LOCALLY-ACCURATE IDENTIFICATION OF NONLINEAR SYSTEMS USING OUTPUT MEASUREMENTS

Cong Wang,David J. Hill
DOI: https://doi.org/10.3182/20060329-3-au-2901.00196
2006-01-01
IFAC Proceedings Volumes
Abstract:A “deterministic learning„ theory was recently proposed for locally-accurate identification of nonlinear system dynamics with full-state measurements. In this paper, it is shown that for a class of nonlinear systems with only output measurements, locally-accurate identification of the underlying system dynamics can be achieved. Specifically, by using a high-gain observer and a dynamical radial basis function network (RBFN), when state estimation is achieved by the high-gain observer, a partial persistence of excitation (PE) condition can still be satisfied, and accurate identification of system dynamics can be achieved in a local region along the estimated state trajectory. The significance of this paper is that it reveals that the difficult problem in nonlinear observer design can be successfully resolved by incorporating the deterministic learning mechanism.
What problem does this paper attempt to address?