Forecasting Framework for Digital Twins

Abstract The ability to forecast system conditions is integral to the definition of digital twins. However, the strategic implementation and associated assumptions underlying the forecasting technique have not yet been fully defined and integrated with a physical system. This research develops a generic forecasting framework and baseline assumptions that can be adapted to a wide variety of applications using digital twins. Experimental validation of the forecasting methodology is demonstrated through a real-time electro-thermal digital twin designed for onboard power distribution cables as an illustrative example. The digital twin utilizes real-time sensor data to forecast thermal profiles, facilitating proactive power management and enabling real-time decision-making. Forecasted insights derived from the digital twin allow for timely adjustments in power flows, thereby preventing the cables from exceeding selected thermal thresholds. The efficacy of the forecasting framework is experimentally verified in a controlled three-bus power system configuration.