Path Planning of Improved RRT* Based on DBSCAN Algorithm

Aiming at the path planning problem of the unmanned aerial vehicle (UAV) under complex and changeable flight environment and flexible missions, the shortcomings of the existing Rapidly-exploring Random Tree* (RRT*) algorithm are improved. The improved method avoids the excessive ineffective search of the RRT* algorithm in space, reduces the computation, improves the planning speed, the search results are more efficient and fit the physical constraints of the vehicle, and the planned path result is better. At the same time, the DBSCAN clustering algorithm is combined with the specific application scenarios for improvement, thus the critical points in the planning results are effectively extracted. The new planning results are smoother and more reliable, and the randomness of the improved RRT* algorithm is avoided. The simulation results show that the new algorithm is universally applicable, has high security, and the planning results obtained are closer to the optimal state that can be directly applied to the path planning scenarios without prior knowledge.