Abstract
Wave energy is nowadays one of the most promising renewable energy sources; however, wave energy technology has not reached the fully-commercial stage, yet. One key aspect to achieve this goal is to identify an effective control strategy for each selected Wave Energy Converter (WEC), in order to extract the maximum energy from the waves, while respecting the physical constraints of the device. Model Predictive Control (MPC) can inherently satisfy these requirements. Generally, MPC is formulated as a quadratic programming problem with linear constraints (e.g., on position, speed and Power Take-Off (PTO) force). Since, in the most general case, this control technique requires bidirectional power flow between the PTO system and the grid, it has similar characteristics as reactive control. This means that, under some operating conditions, the energy losses may be equivalent, or even larger, than the energy yielded. As many WECs are designed to only allow unidirectional power flow, it is necessary to set nonlinear constraints. This makes the optimization problem significantly more expensive in terms of computational time. This work proposes two MPC control strategies applied to a two-body point absorber that address this issue from two different perspectives: (a) adapting the MPC formulation to passive loading strategy; and (b) adapting linear constraints in the MPC in order to only allow an unidirectional power flow. The results show that the two alternative proposals have similar performance in terms of computational time compared to the regular MPC and obtain considerably more power than the linear passive control, thus proving to be a good option for unidirectional PTO systems.
Highlights
Ocean energy and wave energy is nowadays one of the most promising, but still underutilized, renewable energy sources that can be commercially exploited to provide power to the electrical grid [1]
The results show that the two alternative proposals have similar performance in terms of computational time compared to the regular Model Predictive Control (MPC) and obtain considerably more power than the linear passive control, proving to be a good option for unidirectional PTO systems
The PTO damping in passive MPC Proposal 1 of Section 4 (P-MPC-1) presents less abrupt changes and lower values compared to passive MPC Proposal 2 of Section 5 (P-MPC-2). The reason for this is that P-MPC-1 is based on keeping the PTO damping constant over the prediction horizon, while P-MPC-2 estimates a different PTO damping for each time instant over the prediction horizon
Summary
Ocean energy and wave energy is nowadays one of the most promising, but still underutilized, renewable energy sources that can be commercially exploited to provide power to the electrical grid [1]. Energy Converters, because its formulation is optimal in terms of energy extraction and is able to inherently manage physical constraints [6,7] This technique requires a WEC model and an excitation force prediction within a predefined time interval in order to predict the control action on the PTO force [8]. This work proposes two independent MPC control strategies applied to a two-body self-reference point absorber that address this issue from two different perspectives: (a) adapting the MPC formulation to achieve passive loading strategy similar to the work by Tom and Yeung [14], but integrating analytically the dynamic of the WEC inside the objective function; and (b) re-formulating the linear constraints in the MPC in order to only allow an unidirectional energy flow.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
