Comparative Study of Modern Control Techniques for Optimal Dynamic Nonlinear Process Control

Abstract

This paper carries out to review a comparative analysis of modern control with conventional control in search towards an optimal control algorithm for a dynamic nonlinear process control system. Achieving global optimal behavior in closed loop control of a nonlinear system varying considerably in system dynamics over operating range is a challenging target. Improved versions of some of the most effective control algorithms from the modern control, that is Linear Quadratic Regulator (LQR) and Model Predictive Control (MPC), were designed and evaluated by comparison with conventional gain scheduled Proportional-Integral-Derivative (PID) control. The results showed that a new advanced version of MPC that is Switched Model Predictive Control (SMPC) is the most promising controller due to its advanced and benefited features in achieving global optimal dynamic nonlinear process control performance.