Modeling, estimation and control of wave energy converters

Abstract

Wave energy converters (WECs) are devices designed to extract energy from ocean waves and have gained increasing attention as a promising renewable energy technology. In the study of WEC systems, three essential topics are wave estimation, WEC dynamic modeling, and WEC control. Wave estimation provides critical real-time information of the wave environment, enabling adaptive operation of the WEC in response to changing conditions. Accurate WEC dynamic modeling is crucial for system identification and serves as the foundation of reliable numerical simulations. Finally, WEC control plays a key role in maximizing the power output by implementing the force or torque via a power take-off system. These three topics are tightly coupled and together form the basis for the effective design and operation of modern WEC systems. Ocean waves are inherently complex and often exhibit highly nonlinear behavior. Nevertheless, they can be decomposed into a superposition of regular linear wave components (Airy waves) with different wave frequencies. This spectral decomposition enables real-time reconstruction of the wave field using a state-space representation combined with a Kalman filter observer. Both numerical simulations and experimental tests have demonstrated that this approach can accurately reconstruct the wave field with relatively low error. Conventional WEC dynamic simulation typically works in the frequency domain, which relies on three linear assumptions: small wave amplitude assumption, small WEC amplitude assumption, and open ocean assumption. Due to these assumptions, this approach becomes limited in some complex wave scenarios. To address these limitations, this dissertation adopts a time domain nonlinear panel method and implements it under periodic boundary conditions, making it suitable for simulating large amplitude (nonlinear) WEC motion with arbitrary initial conditions of the wave. Numerical simulations demonstrate that the nonlinear panel method yields results consistent with the traditional Boundary Element Method (BEM) under linear assumptions, while also capturing the nonlinear characteristics of WEC dynamics in more realistic wave environments. Finally, WEC control strategies are investigated under two distinct scenarios. In the first scenario, both the wave and the WEC motion are assumed to be small, while the power take-off (PTO) system exhibits quadratic energy losses. To optimize the net energy output under these conditions, a linear quadratic (LQ) controller is designed. This controller is implemented on the Floating Oscillating Surge Wave Energy Converter (FOSWEC) system. The results indicate that it could extract 99% of the total available energy. In the second scenario, the WEC undergoes large amplitude motion near the wave surface, where its submersion depth varies over time. In this case, a saturated control strategy is proposed and implemented using a sliding mode control framework. The performance of this controller is validated using both the nonlinear panel method and direct numerical simulation, confirming its ability to enhance energy extraction by more than 25% compared to the PI controller. In summary, this dissertation introduces novel structures for the estimation, modeling, and control of wave energy converters, making significant contributions to the advancement of WEC research and design, and promoting the adoption of wave energy as a cost-effective and reliable energy source.