USING MEASUREMENT DATA IN BIOPROCESS MODELING AND CONTROL

Attempts to improve quality control in complex bioproduction processes require the efficient use of as much knowledge about the process as is available. Knowledge of the process conditions, and the relationship between process variables and product quality may be expressed by a process model. This article reviews methods that aim to make better use of empirical data, or of process knowledge derived from such data, in order to develop and improve the models. Application of these methods leads to hybrid process models, which combine mathematical models with artificial neural networks and fuzzy expert systems.