MULTIVARIABLE PROCESS IDENTIFICATION FOR MPC - THE ASYMPTOTIC METHOD AND ITS APPLICATIONS

In this work we will introduce the asymptotic method (ASYM) of identif ication and provide two case studies. The ASYM was developed for multi variable process identification for model based control. The method ca lculates time domain parametric models using frequency domain criterio n. Fundamental problems, such as test signal design for control, model order/structure selection, parameter estimation and model error quant ification, are solved in a systematic manner. The method can supply no t only input/output model and unmeasured disturbance model which are a symptotic maximum likelihood estimates, but also the upper bound matri x for the model errors that can be used for model validation and robus tness analysis. To demonstrate the use of the method for model predict ive control (MPC), the identification of a Shell benchmark process (a simulated distillation column) and an industrial application to a crud e unit atmospheric tower will be presented. (C) 1998 Elsevier Science Ltd. All rights reserved.