Parallel Wiener-Hammerstein System

A Parallel Wiener-Hammerstein system is obtained by connecting multiple Wiener-Hammerstein systems in parallel. Each parallel branch contains a static nonlinearity that is sandwiched in between two linear time-invariant (LTI) blocks. The presence of the two LTI blocks, and the parallel branches results in a problem that is harder to identify.

The provided data was part of a previously published Automatica paper available online at Sciencedirect or as an ArXiv preprint. The Parallel Wiener-Hammerstein system, the measurement setup, and the input signals used are detailed in Section 10 of the aforementioned paper. The provided Parallel Wiener-Hammerstein datasets are available for download here. This zip-file contains the multisine estimation data set, the multisine and an increasing-amplitude test data set. The data is available in the .csv and .mat file format.

Special thanks to Johan Pattyn for designing the parallel Wiener-Hammerstein system.

A two-branch Parallel Wiener-Hammerstein system: the red blocks represent linear time-invariant dynamics, while the blue blocks respresent a static nonlinearity.


Please refer to the Parallel Wiener-Hammerstein benchmark dataset as:

M. Schoukens, Parallel Wiener-Hammerstein Time Series, 4TU.ResearchData, 15-Sep-2020, doi: 10.4121/12950081.

M. Schoukens, A. Marconato, R. Pintelon, G. Vandersteen and Y. Rolain, Parametric identification of parallel Wiener–Hammerstein systems, Automatica, vol. 51, pp.111-122, 2015, doi: 10.1016/j.automatica.2014.10.105.