Digital twin as a tool for modeling and optimization of complex natural and technical systems
Abstract
The concept of a digital twin (hereafter referred to as DT) is discussed as a complex cyber-physical system that represents a virtual representation of physical objects, processes, or systems. A retrospective analysis of the evolution of this technology is conducted, starting from its origins in NASA’s practice and culminating in contemporary conceptual approaches, such as the product life cycle model proposed by Michael Grieves and the multi-physics models developed by Glassgen. Unlike simple modeling, a DT ensures dynamic correspondence between the virtual and physical entities through continuous data exchange and feedback. Key methodological aspects of creating and operating a DT are identified, including issues related to the integration of heterogeneous data, the selection of appropriate models, and ensuring interoperability. A critical analysis of the advantages and limitations of this technology is provided, taking into account considerations regarding the need for validation and the limitations associated with the availability and quality of data. The prospects for further development of DT are discussed, particularly the integration with artificial intelligence technologies and big data analytics to address complex tasks related to sustainable development and the minimization of anthropogenic impact on the environment, including aspects of pollution monitoring and natural resource management. It is specifically emphasized that, unlike a simple database, a DT possesses an operational model that enables the interpretation and use of data to solve specific tasks.