Cooperative model predictive control for Wave Energy Converter arrays

Abstract

It is well known that commercial Wave Energy Converters (WECs) are likely to be deployed in arrays, which gives the possibility to enhance the energy harvesting properties of the wave farm as a whole. Advanced control strategies are necessary to exploit the potential and capture as much energy as possible in a limited sea space. Centralized control is a natural benchmark, as it sees the array as a large-scale system, making energy maximization a large plantwide problem. However, centralized control has a high requirement for computational efficiency. In this paper, a cooperative model predictive control (MPC) is proposed to achieve energy maximization while reducing computational cost. A modified array model is constructed to capture the dynamics of interactions in arrays, explicitly showing the linear couplings in the system state, control input, and excitation force of each subsystem. Sensitivity analyses are performed to evaluate the performance for different array layouts, separation distances, and sea states. The results show that cooperative MPC can achieve an energy capture performance that approximates centralized MPC, outperforming decentralized MPC significantly, and is well adapted to sea states. Furthermore, the proposed method has the potential to improve computational efficiency by 14.86%–51.14% for simulations in the study.