Proposal for a Digital Twin Approach Based on Uncertainty Management Interactive Optimisation

Abstract

In multi-criteria mechanical design, it is common to support the decision-making process with algorithms that facilitate understanding the impact of modifications in design variables on the final performance and associated design objectives. This article presents the development of an interactive optimisation method for application in an operational context through the introduction of a digital twin approach. The original method incorporates uncertainty management via the CPM/PDD model to establish a traceability framework, enabling near-optimal solutions that align design objectives in a multidisciplinary context. Furthermore, it includes a graphical tool that allows users to manually adjust variables while assessing compliance with design objectives in real time. For the digital twin application, the tool examines how changes in external conditions influence product performance. By integrating sensitivity analysis with the graphical tool, engineers can improve decision-making by anticipating how external modifications may degrade performance. At the core of the method, a Monte Carlo simulation is combined with traceability graph results to evaluate the impact of variable modifications on design objectives. This innovative approach supports the development of preventive and predictive maintenance processes in mechanical systems.