New tuning and identification methods for unstable first order plus dead-time processes based on pseudoderivative feedback control

The use of pseudoderivative feedback (PDF) in the control and identification of unstable first order plus dead-time (UFOPDT) processes is investigated. Several new methods for tuning the PDF feedback controller are presented. In contrast to known tuning rules for conventional proportional integral derivative (PID) controllers, which result in excessive overshoot in the closed-loop response, the proposed control structure and tuning methods ensure a smooth response to set-point changes, fast attenuation of step-load disturbances, and satisfactory robustness against parametric uncertainty. Moreover, two simple methods for identifying the UFOPDT process parameters, based on this controller structure, are proposed in this paper. Both methods rely on a single experiment on a closed-loop system with a step change in the set point of a PDF controller. They are very accurate, as well as simpler and less sensitive than existing identification methods. Finally, an application of the proposed identification and tuning methods to an open-loop unstable bioreactor with hard input constraints and significant measurement delay is presented.