Data reconciliation & gross error detection: an intelligent use of process data
Data reconciliation & gross error detection: an intelligent use of process data
Hi-index | 0.00 |
This paper compares the techniques of Data Reconciliation and the Diagnostic Model Processor applied to an industrial full scale two-ply paper machine using actual operating data. Data reconciliation optimally adjusts the raw data to satisfy known constraints whilst simultaneously identifying gross errors. The DMP searches for faults that create the observed discrepancies in these constraint equations. The DR strategy worked well, reconciling the raw measurements and correctly identifying gross errors while the DMP was over enthusiastic in its attempts to identify assumption violations.