Automobile automatic driving system modeling method based on MBSE

Abstract

The invention belongs to the technical field of complex system design and model-based system engineering, and particularly relates to an MBSE-based automobile automatic driving system modeling method. The specific process of the method comprises the first step of subsystem-level demand analysis, the second step of subsystem-level functional architecture analysis, the third step of component-level functional analysis, the fourth step of component-level logic architecture design, the fifth step of subsystem-level logic architecture design, the sixth step of component-level physical architecture design and the seventh step of subsystem-level physical architecture design. The method supports full-process modeling of the automatic driving system, and ensures the consistency and integrity of the model on each level.

Description

Technical Field

[0001] The present invention belongs to the technical field of complex system design and model-based system engineering, and specifically relates to a modeling method for an automobile automatic driving system based on MBSE. Background Art

[0002] The automotive autonomous driving system is a complex and usually distributed system. They are highly integrated with other vehicle systems and process information from the vehicle itself and the vehicle environment. The design of the autonomous driving system involves different levels such as the system level, subsystem level, and component level, multiple disciplines such as mechanical engineering, control engineering, electronic engineering, and software engineering, as well as the cross-integration of multiple fields such as perception, decision-making, control, communication and interconnection, human-computer interaction, safety and reliability. It has the characteristics of cross-level, multi-disciplinary, and multi-field. At the same time, the autonomous driving system is a safety-related system, and its design and operation are directly related to personal safety and property safety. It has the characteristics of no-man-in-the-loop and strong autonomy. In the design process, possible situations must be fully and comprehensively considered. Therefore, if the system requirements are incomplete during the design process and the requirements cannot be traced back to the corresponding design, it may lead to design defects or omissions, resulting in catastrophic consequences and endangering road traffic safety.

[0003] With the rapid development of autonomous driving technology, the automotive industry is undergoing a profound transformation. Traditional document-based design methods can no longer cope with the complexity and diversity challenges in the development of modern autonomous driving systems. Model-Based Systems Engineering (MBSE) methodology came into being. MBSE provides full life cycle support from requirements analysis, system design, verification and validation to subsequent stages by building a system model. It effectively solves many problems of traditional document-based methods in complexity management, cross-disciplinary collaboration, and requirements change management, and improves the design efficiency of autonomous driving systems.

[0004] Existing MBSE modeling methods are usually oriented towards a specific modeling language (such as SysML). However, in the design process of complex equipment systems, a single modeling language is often unable to meet the multi-domain and multi-level design of complex equipment systems.

[0005]

[0006] The Chinese invention patent with application publication number CN115826452A discloses a SysML-based intelligent driving system modeling method, device, equipment and medium. The method analyzes the functional requirements of the intelligent driving system to obtain the operating elements of the intelligent driving system; describes the operating elements of the intelligent driving system using natural language; and constructs the intelligent driving system model based on the operating elements of the intelligent driving system described in natural language and the SysML modeling language. The method applies MBSE combined with the SysML modeling language standard to design the intelligent driving system, which can replace document design with model-driven method and improve system design efficiency. However, the modeling method only involves functional requirement analysis and operation analysis, lacks clear logical design steps, and may lead to logical inconsistency in the function implementation process. And the lack of physical implementation details affects the implementation effect of the final system.

[0007] By analyzing the above existing modeling methods for autonomous driving systems, we can find the following problems that need to be solved in the current autonomous driving systems:

[0008] (1) Existing MBSE modeling methods are usually oriented towards a specific modeling language

[0009] Most existing MBSE modeling methods focus on a specific modeling language, such as SysML, UML, or a specific DSML. This single language dependency limits the versatility and flexibility of the method. When faced with complex and diverse requirements in the design of autonomous driving systems, a single language may not be able to fully and efficiently express all necessary information; in addition, different modeling languages ​​often require different tool support, which not only increases the learning cost, but also makes it difficult to reuse models.

[0010] (2) Lack of modeling methods for specific domains of autonomous driving systems

[0011] Several existing modeling methodologies, including OPM, Harmony SE, OOSEM, MagicGrid, and Arcadia, provide a systematic modeling framework, but they are all general methodologies and are not optimized for the specific needs and characteristics of autonomous driving systems. Summary of the invention

[0012] In view of this, the present invention proposes a modeling method for an automobile autonomous driving system based on MBSE, which supports full-process modeling of the autonomous driving system and ensures the consistency and integrity of the model at all levels.

[0013] The technical solution for implementing the present invention is as follows:

[0014] A modeling method for an autonomous driving system based on MBSE. The specific process is as follows:

[0015] Step 1: Subsystem-level requirements analysis: Build an ADS requirements model based on expression rules and constraints, and incorporate the external ADS requirements model; analyze ADS features based on the requirements model and build an ADS feature model; identify external entities and interaction processes based on the ADS feature model, and define the ADS system boundaries and use cases;

[0016] Step 2: Subsystem-level functional architecture analysis: Analyze the behavior and interaction mode of ADS features in different functional scenarios; Based on the behavior and interaction mode, refine the ADS requirements to obtain the refined ADS requirement model and ADS functional architecture model, and build a traceability relationship from requirements to functions;

[0017] Step 3: Component-level functional analysis: According to the ADS functional architecture model, determine the functional activities of each component in the subsystem and obtain the component activity model;

[0018] Step 4: Component-level logical architecture design: Based on the component activity model, determine the logical composition of each component and generate a component logical composition model; determine the data exchange and information transmission within each component, build a component logical connection model, and then obtain a component logical architecture model;

[0019] Step 5: Subsystem-level logical architecture design: Based on the component logical architecture model, the functions of executing each feature of ADS are allocated to the logical components to generate the logical composition model of ADS; the data exchange and information transmission between the components are determined, and the interface model between the logical components of ADS is constructed to obtain the subsystem-level logical architecture model;

[0020] Step 6: Component-level physical architecture design: perform component-level physical architecture design based on the component logical architecture model;

[0021] Step 7: Subsystem-level physical architecture design: Complete the subsystem-level physical architecture design based on the subsystem-level logical architecture model.

[0022] Furthermore, in step 1 of the present invention, a SysML use case diagram is used to define the ADS system boundary and use cases.

[0023] Furthermore, in step one of the present invention, an ADS feature model is constructed based on the vehicle feature model of EAST-ADL.

[0024] Furthermore, in step 2 of the present invention, a BPMN flowchart is used to analyze the workflow and behavior of each feature under specific conditions, and the behavior and interaction mode of the ADS feature in different scenarios.

[0025] Furthermore, in step 2 of the present invention, a SysML module definition diagram is used to describe the ADS functional architecture.

[0026] Furthermore, in step 4 of the present invention, a SysML internal module definition diagram is used to describe the interface model inside each logical component of the ADS.

[0027] Furthermore, in step five of the present invention, a SysML internal module definition diagram is used to describe the interface model between various logical components of the ADS.

[0028] Furthermore, in step five of the present invention, the SysML activity diagram and swimlane are used to perform the process of allocating functions under each feature of the ADS to logical components.

[0029] Furthermore, step six of the present invention, component-level physical architecture design includes the design of component physical composition, component physical interface and component physical parameter model.

[0030] Furthermore, the subsystem-level physical architecture design in step seven of the present invention includes the design of ADS physical composition, ADS physical interface and ADS physical parameter selection.

[0031] Beneficial effects:

[0032] First, at the subsystem level, conduct demand analysis and functional architecture analysis, and at the component level, conduct functional analysis and logical architecture design, and then conduct component-level logical architecture design. The above process supports cross-domain and cross-level modeling and improves system design efficiency.

[0033] Second, in the process of modeling the car's autonomous driving system, this method is compatible with different modeling languages ​​such as SysML, EAST-ADL, and BPMN, as well as the expression rules and constraints of requirement integrity, and supports cross-domain team collaboration. This feature significantly improves the efficiency of autonomous driving system design and reduces the communication and conversion costs caused by using different tools and languages.

[0034] Third, this method is based on the MBSE-based automotive autonomous driving system modeling method, which can support the full-process autonomous driving system modeling from requirements, functions, logic to physics, ensuring the consistency and integrity of the model at all levels. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0036] Figure 1It is a flow chart of the MBSE-based vehicle automatic driving system modeling method of the present invention;

[0037] Figure 2 It is a schematic diagram of the ADS demand model;

[0038] Figure 3 It is a schematic diagram of the ADS feature model;

[0039] Figure 4 A schematic diagram of the interaction between external entities and use cases;

[0040] Figure 5 A schematic diagram of a functional scenario model for vehicle control features;

[0041] Figure 6 This is a schematic diagram of the refined ADS demand model;

[0042] Figure 7 It is a schematic diagram of the functional architecture model;

[0043] Figure 8 For traceability relationship forms;

[0044] Fig. 9 It is a schematic diagram of the component-level logical architecture model;

[0045] Fig.10 Assign the functions under the vehicle control feature to the logic diagram;

[0046] Fig.11 It is a schematic diagram of the logic composition model;

[0047] Fig.12 It is a schematic diagram of the interface model between logical components;

[0048] Fig.13 It is a schematic diagram of the ADS logical architecture model;

[0049] Fig.14 It is a schematic diagram of the physical composition model of the vehicle control system;

[0050] Fig.15 A schematic diagram of the interface interaction relationship between ADS physical components;

[0051] Fig.16 Schematic diagram of the ADS physical parameter model. DETAILED DESCRIPTION

[0052] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0053] It should be noted that the following embodiments and features in the embodiments may be combined with each other in the absence of conflict; and, based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making any creative work are within the scope of protection of the present disclosure.

[0054] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein may be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on the present disclosure, it should be understood by those skilled in the art that an aspect described herein may be implemented independently of any other aspect, and two or more of these aspects may be combined in various ways. For example, any number of aspects described herein may be used to implement the device and / or practice the method. In addition, other structures and / or functionalities other than one or more of the aspects described herein may be used to implement this device and / or practice this method.

[0055] Mainstream modeling methods at home and abroad: Harmony SE, OOSEM, MagicGrid and Arcadia. By analyzing the specific modeling process of these four methods, it can be found that they all follow a set of requirements-function-logic-physical (RFLP) analysis process. The RFLP method is a top-down design method, which starts from high-level requirements and gradually refines downward, helping to ensure that the design plan is consistent with the original requirements and goals, thereby supporting the conversion into specific physical implementation. Therefore, the present invention adopts the RFLP modeling method to design the automobile ADS. Under the guidance of the RFLP method process, the ADS requirement model is expressed based on the requirement integrity expression rules and constraints, and the ADS design process model is expressed using SysML, BPMN, and EAST-ADL modeling languages, and the construction of models based on the above-mentioned modeling methods and modeling languages ​​is implemented in the requirement modeling and evaluation system. The embodiment of the present application provides a modeling method for an automobile automatic driving system based on MBSE, and the specific process is as follows: Figure 1 shown.

[0056] Step 1: Subsystem Level Requirements Analysis

[0057] 1) Build an ADS demand model based on expression rules and constraints, and incorporate external ADS demand models

[0058] According to the National Guidelines for the Construction of the Internet of Vehicles Industry Standard System (Intelligent Connected Vehicles) (2023 Edition) and ADS-related standards DB4403 / T 359.1-2023, DB4403 / T 357—2023, ISO26262, etc., the ADS demand model is constructed based on demand expression rules and constraints in response to the needs of ADS stakeholders, and external related ADS needs are integrated, such as Figure 2 shown.

[0059] 2) Analyze ADS features according to the demand model and build an ADS feature model

[0060] Based on the complete demand model, the analysis obtained seven ADS features: vehicle control, perception and perception fusion, high-definition map and positioning, path planning, decision and control, communication, safety and emergency handling. The ADS feature model is constructed based on the vehicle feature model of EAST-ADL, such as Figure 3 shown.

[0061] 3) Based on the ADS feature model, identify external entities and interaction processes, and determine the ADS system boundaries and use cases;

[0062] In order to clarify the scope of the system and identify external entities and interaction processes, SysML use case diagrams are used to define system boundaries and use cases. In ADS design, external entities include other road users (other vehicles, pedestrians), traffic signals, etc. Based on ADS features, corresponding use cases are defined, and the interaction relationship between external entities and use cases is established, such as Figure 4 shown.

[0063] Step 2: Subsystem-level functional architecture analysis

[0064] 4) Analyze the behavior and interaction patterns of ADS features in different scenarios according to different functional scenarios;

[0065] Based on the ADS feature model obtained in the requirements analysis phase, the scenarios corresponding to the ADS features are further analyzed to reveal the system's behavior and interaction patterns in different situations. The BPMN flowchart is used to analyze the workflow and behavior of each feature under specific conditions. Figure 5 shown.

[0066] 5) Based on the behaviors and interaction patterns of the scenarios corresponding to the features, the ADS requirements are refined to obtain a refined ADS requirement model.

[0067] like Figure 6As shown, for example, in the autonomous driving system demand scenario, the vehicle control requirements are refined into: the system supports acceleration control, the system supports automatic control braking, and the system supports control direction, including turning, lane changing, and lane keeping; for example, in the path planning scenario, the requirements are refined into obstacle recognition and avoidance, specifically identifying obstacles on the road, such as other vehicles, pedestrians, animals, or static obstacles, and planning paths to bypass them, etc.

[0068] 6) Based on the behavior and interaction patterns of the scenarios corresponding to each feature, the ADS functional architecture model is obtained

[0069] Based on the functional scenario analysis corresponding to each feature, the ADS functional architecture is obtained and described using the SysML module definition diagram. The functional architecture corresponding to the vehicle control feature is as follows: Figure 7 The ADS functional architecture includes acceleration control functional module, braking control functional module, steering control functional module, speed adjustment functional module and lane keeping functional module.

[0070] 7) Based on the refined ADS requirement model and ADS functional architecture model, build a traceability relationship from requirements to functions.

[0071] Use the design structure matrix to build a traceability form from requirements to functions, such as Figure 8 shown.

[0072] Step 3: Component-level functional analysis

[0073] 8) Component activity model: According to the ADS functional architecture model, determine the functional activities of each component in the subsystem and obtain the component activity model

[0074] Before designing the ADS logical architecture, you first need to determine the functional activities of each component. Therefore, after analyzing the ADS functional architecture, you need to perform component-level functional analysis to determine the functional activities of each component in the subsystem ADS and support the ADS function allocation to logic. The functional activity models within each component are included in the subsystem-level function allocation to logic model.

[0075] Step 4: Component-level logical architecture design

[0076] 9) Component logical composition model: Based on the component activity model, determine the logical composition of each component and generate the component logical composition model;

[0077] Before designing the ADS logical architecture, it is necessary to first determine the logical composition of each component. Therefore, the logical composition of each component is analyzed to support the construction of the ADS logical architecture model, and the logical composition model of each component is included in the logical architecture model.

[0078] 10) Component logical connection: Based on the component activity model, determine the data exchange and information transfer between components and build a component logical connection model;

[0079] In order to ensure that data exchange and information transfer can be carried out effectively within the ADS logic components, the SysML internal module definition diagram is used to describe the interface model within each ADS logic component, such as Fig. 9 As shown, the component-level logical architecture model is obtained.

[0080] Step 5: Subsystem-level logical architecture design

[0081] 11) Allocating functions to logic: Based on the component-level logical architecture model, allocating functions under each feature of the ADS to logical components;

[0082] To ensure the implementation of the functions, each function is assigned to the corresponding logical component, and the specific functions and tasks that each component needs to perform are determined. SysML activity diagrams and swimlanes are used to perform the process of assigning the functions under each feature of ADS to the logical components. Among them, the functions under the vehicle control feature are assigned to the logical components. Fig.10 shown.

[0083] 12) Logical composition: Generate the logical composition model of ADS;

[0084] Based on the process of allocating ADS functions to logical components, the logical composition model of ADS can be obtained, such as Fig.11 As shown, the logical component model includes vehicle control system, vehicle perception system, high-precision map, communication system, fault management system, and time data recording system.

[0085] 13) Logical interface: Determine the data exchange and information transfer between components, and establish the interface model between the logical components of ADS;

[0086] In order to ensure that ADS logic components can effectively exchange data and transfer information and realize the collaborative work between ADS logic components, under the guidance of the function allocation to logic and logic composition related models, the SysML internal module definition diagram is used to describe the interface model between the ADS logic components, such as Fig.12 shown.

[0087] 14) Logical architecture: Based on the logical composition model and logical interface model of ADS, the subsystem-level logical architecture model is obtained;

[0088] Based on the above logical architecture analysis process, the ADS logical architecture model is obtained as follows: Fig.13 shown.

[0089] Step 6: Component-level physical architecture design

[0090] Before designing the subsystem ADS physical architecture, you first need to determine the physical selection of ADS components. Therefore, after designing the subsystem logical architecture, design each component, including 15) component physical composition, 16) component physical interface and 17) component physical parameter model, and finally perform physical synthesis to form the subsystem ADS physical architecture.

[0091] Step 7: Subsystem-level physical architecture design

[0092] 18) Physical composition

[0093] In order to support the implementation of the logical architecture, the SysML module definition diagram is used to further describe the physical composition of the ADS. The physical composition model of the vehicle control system is as follows: Fig.14 shown.

[0094] 19) Physical Interface

[0095] SysML internal module diagram is used to describe the interface interaction relationship between ADS physical components, such as Fig.15 shown.

[0096] 20)Physical parameter selection

[0097] In order to ensure that the designed ADS can meet the expected performance and requirements and control the ADS design cost, the SysML parameter diagram is used to define the ADS physical parameter model, such as Fig.16 shown.

[0098] The above is only a 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 a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A vehicle automatic driving system modeling method based on MBSE, characterized in that: The specific process is: Step 1: Subsystem-level requirements analysis: Build an ADS requirements model based on expression rules and constraints, and incorporate the external ADS requirements model; analyze ADS features based on the requirements model and build an ADS feature model; identify external entities and interaction processes based on the ADS feature model, and define the ADS system boundaries and use cases; Step 2: Subsystem-level functional architecture analysis: Analyze the behavior and interaction mode of ADS features in different functional scenarios; Based on the behavior and interaction mode, refine the ADS requirements to obtain the refined ADS requirement model and ADS functional architecture model, and build a traceability relationship from requirements to functions; Step 3: Component-level functional analysis: According to the ADS functional architecture model, determine the functional activities of each component in the subsystem and obtain the component activity model; Step 4: Component-level logical architecture design: Based on the component activity model, determine the logical composition of each component and generate a component logical composition model; determine the data exchange and information transmission within each component, build a component logical connection model, and then obtain a component logical architecture model; Step 5: Subsystem-level logical architecture design: Based on the component logical architecture model, the functions of executing each feature of ADS are allocated to the logical components to generate the logical composition model of ADS; the data exchange and information transmission between the components are determined, and the interface model between the logical components of ADS is constructed to obtain the subsystem-level logical architecture model; Step 6: Component-level physical architecture design: perform component-level physical architecture design based on the component logical architecture model; Step 7: Subsystem-level physical architecture design: Complete the subsystem-level physical architecture design based on the subsystem-level logical architecture model.

2. The MBSE-based vehicle automatic driving system modeling method according to claim 1 is characterized in that: In the step 1, the SysML use case diagram is used to define the ADS system boundary and use cases.

3. The MBSE-based vehicle automatic driving system modeling method according to claim 1 is characterized in that: In the step 1, an ADS feature model is constructed based on the vehicle feature model of EAST-ADL.

4. The MBSE-based vehicle automatic driving system modeling method according to claim 1 is characterized in that: In the step 2, a BPMN flowchart is used to analyze the workflow and behavior of each feature under specific conditions, and the behavior and interaction mode of the ADS feature in different scenarios.

5. The MBSE-based vehicle automatic driving system modeling method according to claim 1 is characterized in that: In the step 2, the ADS functional architecture is described using a SysML module definition diagram.

6. The MBSE-based vehicle automatic driving system modeling method according to claim 1 is characterized in that: In the step 4, a SysML internal module definition diagram is used to describe the interface model inside each logical component of the ADS.

7. The MBSE-based vehicle automatic driving system modeling method according to claim 1 is characterized in that: In the step 5, the SysML internal module definition diagram is used to describe the interface model between the various logical components of the ADS.

8. The MBSE-based vehicle automatic driving system modeling method according to claim 1 is characterized in that: In step 5, the SysML activity diagram and swimlane are used to perform the process of allocating the functions under each feature of ADS to the logical components.

9. The MBSE-based vehicle automatic driving system modeling method according to claim 1 is characterized in that: The step six component-level physical architecture design includes the design of component physical composition, component physical interface and component physical parameter model.

10. The MBSE-based vehicle automatic driving system modeling method according to claim 1 is characterized in that: The step seven subsystem-level physical architecture design includes the design of ADS physical composition, ADS physical interface and ADS physical parameter selection.

Images