Biomimetic Planning Method of UAV Based on Multi-Scale Clustering

Overhead distribution lines are a critical component of the new power system, and their efficient inspection is of great significance for ensuring power supply and supporting socio-economic development. With the rapid advancement of unmanned aerial vehicle (UAV) technology, achieving fully autonomous UAV inspection for distribution networks has become a research hotspot. However, the traditional “human-UAV collaboration” inspection mode is limited by UAV battery life, signal coverage, and flight range, making it difficult to meet the inspection demands of large-scale distribution networks. The development of UAV nests provides a new solution to overcome this bottleneck. This paper addresses the UAV nest deployment problem in distribution networks by proposing a bionic planning method based on multi-scale clustering. The method employs a three-stage optimization process—coarse clustering (Canopy algorithm), fine clustering (K-means++ combined with genetic algorithm), and cluster point filtering (Kd-Tree nearest neighbor search)—to adaptively generate an optimal nest deployment scheme without the need for predefining the number of clusters. It ensures that the coverage range of each nest meets the operational constraints of UAVs. Experiments conducted on a region in Zhejiang Province with over 7000 poles validated the effectiveness of the proposed method. Compared to the traditional K-means++ algorithm, the proposed method demonstrates significant advantages in clustering compactness, coverage satisfaction rate, and computational efficiency, ultimately achieving a 99% pole coverage rate. This study provides a scientific and efficient solution for UAV nest deployment in large-scale distribution networks.