Multivariate statistical performance monitoring provides a diagnostic tool for the monitoring and detection of process malfunctions for continuous and batch manufacturing processes. The paper initially reviews the concept of process performance monitoring through an industrial application to a fluidised bed-reactor and a comprehensive simulation of a batch methyl methacrylate polymerisation reactor. Finally some of the more recent work being carried out is described including the development of performance monitoring schemes from minimal process data and the development of generic models for the monitoring of multiple products, grades or recipes.
展开▼
