In the paper a three-stage, semirecursive scheme for the Wiener-Hammerstein system identification is proposed. The algorithm combines both parametric and nonparametric strategies and allows to recover linear and nonlinear subsystems directly from the noisy inputoutput data. As to the nonlinearity, the main idea is based on the recursive kernel censoring of measurements, while linear dynamics are recovered by a special kind of deconvolution. Efficiency of the obtained estimates is justified by numerical example.
CITATION STYLE
Hasiewicz, Z., Wachel, P., Mzyk, G., & Kozdraś, B. (2017). Multistage identification of Wiener-Hammerstein system. In Advances in Intelligent Systems and Computing (Vol. 577, pp. 527–535). Springer Verlag. https://doi.org/10.1007/978-3-319-60699-6_51
Mendeley helps you to discover research relevant for your work.