Further work on self-tuning regulators

Two main disadvantages of the self-tuning regulators are 1. They use a minimum variance controller which produces large control signal movement. 2. The controller exhibits a "burst"4 phenomenon, that is at steady state, the controller output occasionally exhibits large oscillations. Methods to minimize these shortcomings are detailed in this paper. Also a method is presented for minimizing the effect of measurable disturbances on the process output, by including a disturbance in the model and controller equations as a feed forward signal. An iterative method is discussed at the end. Instead of inverting a matrix this method uses an iterative scheme to solve the normal equation. This method does not cause any "burst" phenomenon at steady state and has other advantages. Some real time simulation results and results from real process applications are included.