UAV multi-dynamic target points path planning with obstacles based on SOM-DAPF
Chunhui Xiao, Xiaohu Zhao, Yuanyuan Zou, Shaoyuan Li · 2020
In this paper, the unmanned aerial velhicle (UAV) path planning problem with obstacles is disscussed for an air-ground collaborative system which is composed of a messenger UAV and multiple unmanned ground vehicles (UGVs). The messenger UAV needs to visit each UGV to collect and share information. The aim of this problem is to achieve a shortest path which makes the UAV fly over all UGVs while avoiding obstacles. By the Dubins model, the path planning problem is formulated as a Dynamic Dubins Traveling Salesman Problem with Neighborhood (DDTSPN) under obstacle constraints. To adapt to the change of UGVs' locations and reduce the complexity of tuning global path for obstacle avoidance, a moving horizon optimization strategy is porposed. After the UAV visits a UGV, the rest of UGVs' access sequence and access locations are determined using SOM method according to the real-time locations of UGVs. Further, the DAPF algorithm is presented to replan the path which enters the obstacle regions. Simulation results show the advantages of the proposed path planning algorithm.