| Abstract
| - In this paper, adaptive supervisory predictive functional control is discussed for applications in a fed-batchreactor with slow continuous mixing valve, which acts as an actuator. The optimal operation regime of thereactor is to follow the reference trajectory without significant overshoot and to minimize the overall batchtime. By adding the ingredients during the batch procedure, the volume in the reactor varies, tending toincrease with time, as does the heat-transfer surface inside the reactor. This change causes variations in thebatch-reactor dynamics, which can produce serious problems with the reactor's core-temperature control.Another problem that arises is the disturbance in the temperature inside the reactor, when additional ingredientsadded are at the outside temperature. In the pharmaceutical industry, a huge number of fed-batch reactorssimilar to ours are used for the production of drugs. The quality control is extremely strict, and it is necessaryto follow the exact production recipe. The adaptive algorithm based on recursive identification is used to dealwith varying dynamics of the systems; the predictive functional control algorithm is used because of itssimple implementation, which can be realized also in programmable logic controllers; and the supervisorycontrol level is designed to cope with the physical constraints and the discrete on/off valves that are used asactuators. The possible use of the proposed adaptive control algorithm in a real application was studied bytaking into account a very slow actuator.
|
