Methods for a Transparent Development and Optimization of Biotechnological Processes

The article at hand describes an integrating system for the intelligent control of complex bio technological processes including automatic modelling and model based control strategy generation. Starting with a summary of previously achieved results, some new approaches that provide better transparency to process engineers and operators are discussed. This includes aspects of self-organizing generation of structured dynamic nonlinear process models based upon the ideas of genetic programming as well as the transparent generation of fuzzy rules in a particular NeuroFuzzy approach. The latter is used for the classification of physiological states during batch and fed-batch fermentations and for the long time strategy generation to optimize the achievable product yield