Maynooth University

Maynooth University ePrints and eTheses Archive

Maynooth University Library

Real-time Forecasting and Control for Oscillating Wave Energy Devices

Fusco, Francesco (2012) Real-time Forecasting and Control for Oscillating Wave Energy Devices. PhD thesis, National University of Ireland Maynooth.

[img] Download (10MB)

Abstract

Ocean wave energy represents a signi cant resource of renewable energy and can make an important contribution to the development of a more sustainable solution in support of the contemporary society, which is becoming more and more energy intensive. A perspective is given on the benefi ts that wave energy can introduce, in terms of variability of the power supply, when combined with o shore wind. Despite its potential, however, the technology for the generation of electricity from ocean waves is not mature yet. In order to raise the economic performance of Wave energy converters (WECs), still far from being competitive, a large scope exists for the improvement of their capacity factor through more intelligent control systems. Most control solutions proposed in the literature, for the enhancement of the power absorption of WECs, are not implemented in practise because they require future knowledge of the wave elevation or wave excitation force. The non-causality of the unconstrained optimal conditions, termed complex-conjugate control, for the maximum wave energy absorption of WECs consisting of oscillating systems, is analysed. A link between fundamental properties of the radiation of the floating body and the prediction horizon required for an eff ective implementation of complex-conjugate control is identi fied. An extensive investigation of the problem of wave elevation and wave excitation force forecasting is then presented. The prediction is treated as a purely stochastic problem, where future values of the wave elevation or wave excitation force are estimated from past measurements at the device location only. The correlation of ocean waves, in fact, allows the achievement of accurate predictions for 1 or 2 wave periods into the future, with linear Autoregressive (AR) models. A relationship between predictability of the excitation force and excitation properties of the floating body is also identified. Finally, a controller for an oscillating wave energy device is developed. Based on the assumption that the excitation force is a narrow-banded harmonic process, the controller is e ffectively tuned through a single parameter of immediate physical meaning, for performance and motion constraint handling. The non-causality is removed by the parametrisation, the only input of the controller being an on-line estimate of the frequency and amplitude of the excitation force. Simulations in (synthetic and real) irregular waves demonstrate that the solution allows the achievement of levels of power capture that are very close to non-causal complex-conjugate control, in the unconstrained case, and Model predictive control (MPC), in the constrained case. In addition, the hierarchical structure of the proposed controller allows the treatment of the issue of robustness to model uncertainties in quite a straightforward and eff ective way.

Item Type: Thesis (PhD)
Keywords: Real-time Forecasting; Oscillating Wave Energy Devices;
Subjects: Science & Engineering > Electronic Engineering
Item ID: 4011
Depositing User: IR eTheses
Date Deposited: 27 Nov 2012 13:57
URI:

    Repository Staff Only(login required)

    View Item Item control page

    Document Downloads

    More statistics for this item...