Identification of bioprocesses for adaptive PID tuning

In this work, model based low level control tuning of bioreactors, or, fermentation processes, is studied. The problem at hand is to develop a strategy for the control of a class of reactors of various size and material. Main controlled variables are pH, dissolved oxygen and temperature. Traditional fixed PID control cannot meet the ever-increasing high performance requirements of these processes. Control based on nonlinear first principle model is too costly, due to the problem of diversity. We will propose a linear model based control approach to the problem. The idea is to identify linear local models at various working points of the process and tune local (PID) controllers using the identified model. Closed-loop identification and model based control design are introduced. Laboratory tests and model identification will be performed to verify the feasibility of the proposed methodology.