Virtual Sensing Application Cases Exploiting Various Degrees of Model Complexity

Abstract

Virtual sensing is an emerging technology with a wide application potential for the industry. The combination of sensor measurement data with simulation models allows the estimation of physical quantities, of which a direct measurement is not feasible, for reasons of, e.g., cost and accessible space. The estimated physical quantities can be system inputs and responses. The objective of this work was to investigate and validate the technology on industrial setups. Three virtual sensing demonstrators were developed, showing the validity and application potential of the technology on use cases with different degree of system (model) complexity.