TOWARDS SMART DEVICES: ENHANCING MEASUREMENTS BY AUTOMATED DATA RECONCILIATION

摘要

Data reconciliation is a technique to enhance measurement data by exploiting the a priori structure of the process plant. However, in many cases this structure is not exactly known. In this paper, possibilities for exploiting the multivariate statistical methods are studied for constructing a dynamic data reconciliation scheme applying subspace identification and Kalman filtering. As an application example, X-ray analyzer operating in a zinc flotation circuit is studied.