Process Monitoring and Fault Detection Using Multivariate SPC
Considerable attention has been devoted in recent years to the problem of fault detection and diagnosis in chemical plants. Faults in process equipment/instrumentation, or within the process itself, can result in off-specification production, an increase in operating costs, the possibility of line shutdown, and the possibility of detrimental environmental impact. Furthermore, prompt detection and diagnosis of process malfunctions are strategically important due to the economic and environmental demands required for companies to remain competitive in world markets. Various approaches have been proposed. These can be broadly divided into model-based techniques, knowledge-based methodologies and empirical techniques (Patton et al., 1989). Model-based approaches generally utilise results from the field of control theory and are based upon parameter estimation, state estimation or the parity space concept (see Chapters 2–9 of this book). The philosophy of the approach is that a fault will cause changes to certain physical parameters and measurements which in turn will lead to a change in some of the model parameters or states.
KeywordsFault Detection Statistical Process Control Principal Component Analysis Squared Prediction Error Warning Limit
Unable to display preview. Download preview PDF.