Efficient Pseudo-Decentralized Control of Wave Energy Arrays

In control of wave energy converter (WEC) arrays, since the dynamics of different devices are coupled by radiation effects, there has always been a dilemma between using centralized control, with optimal energy production but usually overwhelming computation, or decentralized control, with acceptable computation but performance degradation. This paper proposes a pseudo-decentralized control strategy that is effective in energy production but efficient in computation. In this method, a centralized estimation-forecasting scheme is applied to provide global information on the wave excitation force (WEF) experienced by each device. Then, the optimal unconstrained array motion is calculated from the WEF information, to serve as an initial guess of the optimal constrained motion. Finally, the constrained control moves of each WEC are optimized decentralizedly using model predictive control (MPC). However, in contrast to conventional decentralized control, each local controller not only maximizes its own energy production, but also takes into account the energy it contributes, through radiation, to other WECs, given their anticipated motion. In this way, the proposed controller enjoys the same computational simplicity as decentralized control, but is demonstrated, through comprehensive testing, to be capable of achieving performance close to the ideal centralized control across a wide range of sea states.