AGENT-BASED MODELING IN MULTI-LEVEL INDUSTRIAL ECOSYSTEMS DEVELOPMENT

Resumo

In industrial ecosystems development, there are currently trends towards deepening vertical and horizontal integration within the innovation processes framework. They entail approaches complication to management functions implementation. Management and development problems solutions in multi-level industrial ecosystems becomes particularly relevant. The study’s purpose is to formulate possible approach to such problems solving that can increase management decision-making efficiency. The work proposes simultaneous agent-based modeling and multi-level digital twins use in order to simulate economic processes. The study proposes multi-level industrial systems conceptual scheme for an agent-based modeling, taking into account its’ vertically hierarchical structure. The proposed model identifies four levels (with their own agents), differing in the nature of the tasks being solved, the responsibility area, organizational and economic mechanisms used. It is proposed to base the model on economic and mathematical tools, in particular computer modeling methods, creating digital twins specifically. Digital twins are used to analyze production chains, assess internal and external factors effect, develop alternatives and select most preferable solutions to emerge management problems. At the same time, it was determined that digital twins structure should be multi-layered, where each subsequent level incorporates digital twins developed on the previous one, endowed with implemented functions certain set. It is substantiated that one of the important tasks is to determine industrial ecosystem digital twin managerial layer optimal configuration. This layer is responsible for modeling the organizational and managerial component and is built on needs to achieve financial and economic activity target indicators. The study proposes and describes the agent-based model operation mechanism, the development of which is allows to produce management strategies based economic and mathematical modeling complex tools, scenario and forecast analysis and digital twins numerical modeling.