Robust Identification in Nonlinear Dynamic Process Models

 

María Rodríguez-Fernández, Antonio A. Alonso and Julio R. Banga

Process Engineering Group, IIM-CSIC (Spanish Council for Scientific Research), C/Eduardo Cabello 6, 36208 Vigo (SPAIN) mrodriguez@iim.csic.esantonio@iim.csic.esjulio@iim.csic.es

ABSTRACT. Parameter estimation is a key issue in the mathematical modelling of nonlinear dynamic processes. Standard (gradient-based) methods for data fitting in nonlinear dynamic systems can suffer from slow and/or local convergence, among other problems. However, this is frequently ignored, potentially leading to wrong conclusions about the validity of a model regarding a certain data set. In order to surmount these difficulties, we present alternative methods based on global optimisation and identifiability analysis.

KEYWORDS: parameter estimation, inverse problem, global optimization, sensitivity analisys, identifiability analysis

1. Introduction

Building sound dynamic models is a core task in modern computer-aided process engineering. Model building is usually divided in two tasks: definition of the model structure, and parameter estimation. The latter, also known as model calibration, is a key step in the development of reliable dynamic models. Given a model structure and a set of experimental data, the objective of parameter estimation is to calibrate the model (looking for parameters which can not be measured directly) so as to reproduce the experimental results in the best ...

Get Taming Heterogeneity and Complexity of Embedded Control now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.