Abstract: [Objectives] Unmanned and intelligent efficient inspections is of significant importance for enhancing the reliability of wind turbine operation and reducing maintenance costs. Addressing the severe challenges posed by the strong turbulent environment of wake flows during turbine operation and the complex multi-obstacle environment of mountainous wind farms for unmanned aerial vehicle (UAV) inspection path planning, this paper proposes an intelligent UAV inspection path planning method that considers wind farm wake effects. [Methods] Firstly, a global inspection problem model was constructed, with the optimization objectives of inspection path cost and drone endurance time, and drone flight characteristics as constraint conditions. A virtual inspection scenario reflecting the complex terrain of mountainous wind farms and the characteristics of turbine wake flows is designed. Secondly, a policy interactive target bias rapidly-exploring random trees (PITB-RRT) algorithm is proposed, which significantly improves path efficiency and computational efficiency by optimizing sampling and expansion strategies. Additionally, the non-dominated sorting genetic algorithm Ⅱ (NSGA-Ⅱ) is introduced to optimize the sequence of inspection points. [Results] For a typical mountainous wind farm comprising 12 turbines, the method efficiently plans 66 feasible paths between inspection points. The solved optimal globally inspection path has a length of approximately 9 342 m, corresponding to a flight time of 25.3 minutes, reducing path cost by 28.8% compared to traditional sequences. [Conclusions] The proposed method exhibits excellent applicability and efficiency in complex wind farm scenarios, providing technical support for the intelligentization and automation of UAV inspection tasks.

Key words: wind power, wind turbine, wind farm, intelligent inspection, unmanned aerial vehicle (UAV), path planning, rapidly-exploring random trees