Abstract
Large-scale array deployment represents a major development direction for Wave Energy Converters (WECs). Nevertheless, optimizing such arrays under real wave conditions and considering maritime restricted area constraints poses challenges, primarily due to the complexities inherent in high-dimensional optimization problems. To overcome this, an adaptive evolutionary algorithm with experience-guided Single-Player Monte Carlo Tree Search (MCTS-GA) is proposed and integrated into a stepwise optimization framework for sequentially optimizing array layouts and Power Take-Off (PTO) damping coefficients. Numerical simulations for arrays comprising 10 to 40 WECs within wave energy farms with different grid cell sizes and under maritime constraints demonstrate that: (i) MCTS-GA achieves best optimization performance, fastest convergence, and strongest stability; (ii) optimal PTO damping coefficients generally increase for WECs located downstream in the direction of wave propagation or near the center of each column; (iii) wave energy farms should preferably be sited in areas without restrictions or with exclusion zones perpendicular to the predominant wave direction; (iv) wave energy farm design requires a careful balance between the number of WECs and the farm's grid cell size. The proposed framework provides an efficient tool for optimizing WEC arrays, elucidating the relationship between farm grid size, maritime area constraints, and energy capture performances.