Adaptive estimation of discrete-time systems with nonlinear parameterization (Q5926192): Difference between revisions

The paper deals with parameter estimation for discrete-time dynamic systems with bounded disturbances and affected in a nonlinear way by the parameters. A new algorithm is proposed which leads to a globally stable estimation procedure. It is based on the solution of a min-max optimization problem which provides appropriate regression vector and step size for the estimator. Under suitable conditions parameter convergence is also guaranteed.