Hybrid Modeling for Nosiheptide Fermentation Process Based on Prior Knowledge and SVM

Fermentation is always characteristic of multiple non- linearity and time-varying. In order to know and control the fermentation process, many process models are used to formulate knowledge about process behavior. They are applied, e.g., to predict the process´ future behavior and for state estimation when reliable on-line measuring techniques to monitor the key variables of the process are not available. The sources of information available for modeling come from two ways. First, some equilibrium equation can be described by first principles or obtained rules of thumb. Secondly, some information may still be hidden in the process data recorded during previous runs of the process. This work proposes a hybrid modeling strategy and a correcting method on-line. This hybrid model is a combination of classical fermentation models and support vector machines modeling based data. In case study, the hybrid model is applied in the Nosiheptide fermentation process. Results show that this method has a good estimation performance.