Abstract
AbstractUsing a novel, continuous flow‐cell mounted in a spectrophotometer, various digital control algorithms have been experimentally investigated to control biomass concentration within a turbidostat. The manipulated variable was a stream of pure, sterile water while disturbances were generated in the glucose feed stream.Biomass concentration was found to be related to the controlled variable by a simple first order relation with maximum least squares parameter deviations of 20% when subjected to a variety of disturbances. After two trial and error attempts, the PI algorithm was found to be poorly tuned and gave unstable and oscillatory control. Using the process reaction curve results, the Dahlin's method was found to be an excellent tool for tuning the PI algorithm and gave good to excellent control for each of two adjustments of its tuning parameter. The IMC algorithm was also successfully used to control biomass but it was necessary to adjust the sampling time up from 5 seconds to 5 minutes to overcome the instability problems generated by the sensitivity of its tuning parameter at low sampling times.Finally, it is shown that a mechanistic model employing Ghose growth biokinetics for Saccharomyces cerevisiae simulated the controlled, transient nature of the bioreactors with a good degree of accuracy.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call