An investigation into the applicability of the Fourier transform to dispersive water waves and their short term prediction
After many years of slow but progressive development, the wave energy industry is on the cusp of breaking through the economic and technical barriers to full scale deployment of wave energy electrical generating devices. As the major obstacles in device design are solved, and with several devices in the water, the scope for increasing their efficiency through advanced control techniques is now becoming clearer. In some cases, it would be advantageous to integrate an advanced prediction of wave behaviour (of some tens of seconds into the future) into these control methods. Past research on wave prediction has focused on utilising the Fourier theorem to deconstruct wave records and then make predictions ahead in space, with published results indicating promise. However, predicting ahead in time has so far not been achieved. This thesis takes the Fourier theorem method of prediction to its logical conclusion by exploring its limitations in predicting over both time and space. A discussion as to why these limits should exist, and possible work into the solution of the wave prediction problem, are also presented. A review of current devices under development, and the history and emergence of the wave generating industry (which is a comparatively recent technology and still in its infancy), are also included as appendices to the main thesis in order to put the work into context.