Stability and Control of Complex Nonlinear Systems with Application to Chemical Reactors

This paper introduces a robust strategy for stabilizing complex industrial processes exemplified by a complex chemical reactor. The design uses a combination of backstepping methods and Lyapunov-based techniques for implementing a robust feedback controller. A model-reference-like virtual system is proposed to accomplish both asymptotic stability and satisfactory transient performance. The controller is robust in the sense that it accommodates uncertainties inherent in complex industrial processes. The suggested design procedure also has the advantage of being applicable to non-linear processes without having to carry out linearization approximations. A simulated continuous stirred tank reactor is used to exemplify the suggested technique and the superiority of the controller is demonstrated by comparing its response to a standard PID controller. Practical considerations are addressed to investigate the causality and feasibility of the proposed controller and its applicability to real-time situations. Tradeoffs between stability and performance are carefully studied.