DIGITAL TWIN TECHNOLOGIES ENABLED BY COMPUTATIONAL INTELLIGENCE

Abstract

Digital twin technology has well emerged as one of the foundational elements of the modern cyber–physical systems, enabling the continuous interaction between the physical entities as well as their virtual form of representations. Since computing intelligence such as machine intelligence, deep intelligence, reinforcement intelligence, and other hybrid intelligence models are rapidly evolving, the digital twins are no longer beginning as a basic simulation tool, but as an adaptive, predictive and independent system. This paper is an in-depth analysis of digital twin technologies that are implemented using computational intelligence. It covers their theoretical background, architecture and capability and proceeds to give an overview of current advances in intelligent digital twins in diverse fields of application. The structure of the methodology of the design and deployment of computational intelligence digital twins is provided because it assumes the combination of information, the selection of models, the approaches to their implementation, and the performance scales. The results of the experimental analysis, conducted on the basis of the representative scenario of a cyber-physical system, indicate the improvement of the quality of predictions, robustness of the system, economy of the operations, and excellence of decisions in terms of the smart digital twins. The results indicate an opportunity to improve the performance by a huge margin, and allude to the challenges in the form of latency, interpretability, cybersecurity and lifecycle management. The paper concludes with a statement on how it can be applied to research and practice in the future, as they must have believable, understandable, and scalable intelligent digital twin ecosystems.