ABSTRACT
Some of the basic problems in process industries are the variations in the plant characteristics, which occur due to changing operating conditions and unforeseen disturbances acting on the plant. An adaptive control system is one in which the controller parameters are adjusted automatically in such a way as to compensate for such variations. Studies by simulation on the model of an MSF desalination plant have shown certain features of the model developed on the basis of physical laws and correlations. The most important of these are related to nonlinearity, which results in changing of process characteristics. This study showed a variation in the parameters of the linearized model with the operating conditions. It is clear that a fixed PID controller cannot be optimal if the operating point changes from the one at which the optimal controller was designed. Therefore, adaptive control schemes are used, where the controller parameters are updated to match the plant parameters. In this chapter, two adaptive control schemes are presented. One is based on a linear parameter scheduling law and the other employs an artificial neural network (ANN) to tune the controller. Appendix 8.A provides a brief introduction to the ANN of interest here.
