Adaptive PID control design based on generalized predictive control (GPC)

Since generalized predictive control (GPC) has been introduced, it has been known as an effective tool for the control of many practical systems. In the mean time, the PID controller has always been considered as the most acceptable control technique in the industrial community. Therefore, the underlying practical problem and the aim of this work is to design a PID controller which is equivalent to a GPC controller and incorporates the advantages of both PID and GPC. The PID parameters are estimated based on a predesigned generalized predictive control (GPC). The presented technique can be used both in the adaptive/nonadaptive context. The new technique provides the same performance when tested with various systems with no restrictions on the system order or type.

[1]  David W. Clarke,et al.  Generalized predictive control - Part I. The basic algorithm , 1987, Autom..

[2]  M. R. Katebi,et al.  Predictive PID controllers , 2001 .

[3]  Ole Ravn,et al.  A PID autotuner utilizing GPC and constraint optimization , 1990, 29th IEEE Conference on Decision and Control.

[4]  J. P. Norton,et al.  An Introduction to Identification , 1986 .

[5]  Michael J. Grimble,et al.  Predictive PID controller design , 2001, Proceedings of the 2001 IEEE International Conference on Control Applications (CCA'01) (Cat. No.01CH37204).

[6]  Tore Hägglund,et al.  Automatic Tuning of Pid Controllers , 1988 .

[7]  David H. Owens,et al.  Delayed generalized predictive control , 1998 .

[8]  S. L. Shah,et al.  Development of a stochastic predictive PID controller , 1995, Proceedings of 1995 American Control Conference - ACC'95.