Non-parametric multi-linear frequency-domain modelling of a wave energy device from experimental data

Abstract

In experimental wave tank tests of wave energy converter devices, uncertainty sources arising from the inherent nature of the dynamical system or from external factors are inevitable. It is essential to employ proper signal processing techniques to mitigate these effects and establish realistic WEC models. Careful estimation of a range of representative linearised models, or determination of a nominal model and associated uncertainty region, overcomes the limitations of a single linear model by covering the complete dynamic operational space of the device. In pursuit of this goal, the current study analyses physical wave tank data using force-input experiments for a scale Wavestar prototype. Non-parametric, multi-linear hydrodynamic WEC models are characterised, serving as a starting point for model-based linear WEC robust controller synthesis. Specific focus is directed towards the quantification of uncertainty and external disturbance, specifically tailored to the wave tank under study. Effective data pre- and post-processing steps, adopted for chirp-type signals, are applied to mitigate disturbances while preserving the underlying physical system behaviour. Moreover, as an application example of WEC models obtained from this study, a single nominal model and uncertainty bound are determined that could provide the basis for (linear) robust WEC control.