In this paper, we present the theoretical development and application of Self-Organizing Map (SOM) based multiple local models (MLLM) for on-line control of fed-batch fermentation process. The recalling and generalization of the model was carried out using simulated data. The prediction of the MLLM was evaluated based on the normalized root mean square error and the result is found to be satisfactory. A feed forward control law was derived for a biomass tracking problem and incorporated with the proposed model to control the fed-batch yeast cultivation. The result observed from the simulation and experimental study illustrates the effectiveness of the model for on-line bioprocess control application.
|Journal||CEUR Workshop Proceedings|
|Publication status||Published - 1 Dec 2007|
|Event||10th International Conference on Engineering Applications of Neural Networks, EANN 2007 - Thessaloniki, Hellas, Greece|
Duration: 29 Aug 2007 → 31 Aug 2007