Adaptive systematic optimization of a multi-axis ocean wave energy converter

Due to the growing global energy needs, renewable energy systems, particularly wave energy converters (WECs), are a feasible solution to satisfy current energy demand. Recently, wave farms with diverse technologies, geometries, and layouts have been developed; however, evaluating the performance of these devices is complicated and requires precise hydrodynamic modeling to efficiently deploy wave farms. This study proposes a multi-scenario model using boundary element method (BEM) solver, NEMOH, integrated with evolutionary many-objective algorithms to evaluate the performance of a multi-axis point absorber WEC with respect to cylindrical, triangular, quadrilateral, and octagonal geometries and varying dimensions, that is, radius, draft, and height. To this end, six objective functions were considered to maximize the energy absorption and significant velocity and to minimize the separation distance, levelized cost of energy, net present value, and q-factor. Accordingly, three EMnO frameworks were utilized: the non-dominated sorings genetic algorithm (NSGA-III), reference point-based NSGA-III (R-NSGA-III), and multi-objective evolutionary algorithm by decomposition (MOEA/D). The results of the three optimization algorithms indicate that R-NSGA-III converges faster than the other two and also found that the cylindrical and octagonal geometries produce more annual energy among other forms. Comparing the performances of the three different layouts for cylindrical and octagonal geometries reveals that the arrow layout with thirty buoys produced more energy and had a lower levelized cost of energy and net present value.