RAGPC control of large delay and retarded systems with optimal model approximant approach

In this paper, we consider the design application of robust anti-windup generalised predictive control (RAGPC) to large delay and retarded systems. These systems are often encountered in industrial and transportation processes and are of practical engineering importance. In order to successfully implement RAGPC scheme on these types of systems, an optimal model approximation technique has been employed to obtain stable approximate rational models for these systems. The optimal model approximant uses a performance index of integral squared error (ISE) between the original delay systems and its approximate rational models. A minimum cost function of the optimization procedure is sought to obtain optimum parameters of the approximate models subject to the constraints that all the poles of the approximate model lie in the open left half of complex plane with zero steady state error in the output responses. Some numerical simulation examples are used to demonstrate the effective application of this newly proposed design scheme.