When process monitoring indicates abnormal process operation, diagnosis activities are initiated to determine the source causes of this abnormal behavior. Experienced plant personnel have good insight in integrating various pieces of information provided by process measurements to determine the cause(s) of a fault. Various fault diagnosis paradigms can automate this effort and provide timely information to plant personnel about the most likely causes for abnormal operation. Reduction of down time, fixing the actual problem (as opposed to a secondary fault), and scheduling regular maintenance as opposed to doing emergency repairs contribute significantly over time to the profitability of the process.
The book introduces many alternative fault diagnosis paradigms and illustrates some in more detail through case studies. It also proposes the integration of monitoring and diagnosis activities by linking SPM tools with a real-time knowledge-based system that acts as a supervisor of process operations and fault diagnosis agent. Fault diagnosis techniques and knowledge-based systems are covered in Chapter 8. A forward-looking proposal for further integration of monitoring and diagnosis with control system performance assessment, supervisory control of batch fermentation process operation, and plantwide decision making systems is presented in Chapter 9.
Was this article helpful?