Grey prediction fuzzy control for pH processes in the food industry

Abstract

Proper regulation of pH value is an important issue in the food industry for quality production. A food pH process usually has non-linear dynamics with system uncertainty. This study treats the pH regulation process of a reactor tank as a grey box with partially known system information. The behaviour of the process is predicted one-step ahead with a first-order grey model. A fuzzy controller takes the prediction together with the current system response to regulate the discharge of base (NaOH) or acid (HCl) solution into the reactor tank to arrive at a desired pH value. The integrated grey prediction fuzzy control (GPFC) strategy is simple in control-law derivation and system implementation and is efficient in computation. The developed GPFC was validated with perform base/acid titration and continuous acidification/deacidification control. The controlled system response error was trivial in the titration and was less than 1% in the continuous control under proper agitation of the reactants. The system was used to control Acetobacter xylinum fermentation for cellulose production. The GPFC scheme exerted smooth control action, achieved a trivial steady-state error in pH control, and yielded more cellulose and acetic acid products but consumed much less material than PID or manual control.