The levellised cost of energy of wave energy converters (WECs) is not competitive with fossil fuel-powered stations yet. To improve the feasibility of wave energy, it is necessary to develop effective control strategies that maximise energy absorption in mild sea states, whilst limiting motions in high waves. Due to their model-based nature, state-of-the-art control schemes struggle to deal with model uncertainties, adapt to changes in the system dynamics with time, and provide real-time centralised control for large arrays of WECs. Here, an alternative solution is introduced to address these challenges, applying deep reinforcement learning (DRL) to the control of WECs for the first time. A DRL agent is initialised from data collected in multiple sea states under linear model predictive control in a linear simulation environment. The agent outperforms model predictive control for high wave heights and periods, but suffers close to the resonant period of the WEC. The computational cost at deployment time of DRL is also much lower by diverting the computational effort from deployment time to training. This provides confidence in the application of DRL to large arrays of WECs, enabling economies of scale. Additionally, model-free reinforcement learning can autonomously adapt to changes in the system dynamics, enabling fault-tolerant control.

To achieve persistent systems in the future, autonomous underwater vehicles (AUVs) will need to autonomously dock onto a charging station. Here, reinforcement learning strategies were applied for the first time to control the docking of an AUV onto a fixed platform in a simulation environment. Two reinforcement learning schemes were investigated: one with continuous state and action spaces, deep deterministic policy gradient (DDPG), and one with continuous state but discrete action spaces, deep Q network (DQN). For DQN, the discrete actions were selected as step changes in the control input signals. The performance of the reinforcement learning strategies was compared with classical and optimal control techniques. The control actions selected by DDPG suffer from chattering effects due to a hyperbolic tangent layer in the actor. Conversely, DQN presents the best compromise between short docking time and low control effort, whilst meeting the docking requirements. Whereas the reinforcement learning algorithms present a very high computational cost at training time, they are five orders of magnitude faster than optimal control at deployment time, thus enabling an on-line implementation. Therefore, reinforcement learning achieves a performance similar to optimal control at a much lower computational cost at deployment, whilst also presenting a more general framework.

The wave excited motion of a body floating on water confined between two semi-infinite ice sheets is investigated. The ice sheet is treated as an elastic thin plate and water is treated as an ideal and incompressible fluid. The linearized velocity potential theory is adopted in the frequency domain and problems are solved by the method of matched eigenfunctions expansion. The fluid domain is divided into sub-regions and in each sub-region the velocity potential is expanded into a series of eigenfunctions satisfying the governing equation and the boundary conditions on horizontal planes including the free surface and ice sheets. Matching is conducted at the interfaces of two neighbouring regions to ensure the continuity of the pressure and velocity, and the unknown coefficients in the expressions are obtained as a result. The behaviour of the added mass and damping coefficients of the floating body with the effect of the ice sheets and the excitation force are analysed. They are found to vary oscillatorily with the wave number, which is different from that for a floating body in the open sea. The motion of the body confined between ice sheets is investigated, in particular its resonant behaviour with extremely large motion found to be possible under certain conditions. Standing waves within the polynya are also observed.

Collisions between two thin floating disks forced by regular water waves are studied for a range of wave amplitudes and lengths, using laboratory wave basin experiments and a mathematical model. Three collision regimes are identified from the experiments in terms of collision frequency and strength, and the collisions are shown to be caused by drift for short incident wavelengths and relative surge motion between the disks for longer incident waves. The model is based on slope-sliding theory for the wave-induced disk motions and rigid-body collisions. It is shown to predict collision frequencies and velocities accurately for intermediate–long incident wavelengths. Incorporating drift and wave scattering forces into the model is shown to capture the collision behaviours for short incident wavelengths.