Inferential control of product properties for fluidized bed spray granulation layering

Abstract

Fluidized bed layering granulation has found application in a wide range of industrial processes, e.g. food, pharmaceutical or fertilizer manufacturing. The quality of the produced granules is a critical factor for subsequent processing steps and depends on individual particle properties like characteristic size, porosity or mechanical stability. Therefore, these properties have to be monitored closely during the process to enable accurately timed intervention by plant operators or the application of suitable feedback control algorithms. In case certain properties are not measurable due to financial or technical limitations, observer techniques can fill the gap and provide reliable estimates for these properties. In this manuscript, observer design is demonstrated and validated for a fluidized bed layering granulation process model represented by a complex set of partial and ordinary differential equations. Furthermore, an inferential integral state feedback controller is designed and applied to the complex nonlinear model.