Abstract: Drone inspection features low cost and high efficiency and is gradually replacing traditional manual methods, becoming the primary means of intelligent inspection for distribution lines. However, the complex structure of distribution networks, with numerous branch lines, leads to slow convergence of single flight path planning and a tendency to fall into local optima. To address these issues, this paper proposes a path optimization algorithm for unmanned aerial vehicle (UAV) inspection. The method uses a binary tree model to split and recombine the distribution network topology and incorporates UAV endurance range for analysis and calculation. Taking the key line sections as the boundary of an outer circle, the center of the circle is calculated inward based on the coverage range and is set as the takeoff and landing point. The number of towers that can be covered within the UAV’s flight radius is calculated and removed from the tower set. This process is repeated until all towers along the line are covered and inspected, thereby improving the efficiency of UAV inspection.