DECENTRALIZED SELF-TUNING CONTROL WITH MULTI-STEP PREDICTION OF LABORATORY APPARATUS

PERŮTKA, K.; BOBÁL, V.; DOSTÁL, P.

Abstract

The paper deals with decentralized autonomous control of MIMO (multi input multi output) systems employing a set of SISO (single input single output) controllers running simultaneously. The algorithm of controller design has incorporated on-line recursive identification, together with the method of evaluating the parameters of controller, so far known as self-tuning control. The instrumental variable approach was used as the method of identification and the linear time-invariant sub-systems were of the second order without time delay. The computation of action signals uses the future estimated values of the action signals as well as the future estimated values of the output signals from the system on the chosen horizon. The verification was realized in MATLAB environment on many systems, one is chosen for this paper. A laboratory model was chosen to verify the used algorithm in pilot plant of real-time world.

Coresponding author e-mail: kperutka[at]ft[dot]utb[dot]cz

Session: Robust and Adaptive Control