Fuzzy rule-based prediction of lovastatin productivity in continuous mode using pellets of Aspergillus terreus in an airlift reactor

Lovastatin production using pellets of Aspergillus terreus wasinvestigated in an airlift reactor. A fuzzy system has been developedfor predicting the lovastatin productivity. Analysis of the effect ofdilution rate and biomass concentration on the productivity oflovastatin was carried out and hence these were taken as inputs forthe fuzzy system. The rule base has been developed using theconceptions of developmental processes in lovastatin production.The fuzzy system has been constructed on the basis of experimentalresults and operator’s knowledge. The values predicted forlovastatin productivity by the fuzzy system has been compared withthe experimental data. The R squared value and mean squared errorhas been calculated to evaluate the quality of the fuzzy system. Theperformance measures show that the rule-based results of the fuzzysystem is in accordance with the experimental results. Theutilization of fuzzy system aided in the increase of lovastatinproductivity by about 1.3 times when compared to previousempirical experimental results.