Brief paper: An algorithm for bounded-error identification of nonlinear systems based on DC functions

  • Authors:

  • J. M. Bravo;T. Alamo;M. J. Redondo;E. F. Camacho

  • Affiliations:

  • Departamento de Ingeniería Electrónica, Sistemas Informáticos y Automática, Universidad de Huelva, Carretera Huelva - La Rábida, Palos de la Frontera, 21071 Huelva, Spain;Departamento de Ingeniería de Sistemas y Automática, Universidad de Sevilla, Camino de los Descubrimientos s/n. 41092 Sevilla, Spain;Departamento de Ingeniería Electrónica, Sistemas Informáticos y Automática, Universidad de Huelva, Carretera Huelva - La Rábida, Palos de la Frontera, 21071 Huelva, Spain;Departamento de Ingeniería de Sistemas y Automática, Universidad de Sevilla, Camino de los Descubrimientos s/n. 41092 Sevilla, Spain

  • Venue:
  • Automatica (Journal of IFAC)
  • Year:
  • 2008

Quantified Score

Hi-index 22.14

Visualization

Abstract

This paper presents a guaranteed method for the parameter estimation of nonlinear models in a bounded-error context. This method is based on functions which consists of the difference of two convex functions, called DC functions. The method considers DC representations of the functional form of the dynamic system to obtain an outer bound of the set of parameters that are consistent with the measurements, the system and the considered bounded error. At each iteration, the proposed algorithm solves several convex optimization problems to discard from the initial search region subregions that are proved not consistent. This operation is repeated while the obtained solution is improved. Four examples are provided to clarify the proposed identification algorithm.