition path set and a linear sequential logic rule set; and generating an executable interaction logic verification file based on the vehicle component state transition path set and the linear sequential logic rule set. The vehicle model information is obtained based on the multi-modal data, the vehicle part state transition path set and the linear sequential logic rule set are automatically generated according to the vehicle model information, and the verification accuracy of interaction logic between vehicle parts is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of formal verification technology, in particular to a vehicle component interaction logic verification file generation method and an interaction logic verification method. BACKGROUND

[0002] With the development of automobile technology, from traditional mechanical systems to modern electronic control unit integrated systems, the coordinated work between automobile components becomes extremely critical. Moreover, the complexity of software and hardware is increasing. This complexity makes it more necessary to accurately understand the spatial relationship and operation sequence between multiple components.

[0003] However, the traditional solution often relies on manually defined rules to verify the spatial relationship and operation sequence between components. The manually defined rule approach is not only inefficient, but also limited by human factors. In addition, it is difficult to exhaustively enumerate all possible interaction situations and complex logical relationships between vehicle components through manual means. Limited by manually defined rules and manually determined logical relationships, the verification accuracy of the interaction logic between vehicle components is not high. SUMMARY

[0004] The purpose of the embodiments of the present application is to provide a vehicle component interaction logic verification file generation method and an interaction logic verification method to obtain vehicle model information based on multi-modal data, and automatically generate a vehicle component state transition path set and a linear temporal logic rule set according to the vehicle model information, thereby improving the verification accuracy of the interaction logic between vehicle components.

[0005] In a first aspect, the embodiments of the present application provide a vehicle component interaction logic verification file generation method, which includes: obtaining vehicle component multi-modal data; analyzing the vehicle component multi-modal data to obtain vehicle model information; generating a vehicle component state transition path set and a linear temporal logic rule set according to the vehicle model information; and generating an executable interaction logic verification file based on the vehicle component state transition path set and the linear temporal logic rule set.

[0006] The embodiments of the present application analyze multi-modal data to obtain vehicle model information. Since the multi-modal data includes vehicle-related data in different forms, the vehicle model information obtained from the multi-modal data is more complete and accurate, laying a foundation for subsequent generation of a vehicle component state transition path set and a linear temporal logic rule set. Furthermore, all possible vehicle component state transition paths and linear temporal rules in the vehicle model information are automatically generated according to the vehicle model information, thereby improving the verification accuracy of the interaction logic between vehicle components.

[0007] In some embodiments, the vehicle model information includes a first component type and a first component attribute parameter; and the manner of generating the vehicle component state transition path set includes: determining, according to the first component attribute parameter, executable actions and a movement range of a vehicle component corresponding to the first component type; and generating the vehicle component state transition path set of the vehicle based on the executable actions and the movement range of each vehicle component.

[0008] In the embodiments of the present application, the executable actions and the movement range of the corresponding vehicle component are automatically analyzed according to the first component attribute parameter, so that any possible single operation path of the vehicle component is not missed, and a basis is provided for subsequent combination of more complex operation paths.

[0009] In some embodiments, the vehicle model information further includes a first component interaction constraint parameter; and the manner of generating the vehicle component state transition path set of the vehicle based on the executable actions and the movement range of each vehicle component includes: determining an interaction constraint relationship between vehicle components based on the first component interaction constraint parameter; and generating the vehicle component state transition path set of the vehicle according to the interaction constraint relationship, the executable actions and the movement range of each vehicle component.

[0010] Based on the single operation path, the embodiments of the present application reasonably combine multiple steps according to the interaction constraint relationship between the vehicle components, and plan a possible complex operation path sequence in advance, so that all possible situations of the vehicle component state transition path are covered, thereby improving the accuracy of the interaction logic verification between the vehicle components.

[0011] In some embodiments, the vehicle model information includes a second component type, a second component attribute parameter and a second component interaction constraint parameter; and the manner of generating the linear temporal logic rule set includes: determining physical constraints of a vehicle component corresponding to the second component type according to the second component type and the second component attribute parameter; determining an interaction logic between vehicle components according to the physical constraints and the second component interaction constraint parameter; and generating the linear temporal logic rule set based on the physical constraints of each vehicle component and the interaction logic between each vehicle component.

[0012] The embodiments of the present application automatically generate the linear temporal logic rule set based on the physical constraints of each vehicle component and the interaction logic between each vehicle component, which not only reduces the workload of manually writing rules, but also improves the accuracy and completeness of the generated linear temporal logic rules, thereby improving the accuracy of the interaction logic verification between the vehicle components.

[0013] In some embodiments, after analyzing the multi-modal data of the vehicle components and obtaining the vehicle model information, the method further comprises: obtaining a user instruction; correcting the vehicle model information based on the user instruction to obtain new vehicle model information; and generating the set of vehicle component state transition paths and the set of linear temporal logic rules based on the vehicle model information, including: generating the set of vehicle component state transition paths and the set of linear temporal logic rules based on the new vehicle model information.

[0014] The embodiments of the present application improve the accuracy of the vehicle model information by correcting the vehicle model information based on the multi-modal data analysis according to the user instruction, thereby improving the accuracy of the generated set of vehicle component state transition paths and the set of linear temporal logic rules.

[0015] In a second aspect, the embodiments of the present application provide a vehicle component interaction logic verification method, which comprises: reading and parsing an executable interaction logic verification file to obtain a set of vehicle component state transition paths and a set of linear temporal logic rules; wherein the executable interaction logic verification file is generated by the method of any one of the embodiments of the first aspect; constructing a state space based on the set of vehicle component state transition paths; and traversing the state space according to the set of linear temporal logic rules to verify the vehicle component state transition paths in the set of vehicle component state transition paths and obtain a verification result.

[0016] The embodiments of the present application improve the accuracy of the verification result by verifying the executable interaction logic verification file, since the executable interaction logic verification file includes all possible vehicle component state transition paths and linear temporal rules.

[0017] In a third aspect, the embodiments of the present application provide a vehicle component interaction logic verification file generation device, which comprises: an acquisition module configured to acquire multi-modal data of vehicle components; an analysis module configured to analyze the multi-modal data of the vehicle components to obtain vehicle model information; a first generation module configured to generate a set of vehicle component state transition paths and a set of linear temporal logic rules based on the vehicle model information; and a second generation module configured to generate an executable interaction logic verification file based on the set of vehicle component state transition paths and the set of linear temporal logic rules.

[0018] In a fourth aspect, an embodiment of the present application provides a vehicle component interaction logic verification device, which includes: a reading module for reading an executable interaction logic verification file and parsing it to obtain a vehicle component state transition path set and a linear temporal logic rule set; wherein the executable interaction logic verification file is generated by the method of any one of the embodiments of the first aspect; a construction module for constructing a state space based on the vehicle component state transition path set; a verification module for traversing the state space according to the linear temporal logic rule set, verifying the vehicle component state transition paths in the vehicle component state transition path set, and obtaining a verification result.

[0019] In a fifth aspect, an embodiment of the present application provides an electronic device, comprising: a processor and a memory, the memory storing machine-readable instructions executable by the processor, and the machine-readable instructions, when executed by the processor, can execute the method steps of any one embodiment of the first aspect, the second aspect, or both aspects.

[0020] In a sixth aspect, an embodiment of the present application provides a computer-readable storage medium, comprising: computer program instructions are stored on the computer-readable storage medium, and when the computer program instructions are executed by a processor, the method steps of any one embodiment of the first aspect, the second aspect, or both aspects are executed.

[0021] In a seventh aspect, an embodiment of the present application provides a computer program product, comprising: computer program instructions, which, when executed by a processor, execute the method steps of the first aspect, the second aspect, or any one of the embodiments of these two aspects.

[0022] Other features and advantages of the present application will be described in the subsequent description, and in part will become apparent from the description, or will be understood by practicing the embodiments of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0024] Figure 1 A flowchart of a method for generating a vehicle component interaction logic verification file provided in an embodiment of the present application; Figure 2 A schematic diagram of a car cabin image provided in an embodiment of the present application; Figure 3 A flow chart of a vehicle component interaction logic verification method provided in an embodiment of the present application; Figure 4 A structural schematic diagram of a vehicle component interaction logic verification file generation apparatus provided by an embodiment of the present application is provided. Figure 5 A structural schematic diagram of a vehicle component interaction logic verification apparatus provided by an embodiment of the present application is provided. Figure 6 An electronic device structural schematic diagram provided by an embodiment of the present application is provided. DETAILED DESCRIPTION

[0025] The embodiments of the technical solutions of the present application will be described in detail below with reference to the drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present application, and therefore only serve as examples, and cannot limit the protection scope of the present application.

[0026] It should be noted that all the technical and scientific terms used herein have the same meaning as that generally understood by the technical personnel in the technical field of the present application; the terms used herein are only for the purpose of describing the specific embodiments, and are not intended to limit the present application; the terms "include" and "have" in the specification and claims of the present application and the above description of the drawings are intended to cover the non-exclusive inclusion.

[0027] In the description of the embodiments of the present application, the technical terms "first", "second", etc. are only used to distinguish different objects, and cannot be understood as indicating or implying relative importance or implicitly indicating the number, specific order or primary and secondary relationship of the indicated technical features. In the description of the embodiments of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly specified.

[0028] In the description of the embodiments of the present application, the term "and / or" is only a description of the association relationship between the associated objects, which means that there can be three relationships, for example, A and / or B, which can represent the three cases of A alone, A and B together, and B alone. In addition, the character " / " in this paper generally represents that the front and rear associated objects have an "or" relationship.

[0029] The state machine model can decompose complex behavior into a series of states and transitions between states. Each state represents a stable working state of the automotive component under certain conditions, and the transition between states represents the response of the component to external events or signals. For example, a controller in an automotive braking system can have "standby state", "brake activation state", "fault state", etc., and through the transition logic between states, it can be clearly described how the braking system works according to factors such as vehicle speed, driver instructions, etc.

[0030] Linear Temporal Logic (LTL) is a formal temporal logic system based on linear discrete time structures. It is mainly used to formally describe the temporal properties of a system.

[0031] Linear temporal logic uses a linear discrete time structure, modeling the timeline as an infinitely extending sequence. Temporal operators are used to describe the temporal relationships between events. In linear temporal logic, formulas are defined on a single, linear future timeline. In other words, time is viewed as a linear sequence along which the system transitions from one state to another.

[0032] The grammatical structure of linear temporal logic contains the following elements: Basic propositions: Atomic propositions (such as p, q) represent the state of the system; Boolean connectives: ∧ (and), ¬ (not), ∨ (or), etc. Temporal operators: such as F (future satisfaction), G (global satisfaction), U (until), X (next moment), etc.

[0033] For example: G(p) means that p must be true at all time points; F(p) means that p will eventually be true at a certain time point; pUq means that p is true until q is true.

[0034] Linear temporal logic rules can clearly define the sequential order, time intervals, and other constraints between vehicle components, ensuring that each component operates at the correct time and in the correct order. For example, during engine startup, fuel injection must occur before ignition. Linear temporal logic rules can accurately describe this process, ensuring a proper engine start.

[0035] Linear timing logic rules can be used to check whether the state transitions in the state machine model meet the expected logic specifications, which helps to discover potential timing problems during the design and development stages.

[0036] Large Language Model (LLM) is a natural language processing technology based on deep learning in the field of artificial intelligence. It processes user-given tasks through a process called "understanding input → reasoning → generating output."

[0037] Common large language models include GPT-4, Qwen, Gemini, DeepSeek, etc.

[0038] With the advancement of automotive technology, from traditional mechanical systems to modern integrated electronic control unit systems, the coordinated operation of automotive components has become extremely critical. Furthermore, the complexity of software and hardware continues to increase. This complexity makes it even more important to accurately understand the spatial relationships and operational sequences between multiple components. Any error in the interaction logic between components can lead to serious safety accidents.

[0039] Traditional automotive manufacturing design verification relies primarily on manual effort to define rules governing the spatial relationships and operational sequences between multiple components. This manual rule-definition approach is not only inefficient but also susceptible to human limitations, such as omissions and errors, resulting in incomplete and inaccurate rules. Furthermore, manual effort is unable to exhaustively enumerate all possible interactions and complex logical relationships between vehicle components. Due to the limitations of manually defined rules and manually determined logical relationships, verification accuracy for the interaction logic between vehicle components is low.

[0040] On this basis, the present application provides a method for generating a vehicle component interaction logic verification file and an interaction logic verification method. By abstracting the interaction logic of automobile components into a state machine model and linear temporal logic rules of the vehicle component state transition path, and mathematically verifying the vehicle component state transition path through linear temporal logic rules, potential errors and defects can be discovered, thereby improving the verification accuracy of the interaction logic between vehicle components.

[0041] Figure 1 This is a flowchart of a method for generating a vehicle component interaction logic verification file according to an embodiment of the present application. It is understood that the method for generating a vehicle component interaction logic verification file according to an embodiment of the present application can be implemented using an existing large language model or a pre-trained file generation model. The large language model or file generation model can be deployed on a terminal device (also referred to as an electronic device) or a server; the terminal device can specifically be a smartphone, tablet computer, computer, personal digital assistant (PDA), etc.; the server can specifically be an application server or a web server.

[0042] To facilitate understanding of the technical solution provided in the embodiments of the present application, a large language model is deployed in a terminal device, and the method provided in the embodiments of the present application is introduced by taking the method of generating a vehicle component interaction logic verification file using the large language model as an example.

[0043] like Figure 1 As shown, the method includes: Step S101: Acquire multimodal data of vehicle components.

[0044] Step S102: Analyze the multimodal data of vehicle components to obtain vehicle model information.

[0045] Multimodal data refers to data in multiple different forms. Multimodal data describes the same object or scenario from different perspectives and levels. Multimodal data includes at least two of the following: text, images, audio, video, and sensor data.

[0046] The vehicle component multimodal data refers to multimodal data used to describe vehicle components.

[0047] Taking the car cockpit as an example, the multimodal data of the car cockpit includes image data such as photos of the car cockpit, CAD screenshots, and three-dimensional model renderings, text data such as product manuals and design instructions of the car cockpit, and sensor data such as seat movement speed and seat status collected by sensors on the car cockpit.

[0048] The inherent semantic associations between data from different modalities reflect the complete information of vehicle components, helping to more accurately understand and recognize complex objects and scenarios.

[0049] During the specific implementation process, the multimodal data of vehicle components are input through the input box of the large language model, so that the large language model can read the multimodal data of vehicle components and parse the multimodal data of vehicle components.

[0050] In order to make the large language model clearly know how to parse the multimodal data of vehicle parts, the user can set prompt instructions in the large language model in advance, so that after receiving the multimodal data of vehicle parts, the large language model will parse the multimodal data of vehicle parts according to the preset prompt instructions.

[0051] Alternatively, the user inputs the prompt instruction and the vehicle component multimodal data simultaneously, so that the large language model parses the vehicle component multimodal data according to the prompt instruction.

[0052] For example, Figure 2 This is a schematic diagram of a car cabin image provided by an embodiment of the present application. The user enters the image data of the car cabin in the input box of the large language model (such as Figure 2 Examples of text data and prompt instructions are as follows:

Text Input

Prompt command

[0053] After receiving the prompt and the image and text data of the vehicle cabin, the large language model performs a parsing process to obtain vehicle model information. Vehicle model information is a structured digital description of the vehicle and its components.

[0054] The first vehicle model information example is as follows: { "entities":[ {"name":"Driver's seat","type":"Seat"}, {"name":"Passenger seat","type":"Seat"}, {"name":"Second row left seat","type":"seat"}, {"name":"Second row right seat","type":"seat"}, {"name":"Third row left seat","type":"seat"}, {"name":"Third row right seat","type":"seat"}, {"name":"steering wheel","type":"control panel"}, {"name":"Center console","type":"Control panel"}, {"name":"Dashboard","type":"Control Panel"}, {"name":"Center Armrest","type":"Seat Accessories"}, {"name":"door","type":"body structure"}, {"name":"Seat slide","type":"Guide rail"} ], "spatial_relations":[ {"from":"Driver's seat","to":"Passenger's seat","relation":"Parallel"}, {"from":"Driver's seat","to":"Second row left seat","relation":"Located in front"}, {"from":"Passenger seat","to":"Second row right seat","relation":"Located in front"}, {"from":"Second row left seat","to":"Third row left seat","relation":"Located in front"}, {"from":"Second row right seat","to":"Third row right seat","relation":"Located in front"}, {"from":"steering wheel","to":"center console","relation":"on the left side"}, {"from":"Dashboard","to":"Steering wheel","relation":"Located at the rear"}, {"from":"Center armrest","to":"Driver's seat","relation":"Located on the right side"}, {"from":"Center armrest","to":"Passenger seat","relation":"Located on the left"}, {"from":"door","to":"driver's seat","relation":"located on the left"}, {"from":"door","to":"second row left seat","relation":"located on the left"}, {"from":"door","to":"third row left seat","relation":"located on the left"} ], "dimensions":[], "angle_info":[], "movable_structures":[ {"name":"Seat slide","possible_motion":"Move forward and backward"} ] } In an alternative embodiment, vehicle component image data and a first prompt can be input into a large language model. Based on the first prompt, the large language model performs preliminary recognition and extracts key information from the vehicle cabin image data, then outputs an image recognition result. The user then inputs text data and a second prompt based on the image recognition result. The large language model then further optimizes the image recognition result based on the second prompt to obtain vehicle model information.

[0055] For example, the first prompt instruction is as follows: (1) Identify all visible physical components in the image and give them standard names (e.g., “driver’s seat,” “rear seats,” etc.); (2) Determine the type of each component (seat, control panel, guide rail, etc.); (3) Describe the relative positional relationships between components (e.g., “in front,” “on the left,” “in parallel,” “in contact,” “with a gap,” etc.); (4) If there is dimension or angle information in the image, extract it; (5) If there is a movable structure (such as a slide) in the image, please determine its possible direction of movement.

[0056] After receiving the first prompt and the image data of the car cabin, the large language model performs a parsing process to obtain preliminary second vehicle model information as follows: Image recognition results { "entities":[ {"name":"Driver's seat","type":"Seat"}, {"name":"Rear seats","type":"seats"} ], "spatial_relations":[ {"from":"Driver's seat","to":"Rear seats","relation":"Front"} ] } The user then enters text data and a second prompt instruction in the input box of the large language model to optimize the above image recognition results and obtain vehicle model information.

[0057] It should be noted that, in order to improve the recognition accuracy of the large language model, the vehicle component multimodal data is preprocessed before being input into the large language model.

[0058] For example, the vehicle component image data is pre-processed by denoising, contrast enhancement, and size unification, and the vehicle component text data is processed by removing invalid characters and performing block processing.

[0059] Step 103: Generate a vehicle component state transition path set and a linear temporal logic rule set based on the vehicle model information; Step 104 : Generate an executable interactive logic verification file based on the vehicle component state transition path set and the linear temporal logic rule set.

[0060] The vehicle component state transition path refers to a finite set of steps consisting of the state sequence of each vehicle component (such as the main driver's seat, rear seats, etc.) and its triggering actions, which describes the evolution process of the vehicle component from the initial state to the target state.

[0061] The key elements of a vehicle component's state transition path include state, action, and path. A state represents the value of a vehicle component's attribute at a specific moment (e.g., driver's seat position = 0 cm), an action represents the operation that triggers a state transition (e.g., driver's seat moves forward 5 cm), and a path represents the alternating sequence of states and actions.

[0062] Exemplarily, the state transition path of the main driver's seat is expressed as: main driver's seat position = 0 cm → (moved forward 5 cm) → main driver's seat position = 5 cm.

[0063] It should be known that since the vehicle model information is a structured digital description of the vehicle and its components, in addition to describing the seat components in the above example, the vehicle model information can also describe information on vehicle components such as the steering wheel, gear lever, air conditioner, window, brake pedal, accelerator pedal, etc.

[0064] Therefore, in addition to describing component state transitions involving positional movement (e.g., seat movement), vehicle component state transition paths can also describe component state transitions involving on / off states. For example, the state transition path for the air conditioner is represented as: air conditioner is off (OFF) → (air conditioner is on) → air conditioner is on (ON). Component state transitions involving the degree of opening or closing can also be described, for example: the volume control button is in its initial position (multimedia volume is 0) → (rotate the volume control button to the right to the target position) → the volume control button is in its target position (multimedia volume is increased to 10).

[0065] The vehicle component state transition path set includes at least one vehicle component state transition path.

[0066] Linear temporal logic rules are a formal logic language used to describe the temporal behavior of vehicle components. They constrain the properties that vehicle components must satisfy on all possible paths through logical operators and temporal operators.

[0067] Exemplarily, the linear temporal logic rule regarding the main driver's seat is: G (main driver's seat position ≥ 0), indicating that the main driver's seat position is always greater than or equal to 0.

[0068] The linear temporal logic rule set includes at least one linear temporal logic rule.

[0069] An executable interactive logic verification file refers to a digital file (machine-readable code) used to define, verify, and execute interactive logic verification. It usually exists in the form of code or script and can be parsed and executed by corresponding formal verification tools.

[0070] For example, the formal verification tool is NuSMV (New Symbolic Model Verifier), and the executable interactive logic verification file is an executable file with the suffix ".smv".

[0071] It should be noted that the formalization tool can also be SPIN (Simple Promela Interpreter). In this case, the executable interactive logic verification file is a Promela language file, that is, an executable file with the suffix ".pml".

[0072] During the specific implementation process, after obtaining the vehicle model information, the user enters the prompt instructions of "generate vehicle component state transition path based on vehicle model information" and / or "generate linear temporal logic rules based on vehicle model information" in the input box of the large language model, so that the large language model analyzes and infers the vehicle model information based on the aforementioned prompt instructions to generate a set of vehicle component state transition paths and a set of linear temporal logic rules.

[0073] Then, after the user enters the prompt instruction "Convert the vehicle component state transition path set and the linear temporal logic rule set into an executable interactive logic verification file that can be recognized by the target formal verification tool" in the input box of the large language model, the large language model automatically generates an executable interactive logic verification file that meets the requirements.

[0074] It should be noted that in addition to the method of using a large language model to automatically generate an executable interactive logic verification file based on the vehicle component state transition path set and the linear temporal logic rule set, after obtaining the vehicle component state transition path set and the linear temporal logic rule set, the user manually writes the executable interactive logic verification file based on the vehicle component state transition path set and the linear temporal logic rule set.

[0075] The specific process of manually writing an executable interactive logic verification file is as follows: based on the format requirements of the executable file required by the target formal verification tool, define variables and write state sequences according to the vehicle component state transition paths in the vehicle component state transition path set to perform state modeling; write LTL formulas according to the linear temporal logic rules in the linear temporal logic rule set to introduce the linear temporal logic rules into the executable file, thereby generating an executable interactive logic verification file.

[0076] The embodiments of the present application analyze multimodal data to obtain vehicle model information. Because multimodal data includes vehicle-related data in different forms, the vehicle model information obtained from the multimodal data is more complete and accurate, laying the foundation for the subsequent generation of a set of vehicle component state transition paths and a set of linear temporal logic rules. Furthermore, all possible vehicle component state transition paths and linear temporal rules within the vehicle model information are automatically generated, improving the accuracy of verifying the interaction logic between vehicle components.

[0077] In some embodiments, the vehicle model information includes a first component type and a first component attribute parameter; the method of generating a vehicle component state transition path set includes: determining the executable actions and movement range of the vehicle components corresponding to the first component type based on the first component attribute parameter; and generating the vehicle component state transition path set of the vehicle based on the executable actions and movement range of each vehicle component.

[0078] The first component attribute parameter is a set of numerical values ​​used to quantitatively describe the static characteristics and dynamic behaviors of the component type.

[0079] The first component attribute parameters include physical parameters, motion parameters, and state parameters. Physical parameters include dimensions (such as length, width, height, thickness, etc.) and material properties; motion parameters include movement range and speed; and state parameters include whether a vehicle component is unfolded or folded.

[0080] Exemplarily, the third vehicle model information is as follows: { "entities":[ { "name":"Driver's seat", "type":"seat", "width":"50cm", "height":"1m", "movement_range":"±10cm" }, { "name":"Rear seats", "type": "seat", "movement_range": "±15cm", "foldable": true } ], "spatial_relations": [ {"from": "driver seat", "to": "back seat", "relation": "in front of"} ] } wherein the first component type includes "driver seat" and "back seat"; the first component attribute parameters of the driver seat include the height of the driver seat 1 meter, the width of the driver seat 0.5 meters, and the movement range [-10 centimeters, 10 centimeters]. The first component attribute parameters of the back seat include the movement range [-15 centimeters, 15 centimeters] and the foldable of the back seat.

[0081] In the specific implementation process, the prompt instruction for generating the vehicle component state transition path according to the vehicle model information is specifically: Please perform the following tasks: 1. Analyze the executable actions (such as moving, rotating, folding, etc.) of each vehicle component; 2. Enumerate all reasonable single-step or multi-step operation paths; 3. Each path should include component name, action type, and parameter value (such as distance, angle); 4. Output at least 5 different candidate paths.

[0082] The user inputs the above prompt instruction in the input box of the large language model, and uses the reasoning ability of the large language model to determine the executable actions and movement ranges of the vehicle components corresponding to the first component attribute parameters based on the provided vehicle model information.

[0083] Taking the above third vehicle model information example as an example, the driver seat has the attribute "movement_range: ±10cm", and its executable actions are moving forward and moving backward, with a movement distance of ±10cm.

[0084] The back seat has the attributes "movement_range: ±15cm" and "foldable: true", and its executable actions are moving forward, moving backward, and folding, with a movement distance of ±15cm.

[0085] Based on the executable actions and movement ranges of each vehicle component, a set of vehicle component state transition paths of the vehicle is generated.

[0086] Taking the above-mentioned third vehicle model information example as an example, all reasonable single-step operation paths are enumerated, including the main driver's seat moving forward 5cm, the main driver's seat moving backward 5cm, the rear seat moving forward 5cm, the rear seat moving backward 5cm, and the rear seat folding.

[0087] The set of vehicle component state transition paths is formally expressed as follows: [ { "path_id":1, / / Vehicle component status transfer path 1 "steps":[ {"component":"Driver's seat","action":"Move forward","distance":"5cm"}, {"component":"Rear seat","action":"Move backward","distance":"5cm"} ] }, { "path_id":2, / / Vehicle component status transfer path 2 "steps":[ {"component":"Rear seats","action":"Fold"} ] } ] Based on the third vehicle model information above, it can be seen that the following parameters are also included: "spatial_relations":[ {"from":"Driver's seat","to":"Rear seats","relation":"Front"} ].

[0088] "spatial_relations" represent the geometric or logical interactions between components within a specific space, such as a vehicle interior, a mechanical system, or a virtual environment. These relationships describe how components exist, move, or interact relative to each other.

[0089] Therefore, the vehicle model information also includes first component interaction constraint parameters, where the first component interaction constraint parameters are used to quantitatively describe the geometric or logical interaction relationship between different components in a specific space (such as vehicle interior, mechanical system or virtual environment).

[0090] The first component interaction constraint parameters include static and dynamic parameters. Static parameters define the relative positions of components in their initial or fixed states, such as the rear seats being 30 cm behind the driver's seat. Dynamic parameters define how components interact with other components during motion, such as the seats automatically moving back 5 cm when a door is opened or the air conditioning not turning on when a window is opened.

[0091] Therefore, in some embodiments, a set of vehicle component state transition paths for the vehicle is generated based on the executable actions and movement range of each vehicle component, including: determining the interaction constraint relationship between the vehicle components based on the first component interaction constraint parameter; and generating the set of vehicle component state transition paths for the vehicle according to the interaction constraint relationship, the executable actions and movement range of each vehicle component.

[0092] On the basis of the above-mentioned single-step operation path, reasonable combinations are made according to the interaction constraints of the vehicle components to further generate a set of vehicle component state transition paths.

[0093] For example, after the main driver's seat moves forward 5 cm, the rear seats move backward 5 cm.

[0094] In this embodiment, the executable actions and movement ranges of corresponding vehicle components are automatically analyzed based on the first component attribute parameters, ensuring that no single possible operation path for the vehicle component is missed, providing a foundation for subsequent more complex operation path combinations. Furthermore, based on the interaction constraints between vehicle components, reasonable multi-step combinations are performed to plan possible complex operation path sequences in advance, covering all possible scenarios of vehicle component state transition paths, thereby improving the accuracy of interaction logic verification between vehicle components.

[0095] In some embodiments, the vehicle model information includes a second component type, second component attribute parameters, and second component interaction constraint parameters; the method of generating a linear temporal logic rule set includes: determining the physical constraints of the vehicle components corresponding to the second component type based on the second component type and the second component attribute parameters; determining the interaction logic between the vehicle components based on the physical constraints and the second component interaction constraint parameters; and generating a linear temporal logic rule set based on the physical constraints of each vehicle component and the interaction logic between each vehicle component.

[0096] For the specific interpretations of the second component type, second component attribute parameters, and second component interaction constraint parameters, please refer to the above specific interpretations of the first component type, first component attribute parameters, and first component interaction constraint parameters, which will not be repeated here.

[0097] The physical constraints are determined by the second component type and the second component property parameters, and describe the geometry and motion restrictions of the vehicle components.

[0098] Interaction logic refers to the dynamic behavior rules derived from physical constraints and interaction constraint parameters that describe how vehicle components work together safely and functionally correctly.

[0099] In the specific implementation process, a specific example of a prompt instruction for generating a linear temporal logic rule based on vehicle model information is as follows: Perform the following tasks: 1. Analyze the physical constraints and interaction possibilities in the spatial model; 2. Generate at least three linear temporal logic specifications to ensure the security and legitimacy of the operation path; 3. Provide natural language explanation for each specification; 4. Use standard LTL syntax without adding additional explanations.

[0100] The user enters the above prompt in the input box of the large language model. Leveraging the large language model's reasoning capabilities, the model determines the physical constraints of the vehicle components corresponding to the second component type based on the provided vehicle model information, the second component type, and the second component attribute parameters. Based on the physical constraints and the second component interaction constraint parameters, the model determines the interaction logic between the vehicle components. Based on the physical constraints of each vehicle component and the interaction logic between each vehicle component, a set of linear temporal logic rules is generated.

[0101] Taking the third vehicle model information above as an example, the generated linear temporal logic rule set is as follows: [ { "ltl_formula":"G(rear seat.x>driver's seat.x)", "explanation":"The rear seats must never exceed the X-axis position of the main driver's seat to avoid collisions." }, { "ltl_formula":"F(Driver's seat.Move → Rear seats.Fold)", If the driver's seat is moving, the rear seats must not be folded. } ].

[0102] It should be noted that the content of the prompt instruction input by the above user is only used as an example to illustrate the embodiment of the present application. In actual application, it can be adaptively adjusted according to actual conditions.

[0103] The embodiment of the present application automatically generates a set of linear temporal logic rules based on the physical constraints of each vehicle component and the interaction logic between each vehicle component, which not only reduces the workload of manually writing rules, but also improves the accuracy and completeness of the generated linear temporal logic rules, thereby improving the accuracy of the interaction logic verification between vehicle components.

[0104] In some embodiments, after analyzing the multimodal data of vehicle components and obtaining vehicle model information, the method also includes: obtaining user instructions; modifying the vehicle model information based on the user instructions to obtain new vehicle model information; and generating a vehicle component state transition path set and a linear temporal logic rule set according to the vehicle model information, including: generating a vehicle component state transition path set and a linear temporal logic rule set according to the new vehicle model information.

[0105] During the specific implementation process, considering that there are certain errors in the model output, the output results can be corrected based on user instructions to improve the accuracy of the vehicle space model.

[0106] The embodiment of the present application improves the accuracy of the vehicle model information by correcting the vehicle model information obtained based on multimodal data analysis according to user instructions, thereby improving the accuracy of the generated vehicle component state transition path set and linear temporal logic rule set.

[0107] Figure 3 A flow chart of a vehicle component interaction logic verification method provided in an embodiment of the present application is shown as follows: Figure 3 As shown, the method includes: Step 301: Read and parse an executable interactive logic verification file to obtain a set of vehicle component state transition paths and a set of linear temporal logic rules; wherein the executable interactive logic verification file is generated by the method of any one of the above embodiments; Step 302, constructing a state space based on a set of vehicle component state transition paths; Step 303 , traversing the state space according to the linear temporal logic rule set, verifying the vehicle component state transition paths in the vehicle component state transition path set, and obtaining a verification result.

[0108] During implementation, the interaction logic of vehicle components is verified using formal verification tools. These tools can be deployed on a terminal device (also known as an electronic device) or a server. Terminal devices can include smartphones, tablets, computers, and personal digital assistants (PDAs). Servers can be application servers or web servers.

[0109] To facilitate understanding of the technical solution provided by the embodiments of the present application, a formal verification tool is deployed in a terminal device and the method provided by the embodiments of the present application is introduced.

[0110] Formal verification tools include NuSMV, SPIN, etc.

[0111] The specific interpretations of the vehicle component state transition path set and the linear temporal logic rule set can be found in the above embodiments and will not be repeated here.

[0112] Import the executable interactive logic verification file into the corresponding formal verification tool. The formal verification tool has a built-in parser that performs lexical analysis and syntactic analysis on the input code. Lexical analysis primarily breaks down character sequences in the code into a series of words (such as keywords, identifiers, operators, etc.). Syntactic analysis verifies the correctness of the code structure according to the grammatical rules specified by the software, parsing the code into a data structure such as an abstract syntax tree (AST) for subsequent processing.

[0113] Based on the parsed code, the formal verification tool constructs a state space based on the set of vehicle component state transition paths to represent the states and behaviors of the object being verified. The tool then initiates a state space search, traversing the state space according to a set of linear temporal logic rules to explore all possible states and transition relationships within the state space. This exhaustive search examines any violations of the prescribed linear temporal logic rules within the vehicle component state transition paths.

[0114] For example, when checking the linear temporal logic property that "a global system variable A is always true," the software will follow every possible execution path in the state space, sequentially checking whether the value of variable A is true in each state. If this condition is met in all possible states and paths, the verification passes. If any violation is found, the corresponding counterexample is recorded and the verification fails.

[0115] During the state space search process, formal verification tools use various verification algorithms to assist in judgment. Common model checking algorithms, such as depth-first search (DFS) and breadth-first search (BFS), are used to systematically traverse the state space, ensuring that no possible cases are missed.

[0116] At the same time, related data structures and algorithms such as binary decision diagrams (BDDs) will be used to efficiently represent and operate state spaces and logical expressions, thereby improving the efficiency and accuracy of the verification process.

[0117] After the completion of the state space search and the execution of the related verification algorithm, the formal verification tool makes a decision based on the verification results. If all the verified properties are satisfied, that is, the vehicle component state transition path meets the expected linear temporal logic rules, the formal verification tool outputs a verification pass result, indicating that the current system is correct under the given operation path and logic rules.

[0118] If there is a case that does not meet the requirements, the formal verification tool outputs the specific path, state, and other information that violates the rules as counterexamples, so that users can understand the problematic part of the vehicle component state transition path and then modify and optimize the design. The modified vehicle component state transition path is verified again by the formal verification tool to determine whether it meets the requirements.

[0119] For example, the formal verification tool NuSMV processes the.smv file as follows: File reading and parsing Reading file content: By executing the read_model command, the text content in the.smv file is read.

[0120] Lexical and syntax analysis: The read content is analyzed for lexical and syntax, which is decomposed into a series of words and syntax structures, checks whether the file syntax is correct, and constructs internal representations such as abstract syntax tree (AST).

[0121] State space generation: The model is constructed according to the vehicle component state transition path, and different partition methods and strategies are used to generate all possible states and their transition relationships of the system, forming a state space.

[0122] Verification property conversion: The verification properties (such as LTL expressed properties) in the input file are converted and processed.

[0123] Verification process: The user selects the verification method (BDD-based model checking or SAT-based model checking) according to the needs. In BDD-based model checking, a BDD-based representation of the finite state machine is constructed; in SAT-based model checking, an internal representation of the model is constructed based on the reduced Boolean circuit (RBC).

[0124] According to the selected verification method, starting from the initial state, the state space is traversed to determine all reachable states. Then, according to the verification properties, it is checked whether these reachable states meet the specified properties.

[0125] Result output and feedback: If all verification properties are true in the state space, the verification success result is output; if there are properties that are not true, the counterexample is output, indicating the specific state sequence or path in the vehicle component state transition path set that violates the properties.

[0126] The embodiment of the application verifies the executable interactive logic verification file, and since the executable interactive logic verification file includes all possible vehicle component state transition paths and linear timing rules, the accuracy of the verification result is improved.

[0127] Figure 4 A structural schematic diagram of a vehicle component interactive logic verification file generation device provided by the embodiment of the application is shown in FIG. 1. Figure 4 As shown in the figure, the device includes an acquisition module 401, an analysis module 402, a first generation module 403 and a second generation module 404. The acquisition module 401 is configured to acquire vehicle component multi-modal data; the analysis module 402 is configured to analyze the vehicle component multi-modal data to obtain vehicle model information; the first generation module 403 is configured to generate a vehicle component state transition path set and a linear timing logic rule set according to the vehicle model information; and the second generation module 404 is configured to generate an executable interactive logic verification file based on the vehicle component state transition path set and the linear timing logic rule set.

[0128] On the basis of the above embodiment, the vehicle model information includes a first component type and a first component attribute parameter; the first generation module 403 is specifically configured to determine executable actions and movement ranges of vehicle components corresponding to the first component type according to the first component attribute parameter; and the vehicle component state transition path set of the vehicle is generated based on the executable actions and movement ranges of each vehicle component.

[0129] On the basis of the above embodiment, the vehicle model information further includes a first component interaction constraint parameter; the first generation module 403 is specifically configured to determine interaction constraint relationships between vehicle components based on the first component interaction constraint parameter; and the vehicle component state transition path set of the vehicle is generated according to the interaction constraint relationships, the executable actions and movement ranges of each vehicle component.

[0130] On the basis of the above embodiment, the vehicle model information includes a second component type, a second component attribute parameter and a second component interaction constraint parameter; the first generation module 403 is specifically configured to determine physical constraints of vehicle components corresponding to the second component type according to the second component type and the second component attribute parameter; to determine interactive logic between vehicle components according to the physical constraints and the second component interaction constraint parameter; and to generate the linear timing logic rule set based on the physical constraints of each vehicle component and the interactive logic between each vehicle component.

[0131] On the basis of the above-mentioned embodiments, the device further comprises a correction module configured to obtain a user instruction; correct the vehicle model information based on the user instruction to obtain new vehicle model information; and generate a vehicle component state transition path set and a linear temporal logic rule set based on the vehicle model information, including: generating the vehicle component state transition path set and the linear temporal logic rule set based on the new vehicle model information.

[0132] It should be understood that the device corresponds to the vehicle component interaction logic verification file generation method embodiment described above, and can perform each step involved in the above method embodiments. The specific functions of the device can be referred to the description in the above, and the detailed description is appropriately omitted here to avoid repetition. The device includes at least one software function module that can be stored in the memory in the form of software or firmware or solidified in the operating system (OS) of the device.

[0133] Figure 5 A structural diagram of a vehicle component interaction logic verification device provided by an embodiment of the present application is shown in FIG. 1. Figure 5 As shown in FIG. 1, the device comprises a reading module 501, a construction module 502, and a verification module 503; wherein, The reading module 501 is configured to read and parse an executable interaction logic verification file to obtain a vehicle component state transition path set and a linear temporal logic rule set; wherein, the executable interaction logic verification file is generated by the method of any one of the first aspect embodiments; the construction module 502 is configured to construct a state space based on the vehicle component state transition path set; and the verification module 503 is configured to traverse the state space according to the linear temporal logic rule set, verify the vehicle component state transition path in the vehicle component state transition path set, and obtain a verification result.

[0134] It should be understood that the device corresponds to the vehicle component interaction logic verification method embodiment described above, and can perform each step involved in the above method embodiments. The specific functions of the device can be referred to the description in the above, and the detailed description is appropriately omitted here to avoid repetition. The device includes at least one software function module that can be stored in the memory in the form of software or firmware or solidified in the operating system (OS) of the device.

[0135] Figure 6 A structural diagram of an electronic device provided by an embodiment of the present application is shown in FIG. 1. Figure 6As shown, the electronic device includes a processor 601 (processor), a memory 602 (memory), and a bus 603; wherein the processor 601 and the memory 602 communicate with each other via the bus 603. The processor 601 is used to call program instructions in the memory 602 to execute the methods provided by the above-mentioned method embodiments.

[0136] Processor 601 can be an integrated circuit chip with signal processing capabilities. The processor 601 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor.

[0137] The memory 602 can be implemented by any type of volatile or non-volatile memory device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0138] An embodiment of the present application provides a computer program product, including: computer program instructions, which, when executed by a processor, execute the methods provided by the above-mentioned method embodiments.

[0139] An embodiment of the present application provides a computer-readable storage medium, including: computer program instructions stored on the computer-readable storage medium, and the computer program instructions execute the methods provided by the above-mentioned method embodiments when executed by a processor.

[0140] Computer-readable storage media can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0141] In the embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interface, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0142] In addition, the units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0143] Furthermore, the functional modules in each embodiment of the present application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0144] The above merely provides an example of the present application, and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for generating a vehicle component interaction logic verification file, characterized in that: The method comprises: Acquire multimodal data of vehicle components; Analyzing the multimodal data of the vehicle components to obtain vehicle model information; generating a vehicle component state transition path set and a linear temporal logic rule set respectively according to the vehicle model information; An executable interactive logic verification file is generated based on the vehicle component state transition path set and the linear temporal logic rule set.

2. The method according to claim 1, characterized in that in, The vehicle model information includes a first component type and a first component attribute parameter; The method of generating the vehicle component state transition path set includes: determining, based on the first component attribute parameters, an executable action and a movement range of a vehicle component corresponding to the first component type; A set of vehicle component state transition paths of the vehicle is generated based on the executable actions and the movement range of each of the vehicle components.

3. The method according to claim 2, characterized in that in, The vehicle model information also includes first component interaction constraint parameters; The step of generating a set of vehicle component state transition paths of the vehicle based on the executable action and the movement range of each vehicle component includes: determining an interaction constraint relationship between the vehicle components based on the first component interaction constraint parameters; A set of vehicle component state transition paths of the vehicle is generated according to the interaction constraint relationship, the executable action and the movement range of each vehicle component.

4. The method according to claim 1, wherein in, The vehicle model information includes a second component type, a second component attribute parameter, and a second component interaction constraint parameter; The method of generating the linear temporal logic rule set includes: determining a physical constraint of a vehicle component corresponding to the second component type according to the second component type and the second component attribute parameter; determining an interaction logic between the vehicle components according to the physical constraint and the second component interaction constraint parameter; The linear temporal logic rule set is generated based on the physical constraints of each of the vehicle components and the interaction logic between each of the vehicle components.

5. The method according to any one of claims 1 to 4, characterized in that: After analyzing the multimodal data of the vehicle components to obtain vehicle model information, the method further includes: Get user instructions; Modifying the vehicle model information based on the user instruction to obtain new vehicle model information; Generating a set of vehicle component state transition paths and a set of linear temporal logic rules based on the vehicle model information includes: According to the new vehicle model information, a vehicle component state transition path set and a linear temporal logic rule set are generated respectively.

6. A vehicle component interaction logic verification method, characterized in that: The method comprises: Reading and parsing an executable interactive logic verification file to obtain a vehicle component state transition path set and a linear temporal logic rule set; wherein the executable interactive logic verification file is generated by the vehicle component interactive logic verification file generation method according to any one of claims 1 to 5; constructing a state space based on the set of vehicle component state transition paths; The state space is traversed according to the linear temporal logic rule set, and the vehicle component state transition paths in the vehicle component state transition path set are verified to obtain a verification result.

7. A vehicle component interaction logic verification file generation device, characterized in that: The device comprises: An acquisition module, used to acquire multimodal data of vehicle components; An analysis module, configured to analyze the multimodal data of the vehicle components to obtain vehicle model information; A first generating module is configured to generate a vehicle component state transition path set and a linear temporal logic rule set according to the vehicle model information; The second generating module is used to generate an executable interactive logic verification file based on the vehicle component state transition path set and the linear temporal logic rule set.

8. An electronic device, characterized in that: include: A processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and when the machine-readable instructions are executed by the processor, the method according to any one of claims 1 to 6 is performed.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the method according to any one of claims 1 to 6 is executed.

10. A computer program product, characterized in that include: Computer program instructions, which, when executed by a processor, perform the method according to any one of claims 1 to 6.

Images