Process systems engineering provides automation and decision making capabilities by using mathematical models to formulate smart policies on the basis of predictions of future outcomes. A broad range of tools and techniques are available for this type of analysis and selection is determined by the analytical maturity as well as the specific requirements of the problem being solved.
Although predictive controls and estimators can be put to use in many applications, we outline a few examples where predictive applications have shown positive impact in recent years.
Advanced Process Control
APC applications adjust manipulated variables to maintain target levels of controlled variables. Improving the accuracy of infrequent measurements, noisy measurements, or inferred process states allows the APC application to better meet the defined objective. APMonitor provides optimized information to the APC application.
Differential algebraic equation models of the blood-glucose response have enabled researchers to better tailor medical treatments for insulin dependent patients. Infrequently measured blood glucose levels are combined with information about recent food consumption to gain a real-time estimate of the blood glucose level.
Chemical and Refining Facilities
Modern chemical plants are well instrumented to provide alarms, operators, advanced control applications with real-time information about process conditions. The operators or applications adjust manipulated variables to maintain product quality specifications, reject disturbances, and avoid major process upsets. The measurements provide insight into a process that is constantly changing due to unmeasured disturbances, product grade transitions, or process upsets. In these circumstances, the person or application that is controlling the process needs accurate information about key process variables. APMonitor is an optimization-based application to estimate the current state of the plant. This information is presented to operators or applications to help them gleen the most from available measurement information.