Adaptive planning of multi-UAV refined inspection path for complex and irregular building clusters

Amid rapid global urbanization, cities have shifted into a predominantly building maintenance-oriented phase. Therefore, given that existing studies focus on inspecting simple standalone buildings with single UAV, this paper proposes an automatic path planning method for the refined inspection of complex, irregular building clusters. First, an adaptive layering mechanism is introduced to generate full coverage inspection points based on the structural characteristics of the 3D building cluster model. Initial obstacle free flight paths are then derived by integrating A* and greedy algorithms. Further path optimization is conducted by applying the 2-opt algorithm to eliminate intersections and reduce flight distance, while the DP (Douglas Peucke) algorithm is employed simplified the trajectory by reducing redundant waypoints. Experimental validation on six irregularly shaped buildings demonstrates a 9.6% reduction in flight path length and a 47.7% decrease in intermediate waypoints. The proposed framework enables refined inspection path planning for building clusters, improving the automation level and practical applicability of multi-UAVs based building operation and maintenance.