Multiple model-based controller design applied to an electrochemical batch reactor

This study demonstrates the control of an electrochemical batch reactor, that produces the desired product in a competing chemical/electrochemical reaction network, using a multiple model-based controller design. Since this type Of control framework requires a process model to provide predictions of the controlled variables, and because batch reactors are highly nonlinear and nonstationary in nature, a bank of linear, dynamic, state-space models rather than a single; linear state-space or convolution model is developed to represent the nonstationary behaviour of the batch process. It is shown that the performance of this model/controller design can provide good reactor performance in the face of known disturbances.