Measuring the fidelity of digital twin systems

Abstract

A digital twin is a virtual replica of a system at a certain level of fidelity, synchronized at a specific frequency. Digital twins often replicate physical systems whose simulations are usually computationally costly. One of the solutions to this problem proposed in the literature is to define a hierarchy of multi-fidelity digital twins, where we use one twin or another depending on the specific purpose. However, one of the challenges of this proposal is the need to determine whether the different twins are equivalent to each other and the physical system. In this thesis, we explore different methods to measure this equivalence by analyzing the state and behavior of the twins with the aid of high-level software models.