Proposal of novel analytical wake model and GPU-accelerated array optimization method for oscillating wave surge energy converter

Abstract

Oscillating wave surge converter (OWSC) is widely utilized to exploit energy from waves. Due to the wake effects, the layout of OWSCs will significantly affect the energy output of an OWSC farm. However, up to now the layout of OWSC farm was only determined empirically, and there was no quantitative layout optimization method for OWSC farm. The quantitative layout optimization needs analytical wake model for OWSC to conduct fast optimization iterations. However, there was still no analytical wake model for OWSC. Therefore, in this study, we proposed an analytical wake model regarding regular waves, based on which a quantitative method for optimizing the layout of OWSC farm was also proposed. It was found that the total energy outputs of the optimized layouts using the proposed method are significantly larger than those of the empirical ones, and increasing the spacing between the OWSCs in the bi-direction of wave propagation can greatly increase the total energy output. The proposed OWSC analytical wake model can predict the wave height changes with an accuracy of 94.73%. Innovatively, GPU-acceleration technology was utilized, and the GPU-based codes can run at least 1479 times faster than the CPU-based ones. Those implemented codes are opened for other researchers.