Control of Batch Evaporative Crystallization of Sugar by Means of Adaptive Simulation

The on-line simulation presented in this paper involves balance orientated modeling principles combined with suitable phenomenological expressions, which tie the balance contributions together, thus reducing the number of needed measurement variables. However, as the phenomenological expressions often are inaccurate, they can be provided with open parameters, which continuously are estimated from the real process environment. In this way the model has become an adaptive on-line model. As an adaptive model is able of predicting future behaviour very accurately even of strongly non linear processes, it can be utilized for computing the set-values for control loops, which may be strongly coupled. The use of the presented adaptive simulation is demonstrated by means of a developed adaptive on-line simulator connected with a batch evaporative pilot plant crystallizer for sugar. Special attention has been paid to the automation of this non linear process.