ON LINE COMPUTER CONTROL OF A HEATED STIRRED VESSEL

Abstract

The present work deals with the application of Self tuning techniques for reference temperature tracking of a C.S.T.R. In industry often a C.S.T.R.( Continuous stirred tank reactor ) is used to carryout series and parallel multiple reactions with a number of products and by-products. Out of these the concentration of the desired product can be maximised by choosing a suitable temperature-time strategy. The adaptive controllers can play an effective role in carrying out the above task. As the parameters of the CSTR change with load and system operating conditions, its dynamic behaviour also changes. The conventional controllers are designed to work most satisfactorily at a particular operating point. For optimum performance it is desirable to .track the system and compute control according to the operating conditions. It is with this aim that the present study on simulation of reference temperature tracking in C.S.T.R. using adaptive control techniques has been undertaken. The algorithms developed can be used On-line to meet the above objective. The Implicit Self tuning controller developed by Clarke and Gawthorp along with Self tuning controller with self searching pole shift developed by Chandra were used mainly to develop the algorithm. The algorithm was tested off-line to tune the controller. The results of the Self tuning controller were compared with conventional PID controller under a sequence of -'dentical step changes. It was found that Clarke and Gawthorp's i R controller was the fastest to acheive the desired temperature but showed more oscillation. Chandra's controller was slower than Clarke's controller but showed less oscillation. The conventional PID controller also showed a reasonable controlability but failed to settle when large step changes for shorter duration was given.