Dry goods sharing|Introduction to Model-Based Systems Engineering (MBSE) (Part 1)
Edited from: https://icv.51cto.com/posts/427
foreword
Automobile intelligence is mainly reflected in the digitization and automation of the vehicle's own functions, as well as the information interaction and deep coupling between the vehicle and the outside world (roads, clouds, networks, maps, etc.). Its essence lies in: deeply integrating the powerful real-time mapping and computing deduction capabilities of the "digital virtual body" into the "physical entity" composed of traditional vehicles and roads, forming a complex large system combining "virtual and real". This system is a typical feature of cyber-physical systems (Cyber-Physical Systems, CPS) with the characteristics of interdisciplinary, cross-field, cross-regional high complexity and heterogeneous interoperability, and can be called intelligent networked vehicle cyber-physical system (ICV CPS).
As a typical complex system engineering problem, ICV CPS needs a systematic approach to solve it. The Model-Based Systems Engineering (MBSE) method and corresponding modeling software tools are considered to be effective application methods of ICV CPS.
This article will be divided into two parts, the first and the second, to introduce MBSE systematically. This article is the first part, and will analyze the background significance, method advantages, and domestic and foreign research status of MBSE.
1. Background and Significance
As a typical complex system engineering problem, ICV CPS has many behavioral characteristics and requirements that each member system does not have. If the traditional design mode independently developed by each member system is used for the research and development of each member system, it may not only lead to the failure of each member system to meet the requirements of ICV. The requirements of CPS may also easily cause member systems to have coordination failures with other member systems in ICV CPS, and even lead to failure and disintegration of ICV CPS.
Systems Engineering (SE), derived from the Manhattan Project and Qian Xuesen in the United States, proved to be one of the most effective methods to solve the difficulties in the development of complex systems. This method was first successfully applied to aerospace systems, and then Model-Based Systems Engineering (MBSE) was established. method, and built the corresponding modeling software tools.
MBSE has been initially applied in the automotive field. Ford of the United States is applying MBSE to complete user demand analysis, system architecture design, and vehicle-level system simulation verification for smart cars; foreign industrial software suppliers, such as Dassault Systèmes of France and Siemens of Germany, are all in The construction covers the entire process of automobiles from system requirements, system development, detailed design to integration testing; China has proposed ICV-oriented The R&D and design technology path of CPS.
2. MBSE is an effective application method for systems engineering
As a branch of systems science, systems engineering can better achieve the purpose of the system by conducting scientific and methodical research and analysis on the composition, structure, and information flow of the system.
The International Institute of Systems Engineers (INCOSE) defines systems engineering as: systems engineering is a method and means that can make systems interdisciplinary. Systems engineering focuses on defining and documenting customer requirements in the early stages of system development, and then considers issues such as system operation, cost, schedule, performance, training, support, testing, and manufacturing, and then conducts system design and validation.
It can be seen that systems engineering can be applied to establish interdisciplinary complex large-scale systems, so that the coordination and cooperation between disciplines, subsystems and subsystems, and the whole and part of the system can optimize the operation of the system. .
With the growth of demand and technological innovation, traditional industries are gradually transforming to intelligent and digital. In the new industrial environment, the huge amount of information and data generated by the increase of system complexity has brought unprecedented challenges to traditional document-based systems engineering (TSE). In this context, Model-Based Systems Engineering (MBSE) emerged and developed gradually. In fact, the transformation from TSE to MBSE is the transformation from "document" to "model", and the main differences are shown in Table 1.
First, the use of models can largely solve the problem of non-standard and inconsistent documentation standards; second, the model can more comprehensively and three-dimensionally describe the composition, function, operation and other characteristics of complex systems, and strengthen the understanding and understanding of engineering personnel. Cognition, and eliminate possible ambiguity in communication; Finally, compared with documents, model maintenance and update are more convenient and efficient, so MBSE can speed up the iteration speed of the system.
Table 1 Differences between the traditional complex system R&D model and the MBSE R&D model
3. Advantages of MBSE method
The significance or advantages of the MBSE method to develop complex systems can be summarized in three words: "fast", "saving", and "good".
● **"Fast" is reflected in: **In the initial stage of complex system development, MBSE can realize the loop iteration of scheme design, which can accelerate the confirmation of requirements at each stage and the improvement of system scheme to the greatest extent;
● "Province" is reflected in : in the process of complex system development, on the one hand, multi-discipline and multi-field models can be reused; The time cost is greatly reduced, and the economic cost of program modification, improvement and optimization is close to zero;
● "Good" is reflected in : the complex system is verified by physical verification to multi-disciplinary joint simulation verification based on the model, which ensures continuous joint simulation evaluation at each stage of the design, and avoids "over-design" and "under-design".
4. Research status
Systems engineering originated in the early 20th century and was first applied in World War II. It was not until 1951 that Bell Corporation of the United States formally proposed the term "system engineering" after it built a microwave relay communication network. In 1972, the Apollo manned moon landing project in the United States achieved great success using the method of systems engineering, which made systems engineering known worldwide for the first time. Later, under the leadership of the US Department of Defense, contractor standards were introduced, and systems engineering was gradually adopted in the field of civil aviation. Until 1990, in order to meet the development needs of high-complexity industries and products, INCOSE was established under the strong advocacy of aerospace and defense companies such as Boeing and Lockheed, which also laid a solid foundation for the development of system engineering in academia and industry. good foundation.
Elm et al. have conducted research on the impact of system engineering capabilities on overall project effects. They pointed out in the "Survey of System Engineering Effectiveness" that improving system engineering capabilities can obviously improve the overall effect of the project. Especially when the overall challenge of the project is low, improving the system engineering capability can even double the overall effect of the project.
In Systems Engineering Effectiveness: A Complexity Point Paradigm for Software-Intensive Systems in the Information Technology Domain, Buckle and Wilma present a method for evaluating the effectiveness of using systems engineering methods and practices in the information technology sector of large integration projects. Technology.
They concluded that applying systems engineering to project development, testing, and management can greatly improve project productivity and help projects meet cost, schedule, and technical scope requirements.
Research Status Abroad
As an important branch and future trend of system engineering, the origin of MBSE can be traced back to the middle and late 20th century. The emergence of system model theory, such as Nam's axiomatic design theory and Tarski modeling theory, has accelerated the development of methods for applying model theory to system engineering.
In 1993, Wymore proposed a modeling method based on mathematical expressions in systems engineering in "Model-Based Systems Engineering". This method establishes the connection between various elements by abstracting the elements in system engineering. In 1997, the Object Management Group (OMG) released the Unified Modeling Language (UML), and based on UML, the System Modeling Language (SysML) was proposed in 2003. Relying on its visualization features, SysML is applied to large-scale or cutting-edge engineering projects to improve development efficiency and reduce development costs. In 2007, INCOSE proposed the definition of MBSE in the "Vision of Systems Engineering 2020": "The application of formalization and standardization of modeling behaviors for system requirements, design, analysis, verification and validation, etc. The conceptual design stage begins and runs through the system development and subsequent life cycle.
"MBSE, as a new paradigm, has been applied to various major projects by government organizations such as the US Department of Defense and the European Space Agency (Figure 1 shows that the US Department of Defense lists MBSE as a research focus). German Industry 4.0 is also List "Using Models to Master System Complexity" as one of the important areas of activity in the future.
Compared with traditional text-based system engineering, the unique advantages of MBSE have also led many government organizations and enterprises to transfer their technical routes from TSE to MBSE. NASA explicitly requires that the system demonstration deliverables must be models. Lockheed Martin's submarine design team spent a year converting all original documents into system models during the design process of a new submarine electronic system. In addition, in the automotive field, the application of MBSE in the development of automotive electrical systems has become a mainstream solution for Tier 1 suppliers such as Siemens. Automobile companies such as Ford have also begun to apply MBSE in the entire product development process.
Figure 1 Department of Defense Systems Engineering Research Center (SERC) lists MBSE as a research focus
(Image source: Industrial Internet Innovation Center website)
Research state in China
Compared with foreign research on systems engineering and MBSE, domestic research is mainly concentrated in the early 21st century. These studies are more focused on how to combine China's national conditions to carry out system engineering and the practical application of MBSE. Common MBSE methods are mainly used in the aerospace field in China, and systems engineering and MBSE are mainly used. For example, the Aviation Industry Group cooperated with IBM to introduce the IBM Harmony-SE method and a complete tool system to carry out aviation system engineering, and through simulation tests, verification, and optimization of the system design (Figure 2 shows the 2016 application). COMAC uses the Harmony-SE method to carry out aircraft system design requirements and functional logic verification to solve the problems of complex aircraft requirements and difficult verification. With the continuous update and iteration of domestic industrial technology, the complexity of systems in some other fields has increased, such as smart cities, smart transportation, and intelligent networked vehicles, etc., making MBSE more widely used.
Figure 2 MBSE application of Aviation Industry Group
(Photo source: Public Account of First Aviation Academy of Aviation Industry)
The previous part has come to an end, and the next part will introduce the modeling process, methods, tools and future development of MBSE, so stay tuned.
Dry goods sharing|Introduction to Model-Based Systems Engineering (MBSE) (Part 2)
Edited from: https://icv.51cto.com/posts/426
Automobile intelligence is mainly reflected in the digitization and automation of the vehicle's own functions, as well as the information interaction and deep coupling between the vehicle and the outside world (roads, clouds, networks, maps, etc.). Its essence lies in: deeply integrating the powerful real-time mapping and computing deduction capabilities of the "digital virtual body" into the "physical entity" composed of traditional vehicles and roads, forming a complex large system combining "virtual and real". This system is a typical feature of cyber-physical systems (Cyber-Physical Systems, CPS) with the characteristics of interdisciplinary, cross-field, cross-regional high complexity and heterogeneous interoperability, and can be called intelligent networked vehicle cyber-physical system (ICV CPS).
As a typical complex system engineering problem, ICV CPS needs a systematic approach to solve it. The Model-Based Systems Engineering (MBSE) method and corresponding modeling software tools are considered to be effective application methods of ICV CPS.
This article is divided into upper and lower two parts, and introduces MBSE systematically. The first part analyzes the background significance, method advantages, and domestic and foreign research status of MBSE (for details, see: dry goods sharing | introduction to model-based systems engineering (MBSE) (on) ). This article is the next one, and will introduce the modeling process, methods, tools and future development of MBSE.
Figure 1 V flow of MBSE method
1. Three pillars of MBSE
MBSE is a methodology and a collection of related processes, methods and tools. The MBSE method combines system engineering ideas, and runs through the entire process of the system's entire life cycle through the model (as shown in Figure 1). The model is the core of the entire MBSE method and is also the basis for the MBSE method to achieve efficient R&D and high-quality design in system R&D. The foundation, or the accumulation and embodiment of system technology and engineering experience in the process of system research and development, is the core asset of an enterprise.
Therefore, the application of MBSE method should focus on the construction of the system model, especially the modeling language, modeling tools and modeling methods involved in the modeling process, these three are also called "the three pillars of MBSE". The following is a brief introduction to the "three pillars".
modeling language
A standardized and robust modeling language is considered as a key factor to achieve MBSE. The system modeling language (OMG SysML™ System modeling language) is such a general modeling language, which can meet the multi-level modeling needs (behavior, structure, performance...), and supports many Subject exchange and multi-field cooperation are intuitive, the whole process is measurable and controllable, and it has the advantages of strong scalability and support for multiple tools.
SysML is a graphical modeling language that supports the specification, design, analysis, and verification of systems, which may include hardware and equipment, software, data, people, processes, and facilities, for modeling system requirements, behavior, structure, and Parameters provide a semantic basis for integration with other engineering analysis models.
It is the language that MBSE practitioners "speak" when creating system models, allowing them to visualize their system design views and communicate them to stakeholders. The visualized "communication" method of SysML is completed through the following nine pictures (as shown in Figure 2).
Figure 2 Nine pictures of SysML language
modeling tool
Modeling is a direct technical means to implement MBSE applications. The tools used are many commercial tool software on the market. The system architecture modeling tools of foreign manufacturers include Rhapsody of IBM, Capella of Tales tool promoted by Siemens, MagicDraw of Dassault Systèmes, Sparx Systems' Enterprise Architect (EA) and other related tools (the modeling schematics of two of the software are excerpted, as shown in Figure 3). In addition, domestic manufacturers have also launched domestically produced SysML system modeling tool software, such as Modelook, in recent years.
Figure 3 Schematic diagram of system architecture modeling tool software
modeling method
With the application and development of MBSE, different methodologies of MBSE have emerged to realize the application and modeling of MBSE. Among them, INCOSE has recognized more than 7 kinds, and the common one is IBM's Harmony-SE (as shown in Figure 4(a)). , INCOSE's OOSEM (as shown in Figure 4 (b)), Dassault Systèmes' MagicGrid, etc.
Figure 4 Two common MBSE methods (schematic)
2. Opportunities and challenges
INCOSE put forward the long-term plan of MBSE for the first time in the "Vision of Systems Engineering 2020". By 2020, a complete MBSE theory and practice system will be defined, and by 2025, a mature cross-domain model library and integrated development environment will be established. This provides a solid theoretical basis for domestic MBSE applications. In addition, based on the many advantages of MBSE, its application can provide support for the establishment of increasingly complex cross-domain system systems in China, so that the iteration of the system system can keep up with the speed of technological updates.
At the same time, the overall promotion and application of MBSE in China is facing many challenges. Although MBSE's model iteration speed is very fast, it takes a lot of time to standardize, standardize, and model complex large systems at the beginning. In addition, traditional system developers will also spend a certain amount of time on learning MBSE and transformation. There are too few practical applications of domestic MBSE in non-aerospace fields, which also limits the public's acceptance to a certain extent.
In the field of intelligent networked automobile industry, the member systems of car-road-cloud-network-map are closed and independent, which makes their traditional independent research and development models difficult to meet the research and development needs of ICV CPS. Therefore, if it is necessary to clarify the ICV CPS-specific design method and R&D process, domestic ICV CPS-related companies and research institutions need to build a systematic method system and design tools based on MBSE on the basis of combining the traditional R&D models of ICV CPS member systems Chain prototype ( see the description of tool chain construction in "Intelligent Connected Vehicle Cyber-Physical System Reference Architecture 2.0" for details ) to quickly support the R&D and development of ICV CPS, build a Chinese solution for intelligent connected vehicles, and get rid of the process of foreign technology research Multi-dependence forms the global industrial advantage of intelligent networked vehicles.
3. Summary and Outlook
The new generation of information and digital technology promotes the deep integration of automobiles, transportation, communications and other fields, forming an intelligent networked vehicle cyber-physical system (ICV CPS) with complex system characteristics. In this complex system, vehicle-road-cloud-network-map are closed and independent R&D models, which cannot adapt to the R&D and development of ICV CPS, and there is no clear ICV CPS design method and R&D process at this stage. Therefore, the MBSE method, a complex system engineering method derived from the aerospace field, has become a solution to the difficulties in the development of ICV CPS.
Compared with the traditional R&D model, MBSE has the advantages of improving system development efficiency and reducing development costs, and is the development direction of future system engineering. However, the MBSE method needs a lot of resources and time investment in standardization application, relevant personnel transformation, model library construction, etc., which greatly hinders its promotion in China. An effective means to overcome these obstacles is to construct a systematic method system and design tool chain prototype based on MBSE, so that all engineers and technicians can properly complete the collaborative design of complex systems in a short period of time without being proficient in MBSE methods.
At this stage, in the face of problems such as ICV CPS design, research and development, simulation, verification, deployment and application, there are still no perfect basic theories, core methods and special design tools in China. The National Intelligent Connected Vehicle Innovation Center will combine MBSE methodology to research and standardize Key basic technologies such as architecture construction, integrated unified modeling, digital twin design, verification evaluation, implementation confirmation, etc., build a set of theoretical system, technical route and implementation method for the whole life cycle of ICV CPS, and then in the ICV CPS system architecture design and configuration Optimization, ICV CPS reference architecture model library, interdisciplinary model fusion, data traceability design, overall verification and evaluation have formed innovative technological achievements, providing a scientific basis and implementation path for the technical development and application realization of the Chinese scheme ICV CPS. ****