Techno-Economic Optimization Under Uncertainty of a Wave Energy Converter

In the field of renewable technologies, the possibility to obtain energy exploiting seas and oceans’ wave motion has been known for a long time. Devices called Wave Energy Converters (WEC) have been developed with this purpose, thanks to which it is possible to transform wave energy into electric energy. Following the design studies carried out in recent years, the research now proceeds towards the development of useful processes for the optimization of these devices. The purpose of this thesis work is to study, analyze and then apply a robust optimization process to the WEC system, in order to increase its reliability and robustness. Robust optimization is a probabilistic solving method for real-life optimization problems, in which there are uncertain data that, due to their stochastic nature or uncertainty (linked to design condition changes or to wear), can float around their project value. This method studies these parameters and finds suitable solutions with the aim to avoid unsatisfactory system performances and designs which can compromise their functioning, studying and highlighting at the same time the influence that specific parameters can have. Robust optimization considers uncertain data belong to a specific set called “uncertainty set” (with specific constraints), that must be defined by the designer. Therefore, the robust optimization process final purpose is to obtain a set of solutions able to limit, below specific thresholds, the variation of such uncertain data (and consequently their own uncertainty), finding a robust optimum instead of a global optimum in order to be able to increase the reliability and robustness of the system. In this work a bibliographic search is at first carried out to describe the state of the art for this field, bringing also examples of engineering design applications in which this optimization technique is employed and analyzing the most used algorithms, with greater attention to evolutionary algorithms, family that also includes genetic algorithms, which are used in this thesis during for the optimization process. Then the case of study is described, in particular this thesis will examinate the PeWEC’s robust optimization design problem. The device is considered to be located on the Island of Pantelleria. Then we proceed with the application of some chosen frameworks for the WEC robust optimization problem, all of which differ in the chosen objective functions and that are all defined by a specific uncertainties’ probability distribution model. The results of each framework are then compared with those obtained using a non-robust optimization one via post-process analysis. This is done exploiting all the information given by a selected robustness index, employed to study how the system response changes with the variation of some input data, in order to discriminate between influencing and non-influencing factors and understand what their degree of influence is. In this way, an evaluation of the model robustness that allows to compare the different optimization processes is obtained.