Model predictive controller design for uncertain bilinear systems and an application to control a continuously stirred tank reactor with Hydrolysis Acetic Anhydride ☆

The paper proposes a predictive control method for a bilinear MIMO system with slow varying uncertain parameters. This proposed predictive control method operate under the principle setting predictive model based on a linearized along the trajectories of the bilinear system and estimated uncertain parameters. To increase the quality of follow the setpoint signal, a compensation input control signal component has also been added. The input compensation signal is added to improve the tracking problem. The input compensation is designed based on an adaptive mechanism. An application of the control algorithm in continuous stired tank reactor with hydrolysis acetic anhydride is given to illustrated the tracking performance.