MPC based on ARX-Chebyshev model for temperature trajectory control in a batch reactor

The control systems used in batch reactors for heating processes can be optimized by using advanced control techniques to allow tracking set-points when they are changing in time. In this sense, we propose to use one scheme Model Predictive Control (MPC) based on Auto Regressive with Exogenous Input (ARX) model projecting the parameters of the classical ARX model onto a set Orthogonal Basis Functions (OBF), thus obtaining a better performance of the new model. This approach enabled to identify the reactor system with a drastically reduction of involved parameters. In addition, the most important achievement was obtained to the represent the nonlinearities and variations of chemical components inside batch reactor. Finally, the theoretical analysis and simulations of the hybrid model have proved that the MPC based on ARX-Hybrid model ensures the trajectory tracking of temperature inside reactor with excellent accuracy and moreover reduces the batch cycle time.