On-Line State Estimation and Parameter Identification for Batch Fermentation

Abstract

A sequential state and parameter identification technique is applied to batch beverage fermentation. The algorithm uses an extended Kalman filter (EKF) for state estimation and a recursive prediction error method (RPEM) for model parameter identification. This adaptive algorithm is tested using both simulated and real process data. Results show that, even with imperfect models, both states and parameters are estimated well enough to reliably track the dynamics of the true system.