FRACTAL CHARACTERIZATION OF FERMENTATIONS WITH BUDDING YEAST UNDER DETERMINISTIC AND CHAOTIC CONDITIONS

Abstract

Time series data of many biological processes are amenable to fractal analysis. In this study, fractal dimensions were calculated for continuous cultures of Saccharomyces cerevisiae with (1) no noise in the feed stream, (2) noise inflow but no filtering, and (3) with noise and an extended Kalman filter. Each situation was analyzed for dilution rates and the gas-liquid mass transfer coefficients for oxygen under which monotonic, oscillatory and chaotic outputs were generated. The fractal dimensions were compatible with the nature of the time-domain profiles and were sensitive to changes in the profiles and to the presence or absence of noise and a filter. Together with similar results for Escherichia coli, the fractal dimension thus appears to be a convenient and reliable index of fermentation performance.