Abstract

Batch bioprocesses are difficult to model due to strong nonlinearities, dynamic behaviour, lack of complete understanding and unpredictable disturbances. A cell produces more cells, chemical products and heat from chemical substrates. Typical growth characteristics include several phases whose appearances and lengths depend on the type of organisms and the environmental conditions. Large differences exist between different fermentation runs. The simulator developed for fed-batch fermentation processes consists of three interacting dynamic models, each with three phase specific versions. The models predict dissolved oxygen concentration, oxygen transfer rate and concentration of carbon dioxide in the exhaust gas through the whole process, by using only the control variables as inputs. A decision system based on fuzzy logic to provide smooth gradual changes between phases. The detection of the changes between process phases is improved by using the intelligent trend analysis. The dynamic simulator is suitable for an online forecasting tool in connection with the real process. The operation is based on the ideas of model predictive control (MPC): the previous online measurements on a chosen horizon are used for constructing a starting point and the simulator predicts the operation on a chosen prediction horizon by using the planned control actions. The simulation is started on fairly long time intervals.