With the rapid development of unmanned aerial vehicle (UAV) technology, its applications in fields such as inspection, monitoring, and logistics have increasingly expanded. Meanwhile, UAV can significantly enhance buoy inspection efficiency due to their flexibility and effectiveness. However, optimizing inspection paths to achieve the shortest possible flight distances for all buoys remains an urgent issue. This paper conducts an in-depth study on UAV buoy inspection path planning using the Self-Organizing Map (SOM) algorithm. By converting the geographic coordinates of buoys into planar coordinates and optimizing the path within the SOM framework, the algorithm minimizes the total flight distance between buoys. To verify the effectiveness of the SOM algorithm, comparative experiments were conducted against Genetic Algorithms (GA) and Particle Swarm Optimization (PSO). The results show that the SOM algorithm significantly outperforms in terms of path length, runtime, and computational efficiency. The average path length for SOM was 413.562 km, reducing the distance by 4.52% compared to GA and by 29.01% compared to PSO; the average runtime for SOM was 85.396 s, markedly better than both GA and PSO. These findings provide an effective path planning method for UAV buoy inspections and offer a theoretical and technical foundation for future research.