Nonlinearity of dynamic characteristics causes various difficulties for the control system of microorganism fermentation system. The fed-batch fermentation method has the advantages that it can prevent substrate inhibition and obtains high cell density.
The objectives of this study are to estimate the cell mass on-line with the data from gas chromatography and to manipulate automatically the feed rate for the cells to grow at the desired growth rate. To accomplish this, it is desirable that no component other than the major carbon substrate is the limiting component for the cell growth. Therefore, the PI controller tuned by fuzzy logic is applied to control the dissolved oxygen concentration, which can be another limiting component if it is not properly controlled.
The specific growth rate could be maintained constant by manipulating the substrate feed flow rate. A manual feed rate control method which utilized the off-line analysis(here, optical density) of samples was applied. On the basis of these results, three control strategies were tested: (i) the estimated specific growth rates were used as the measurements, (ii) the estimated cell masses and specific growth rates were updated to modify the feed rate, and (iii) the estimated cell masses are controlled to follow the constant specific growth rate curve. The third approach is the new one proposed in this study and gave the best result.
