Data Reconciliation for Industrial Experiments

Abstract

The paper presents the use of data reconciliation to bettercharacterize the operating state of experimentalinstallations in an industrial context. The paper focuseson the development of a data reconciliation approach usingthe OpenModelica prototype to study first the detection ofthe defects in an experimental hydraulic test loop, butalso the characterization of good measurement data in anHVAC testing facility. Data reconciliation is shown to beeffective for the most pronounced defects intentionallyintroduced in the system. Regarding the characterization ofmeasurements, data reconciliation identified two initiallyunnoticed invalid experiments.