ABSTRACT
This chapter discusses the minimum-variance approach to multivariable self-tuning as proposed by H. M. Koivo and deals with the pole assignment method of D. L. Prager and P. E. Wellstead. M. M Bayoumi and M. A. El-Bagoury introduce a multivariable self-tuning augmented regulator. The multivariable controller is applied to such a complex process as product composition control in a distillation column. Grimble reports about minimum-variance interpretation of linear-quadratic-Gaussian multivariable controllers for self-tuning. Wellstead and Prager discuss the various methods available for multivariable self-tuning with special reference to their relative merits. Prager and Wellstead deal with two examples: one involving simulation of a non-minimum-phase plant with different time delays in the two loops and the other involving laboratory experiments for the control of a hydraulic system comprising two water tanks coupled by an interconnecting orifice.
