A preliminary study into optimising the shape of a wave energy collector using a genetic algorithm

This paper describes an initial study into the optimisation of the shape of a wave energy collector using a genetic algorithm. The study investigates three descriptions of the surface geometry of a surging device, which are both compatible with panel-based hydrodynamic analysis software and form a suitable representation in the genetic algorithm. The analysis has been simplified by considering only the two-dimensional problem, that of optimising the cross-sectional shape of the surging body. Real-valued representation and continuous search space recombination and mutation are used in the genetic algorithm. A basic power absorption cost function is used to assess the relative performance of each candidate shape. The contribution of the results to the overall understanding of the optimization procedure and the direction of future research is discussed. In addition, the relative merits of the different geometric descriptions, and their value for further development are considered.