An investigation of self-learning and self-protection for Adaptive Digital Twins

Abstract

Adaptive Digital Twins are applicable to a number of fields, including the cybersecurity of industial control systems. This thesis prototypes a Self-Learning adaptive digital twin and posits an architecture for the creation of digital twins based on the learnings gained from the prototype. The prototype shows the efficacy of control theoretical approaches for adaptive digital twins for both modelling and protecting a system, and the architecture posits a generalised method for developing adaptive digital twins.