An adaptive controller for batch feed bioprocess application to lysine production

This paper presents an adaptive algorithm designed for batch feed bioprocess control, The proposed algorithm is based on a non-linear ‘grey box’ model, which is adapted on-line to the process behaviour by the estimation of a key parameter. Stability and convergence analysis of this algorithm is presented, and from this, a practical tuning method is given. The proposed algorithm is compared to a standard adaptive generalized predictive controller approach, and is shown to exhibit similar performance while being easier to use. A real application of this adaptive algorithm to a batch feed lysine process is presented. Good results are obtained, which show that this control scheme is a worthwhile alternative for batch feed bioprocess control.