Authors: M. Keerthana, T.K. Radhakrishnan and N. Sivakumaran
Abstract: The coupling of reaction and separation in reactive distillation has lead to high nonlinearity in the reactive distillation. Model based predictive control schemes are preferred for the control of nonlinear process. Orthonormal basis filter (OBF) models are more flexible and comprehensive, to capture the dynamics of a process with few numbers of terms. Among the various types of orthonormal basis filters, laguerre filters are more appropriate for well damped processes with incorporation of a single real pole. These filters cannot be used when the system has complex poles. Generalized Orthonormal Basis Filter (GOBF) models can overcome the drawback of laguerre networks by allowing non-identical poles and complex poles in the construction of the networks. The response of GOBF model based predictive controller for set point and load changes has been studied and a satisfactory performance is ensured with the controller.
Keywords: generalized orthonormal basis filter, model predictive control, orthonormal basis filter, reactive distillation