Automatic Linguistic Inversion of a Fuzzy Model for Fed-Batch Fermentation Control

The goal of the paper is to analyze the potential of linguistic inversion of a fuzzy model to obtain a working controller, which is tested upon a benchmark fed-batch fermentation process. For the automation of the inversion procedure a simple decision mechanism is designed that dramatically increases the efficiency of the approach and results in improved performance compared to our earlier work. Moreover, automation of the decision process allows us to conduct a larger set of experiments to provide evidential material to the statement that approximate reasoning can produce surprisingly good results in the environment resembling real life conditions, however, these results also indicate that for guaranteed performance level, dynamics of the systems must be taken into account in the inversion process