CHALLENGES OF MODEL PREDICTIVE CONTROLLERS FOR INDUSTRIAL CONTROL SYSTEMS

Model Predictive Control (MPC) has become one of the most effective methods for solving complex problems in industrial control systems because it performs well with multivariable systems, handles constraints and objectives over a control space, and enhances performance over a given time interval. This is made possible by the fact that it allows to foresee how the system will behave in the near future which is an ideal situation for complicated processes where conventional control techniques may fail. However, value addition does not come without challenges. Those challenges are associated with, existence of an accurate system model, availability of computing power to enable real time optimization routine, disruption and changing environmental uncertainties. Integrating MPC in previously established control systems in large industries can at times be an obstacle on its own. This article, seeks to address the gap that touches core areas of MPC for complex industrial systems i.e modeling issues, limitations that exist in computational resources, real time requirements, how appropriate constraints can be put in place, feasible scalability of the system in place, connecting the new system to the existing one in a seamless manner among other elements. In addition, these papers consider also possible ways out and the latest trends of the development of MPC methods, robust and adaptive MPC, parallel computing, hierarchical control frameworks and neural networks based system identification are examined. Keywords: Model Predictive controller, İndustrial Systems, Stability ,Robustness.