Digital control of a rotary dryer in the sugar industry

In order to utilize pulp, which is a by-product of the extraction of sugar from beet, one has to dry it. This is usually done with a rotary dryer. Generally the dry substance percentage of the dried pulp is controlled manually; this, however, is not sufficient. The aim was to improve the quality of control and thereby to reduce the fuel consumption. To reach this goal, first a linear, time-invariant, discrete-time mathematical model of the dynamic behaviour of the drying process was identified off-line. A correlation analysis method with least-squares parameter estimation was used. Based on the identified model, the parameters of the control algorithms of a digital multicascaded control system were designed with computer aid. The control behaviour was simulated and compared with the control behaviour achieved by an optimal state feedback control system. Although the latter one gave a better control performance, the cascaded control system was implemented on a process control computer—this was done for practical reasons. The results and experience gained during three campaigns show that the goal, a significant improvement of the control performance and the resulting saving in fuel oil, was achieved.