Abstract

The problem of system identification from a time series of measurements is solved by using nonparametric additive models. Having only few structural information about the system, a nonparametric approach may be more appropriate than a parametric one for which detailed prior knowledge is needed. Based on nonparametric regression, the functions in the additive models are estimated by a penalized least-squares approach using backfitting. The optimal smoothing parameters are determined via generalized cross validation, making this approach completely adaptive to the data. The procedure is applied to identify the nonlinear restoring force of vibrationally excited helical wire rope isolators.

Citations