On-Line Adaptive, Optimal Control of a Fed-Batch Fermentation of Streptomyces C5

Streptomycetes are microorganisms of commercial importance, but much of their biology is poorly understood. Traditional techniques of growing them in submerged culture have depended upon reproducing previous successful fermentations by following preestablished, empirically-determined paths. Inadaquate instrumentation and such open loop control approaches have left little ability to adapt to process changes. This research has addressed these problems through application of on-line optimal control based upon estimates of eight state variables and parameters obtained from an extended Kalman filter. Simulation and experimental results are described and compared.