| 초록 |
Bioreactors are not easy to model like simple chemical reactors due to the nature of the system. In particular, because modeling the kinetics of reacting organisms depends a lot on empirical equations such as monod equation, resulting in unsatisfactory results in many aspects such as generality and performance of the model. In this study, kinetic models of organisms were obtained from reactor operating data using an artificial neural network and implemented in a black box model. Combining the model with a knowledge based white box model, the bioreactor was implemented as a hybrid gray box model. After validating the model with test sets, dynamic optimization of the bioreactor feed rate was performed using the control vector parameterization technique using the genetic algorithm. The results were compared with those obtained using the white box model. |
