%0 Journal Article
%@ 1312 – 451X
%A Patnaik, P R
%D 2006
%F open:1029
%I BAS, Centre of Biomedical Engineering
%J Bioautomation
%K Microbial oscillations, Saccharomyces cerevisiae, Bioreactor, Noise inflow,  Neural filters.  Introduction  Many microbial processes exhibit sustained oscillations over long durations. Under controlled  conditions, clear oscillations are observable. In more realistic natural environments and  production processes, however, the influx of noise obfuscates the intrinsic oscillations from  the aberrations caused by noise. Since the occurrence of oscillations is linked to the reactions  inside the cells and to transport processes across the cell walls [22], the identification of the  oscillating signals is important for the understanding and control of large bioreactors [6, 15,  34].  The bacterium Zymomonas mobilis and the yeast Saccharomyces cerevisiae have been the  work-horses of most studies of oscillating phenomena. S. cerevisiae is more popular in view  of its ease of cultivation, well-understood physiology and industrial importance [7, 22]. Two  recent publications [23, 24] have addressed the issue of recovering smooth oscillations from  noise-distorted concentration profiles during continuous fermentations with S. cerevisiae.  Both studies were based on experimental observations [1, 7, 16, 27] that, in certain ranges of  the dilution rate and the gas-liquid mass transfer rate of oxygen, continuous cultures of S.  cerevisiae display oscillating profiles for some key concentrations such as those of the  biomass, carbon substrate (glucose), product (ethanol), storage carbohydrate and dissolved  oxygen. Different types of oscillations occur in different ranges of these two manipulated  variables, and some oscillations may comprise a superposition of two or more simple  unimodal oscillations of different amplitudes.
%P 45-56
%T Neural network configurations for filtering of feed stream noise from oscillating continuous microbial fermentations
%U http://crdd.osdd.net/open/1029/
%V 4
%X Some microbial systems exhibit sustained oscillations under certain conditions.  The maintenance and the suppression of oscillations are both important in different  situations. While oscillations are clearly identifiable in small bioreactors, the influx of noise  fuzzifies the oscillations in larger vessels. So, noise-filtering devices are employed to recover  clear oscillating profiles. Recent work has shown that an auto-associative (AA) neural  network is a better than standard algorithmic filters. In this study, nine neural network  designs are compared for their ability to filter Gaussian noise in the substrate inflow rate of  a continuous fermentation containing Saccharomyces cerevisiae. While the AA network is  the best overall, specific performance criteria favor other designs. Thus the choice of a  neural filter depends on the evaluation criterion, which is guided by the application.
%Z OPEN ACCESS