Control of a continuous fluidized bed dryer using neural nets

A neural network model of a continuous fluidized bed dryer (FED) is used to synthesize an internal model control (IMC) strategy to control the FED. The controlled variable is exit air temperature and manipulated variable is inlet air temperature. Alternatively an IMC structure is formulated using a linear process model. A conventional controller performance is also studied. Using an error criterion the performance of these controllers is compared. The IMC structure with the neural network performs better.<>