Title :
State space identification of bilinear canonical forms
Author_Institution :
Duisburg Univ., Germany
Abstract :
This paper deals with the identification of nonlinear state space models from input/output data. For bilinear systems four canonical forms are introduced which are bilinearizations of the general nonlinear forms. Canonical forms are well suited for the automated computer aided controller design. For these bilinear forms parameter identification methods are derived, which also enable the identification of the output vector. These methods use a technique to identify the parameters of the canonical forms directly. The convergence is improved by using a two-step procedure. The results of the identification of an electro-hydraulic drive are presented.
Keywords :
State estimation; linear systems; nonlinear systems; parameter estimation; state estimation; state-space methods; I/O data; bilinear canonical forms; computer aided controller design; control system CAD; electrohydraulic drive; input/output data; nonlinear state-space models; parameter identification methods; state-space identification;
Conference_Titel :
Control, 1994. Control '94. International Conference on
Conference_Location :
Coventry, UK
Print_ISBN :
0-85296-610-5
DOI :
10.1049/cp:19940241