Collect. Czech. Chem. Commun. 1994, 59, 595-602

Bioengineering Analysis of Incomplete Data from Large-Scale Anaerobic Bioreactor for Waste Water Treatment by Fuzzy Expert System

Vu Lan Huonga, Jaroslav Votrubaa and Ivan Stuchlb

a Institute of Microbiology, Academy of Sciences of the Czech Republic, 142 20 Prague 4, Czech Republic
b Sugar Research Institute, Cukrspol, a.s., 140 00 Prague 4, Czech Republic


The knowledge-based model of an anaerobic waste water treatment plant was designed when the incompleteness of industrial records excluded the application of mass balances. New two step information technology that permits to excerpt the knowledge from industrial data as a set of heuristic rules is described. In the first step, the real numbers are converted to the set of fuzzy ones and arranged as a knowledge data base of an expert system. Secondly, the fuzzy expert system SENECA was used for fuzzy modelling of the process. Three examples of bioengineering analysis are presented. First, the average efficiency of biological degradation was determined. Second, the effect of pH in the input stream was evaluated. Third, the process dynamics and time required to reach the maximum activity of microbial population were estimated. It was possible to detect and evaluate the decrease of the biodegradation efficiency caused by unexpected leakage of xenobiotics.