Do I need to standardize modeling?

Prerequisites and world practice

Considering the general technological and political, extremely unstable and rapidly changing environment, it is necessary to ensure that Russian enterprises can not only develop advanced technical solutions or products, but also create an opportunity for qualitative growth. The basis of this strategy can be the development of standards that will be the foundation for universal inter-sectoral exchange in the life cycle of any object under study. One of the directions can be the introduction of computer-engineering technologies and the centric approach model into practice. At the same time, it should be noted that the standardization system developed in Soviet times in the field of mechanical engineering and construction was considered one of the best, due to the fact that within these standardization systems it was possible to implement a systematic approach to solving the problem. However, modern challenges and imminent technological development require the use and active use of integrated modeling approaches, which are based on numerical methods and algorithms.



In order to develop a single holistic concept of standardization in the use of numerical modeling, we suggest to proceed from the following prerequisites:



1. The developed group of standards should cover all industries and form a supra-branch “umbrella”.

2. The developed group of standards should ensure the possibility of a qualitative leap in technological and innovative development of the state, and first of all, priority development areas, such as:

2.1. Computer engineering, industrial design and innovative engineering of automated systems for process management of activities and production.

2.2. The best available technologies, including additive manufacturing technologies.

2.3. "Green" and resource-saving technologies.

2.4. Integral management of large socio-technical systems.

3. Solve strategic industry problems, such as:

3.1. Development of a comprehensive development strategy for the Best Available Technologies;

3.2. Formation of requirements for product lifecycle management;

3.3. Creating an ecosystem in the field of engineering.

4. To form the basis for international and interstate standardization.

5. The ideological basis for the development of standards should be a methodological approach based on “Systems Analysis and Engineering”.

6. Providing a comprehensive analysis within the life cycle of the objects of study, such as for example:

6.1. Complex engineering complexes and facilities of vital, critical energy infrastructure and defense complex;

6.2. Electronics, precision engineering, bioengineering and laser technology;

6.3. Construction of infrastructure facilities;

7. Providing analysis and modeling of cyber-physical systems and bio-cybernetic systems.



Before proceeding to the designation of areas of standardization, it is necessary to briefly review international experience in standardization of modeling and computer engineering.

')

As an example of a comprehensive solution to this problem, we can cite the project Complex Adaptive Systems-of-Systems Engineering , which took place in America from 2008 to 2012, and as a result, a platform was developed that combined analysis, simulation, and accumulation into an interdisciplinary life cycle. knowledge. This is presented in the form of the following diagrams (left to right): CASoS Engineering theory and experiment, developing through application to real problems and the Structural Integration Scheme of the theoretical platform CASoS Engineering, technical

environment and stimulating reality of super-real applications.











More information about the CASoS approach can be found in the documents:

• A Roadmap for the Complex Adaptive Systems of Systems (CASoS) Engineering Initiative
• Complex Adaptive System of Systems (CASoS) Engineering Environment Version 1.0
• Complex Adaptive System of Systems (CASoS) Engineering Framework Version 1.0

You can also give examples of other similar projects, the European - Next Generation Infrastructure (ISNGI) and the UK Infrastructure Transitions Research Consortium . Other fundamental research is research in the field of modeling languages ​​and systems description, as well as research and development in the field of interfaces for the exchange of information models. For example, research of innovations in the field of cyber-physical systems ( NIST Foundation for innovation in Cyber-Physical Systems Workshop ), research of the Mathematics Department of the Polytechnic University of Milan ( The Interface Control Domain Decomposition (ICDD) ) or standards developed by the IEEE “1516 WG - HLA Evolved Working Group .



These examples show that the work on the formation of standards in the field of information modeling should be carried out systematically.

Analysis of standardization examples

The starting point for starting the development of standards in the field of mathematical modeling is to consider the experience of standardization in the construction, oil and gas and cartographic industries, due to the fact that these industries were forced to integrate various approaches and technologies within their activities, to use interdisciplinary solutions that provided industry move forward. As examples, examine the federated model from the IFC ISO 16739 standard, the RDL Semantic model from the ISO 15926 standard, and the Eurocode system. It also makes sense to consider the standards “GOST R ISO / IEC 10746 - Information technology. Open distributed processing "and" GOST R ISO / IEC 10000-3 - Information technology. Basis and taxonomy of international functional standards ".

Strategic approaches and directions for standardization of modeling

Based on the analysis performed, an approximate list of tasks for standardization can be defined:

1. Development of a model-oriented approach development strategy

2. Development of a universal language for modeling systems and branch toxanomas

3. To ensure the universality and standardization of information exchange between information models for different purposes, as well as to ensure bi-directional unambiguous exchange between business requirements for the product and the final engineering developments, it is necessary to develop a specialized language . This language should have the characteristic of reading and understanding from the sheet specialists of different specialties (from business to technical tasks) and be machine-readable. On the basis of this language, it will be possible to bring courses in systems engineering to a single format, as well as to form an interdisciplinary connection within the framework of educational programs. The context of a language is formed from sectoral taxonomies (an ontological subset of semantic constructions defining the sectoral orientation of activity), the dimension of which does not exceed 2000 lexemes. At the same time, an essential thesaurus is being formed, which is essentially a “hypercube” of modeling ontology.

4. Formation of the list of required tools for numerical modeling.

5. Formation of a list of required tools and methodologies for ensuring stability, speed of convergence and resource consumption when executing algorithms on various hardware computing systems.

6. Formation of a list of promising tools for numerical modeling.

7. Formation of the library system of verified modeling tools.

8. Formation of interfaces for data exchange between simulation systems.

9. Formation of software and hardware infrastructure

10. Forming a marketing strategy to promote the use of standards.



Such an approach allows implementing the principles described above and ensuring further strategic development. At the same time, proceeding from the law on standardization of standards, in order to speed up the possibility of “rolling in” the proposed versions of standards, it is possible to form a set of rules that will later serve as the basis for the development of final standards.



Accordingly, within the framework of standardization, it is proposed to consider the following standardization scheme (Figure 1).

Core

Standards that define the following areas should be in the core of the platform:

• Terms and definitions, including a multilingual glossary of terms and definitions.
• A language describing models, process modeling and other processes.
• Principles of describing the life cycle of systems / objects.
• Guidelines for the collection and processing of information in the framework of the reference model describing the life cycle of an object / system.
• Interfaces of exchange between systems of modeling and analysis.
• Principles of separation of mathematical methods and approaches to the "Basic" and "Subject" ("Industrial").
• Guidelines for checking models for adequacy.
• Guidelines for the application of numerical methods using computing systems, including parallel computing systems.

This will determine the boundaries of the core of the platform and lay a solid foundation for the further development of sectoral areas with the possibility of organizing intersectoral and supra-sectoral research.

Basic algorithms and methods

As the basic algorithms and methods, we propose to define such mathematical algorithms that can be used in at least two different branches or are used by any two or more algorithms or methods of a higher level. Also, standard methods for assessing and testing, as well as standards for checking models for adequacy, i.e. standards for verification and validation of models. However, it is worth noting that later, due to interdisciplinary research, a situation is possible when industry standards can be incorporated into the basic algorithms, for this purpose a procedure for analyzing and updating standards should be provided.

Subject specialization

Each industry has its own specific areas for research and specialized tasks that are solved on the basis of basic methods and algorithms.



In general, an industry standardization scheme can be represented as the following diagram 2.







Materials on the topic:

1. Artificial Intelligence. Modern Approach (Artifical Intelligence: A Modern Approach) Authors: Stuart Russell, Peter Norvig Publisher: Williams ISBN 0-13-790395-2, 978-5-8459-0887-2; 2007
2. Artificial Societies and Social Simulation Vol. 7, no. 3 < jasss.soc.surrey.ac.uk/7/3/5.html > Published: 30-Jun-2004
3. Using the languages ​​of Ines Čeh, Matej Črepinšek, Tomaž Kosar, Marjan Mernik, Faculty of electrical engineering and computer science, Smetanova 17, 2000 Maribor, Slovenia
4. Spatio-temporal modeling of situations in the ACS. Grigan A.M. UDC 004.023 SWorld - 1-12 October 2013
5. O. E. Barndorf-Nielsen, J. Schmiegel, “Spatio-temporal modeling based on Levy processes and its applications to turbulence”, UMN, 59: 1 (355) (2004), 63–90
6. S.V. Smirnov. Pragmatics of ontologies: an object-oriented model of knowledge about the domain // KII 2008
7. Proceedings of the International Scientific Conference "MATHEMATICAL METHODS IN TECHNOLOGY AND TECHNOLOGIES" (MMTT-27) June 3 - 5, 2014
8. OPEN Process Framework (OPF) - opfro.org
9. Standard Reference Data Act [PL 90-396; 15 USC 290-290f] www.nist.gov/srd/upload/SRDAct-2.pdf
10. Gellish Modeling Method Part 3B Ir. Andries van Renssen
11. "Trends in Multiple Criteria Decision Analysis", Editors Matthias Ehrgott