[MA] Consistent Simulation Management for Digital Twin Families

A digital twin is an evolving technology that allows advanced simulation reinforced by real-world data in real time. Considering large-scale industrial objects like road systems (the focus point of this thesis), running a digital twin requires thorough simulation management. In this case, complexity is defined by two main factors: multi-physics simulation and intensive variability. A motivating example is a thermo-mechanical simulation for different road states. This model includes two-step simulation: thermal based on sensor value and mechanical based on thermal simulation results. Variations of this simulation are driven by weather state and road structure. Using techniques for reducing computational cost, e.g., ALE (Arbitrary Lagrangian Eulerian) formulation, allows the specification of a structured mesh for an area of interest, which can move according to object dynamics. Practically, this means that instead of one complex simulation for the whole road system, managing multiple vehicle-wise smaller simulations that cover only a small area around every vehicle is worthwhile.

Betreuer: Volodymyr Prokopets