Cybernetics and Information Technologies
Volume 5, No 2. Sofia, 2005, Bulgarian Academy of Sciences

Predictive Functional Control Using a Blending Approach

Mincho Hadjiski*, Ventsislav Asenov**

* University of Chemical Technology and Metallurgy, 8 Kl. Ohridski Blvd, 1756 Sofia
** SyscontLtd. Iskarsko Chaussee 14, 1592 Sofia

Abstract: A new strategy for controlling nonlinear parametrically dependent plants is proposed in order to ensure a good control performance and keep up the simplicity of Predictive Functional Control (PFC). Two alternative approaches are considered to realize a quasi linear PFC: manipulated variable blending and neighboring models blending. In this way it is avoided a nonlinear (NP) problem solving at each sampling time instant. Very simple no iterative calculations are needed. This implies that the computational load is slightly heavier in comparison with linear PFC (LPFC). The modified quasi-linear PFC algorithms are applied to a pH waste water neutralization laboratory stand. A comparative analysis is carried out between the proposed blending alternative approaches. Their performance is also compared with PFC and PID control schemes.

Keywords: Predictive Functional Control, blending, pH neutralization, nonlinear parametrically dependent plant.