Performance Evaluation of a Distributed MPC Strategy Applied to the Continuous Stirred Tank Reactor

Abstract

The search of more robust solutions for multiple-input multiple-output (MIMO) process control has motivated the development of complex modeling and control techniques for such systems. In this work the performance of a distributed controller (Distributed Model Predictive Control DMPC) versus a centralized solution (Model Predictive Control, MPC) are compared when faults in sensors and actuators occur. Preliminary results show that for faults in actuators the distributed solution has a better response but it requires more calculation time.