On the role of modelling uncertainty in optimal control and performance assessment of wave energy conversion systems

Modelling uncertainty can play a fundamental role in optimal control design for wave energy converters (WECs), with poten tially drastic results predicted within the literature, if not consid ered and incorporated into the synthesis procedure appropriately. Most of these hypothesised situations, nonetheless, are conducted on the basis of modelling uncertainty generated synthetically, e.g. by assuming numerically produced parameter/system variations, precluding a realistic assessment of the effect of uncertainty aris ing from actual experimental conditions. To fill this gap, within this paper, modelling and uncertainty characterisation are per formed fully based on experimental data, with tests designed specifically to extract information on the response of a proto type WEC system under different operating conditions. With the computed family of models, the energy-maximising control problem is solved by leveraging so-called moment-based theory, in both nominal and robust control frameworks.