基于RRT森林算法的高层消防多无人机室内协同路径规划

Cooperative Indoor Path Planning of Multi-UAVs for High-rise Fire Fighting Based on RRT-forest Algorithm

在多无人机(Multi-unmanned aerial vehicles,Multi-UAVs)协同执行高层消防救援任务的场景中,室内复杂火场环境下路径规划是亟待解决难题之一.针对快速搜索随机树算法(Rapidly-exploring random tree,RRT)搜索区域受限、耗时较长、结果可行性差等问题,提出RRT森林算法.通过随机选取根节点、生成随机树、连接合并随机树,使高层消防多无人机在复杂室内环境下协同路径规划效率显著提高.此外,采用两次动态规划(Dynamic programming,DP)以及改进障碍物接近检测方法,进一步提高路径的可行性.最终,通过仿真验证算法的有效性.

In the scene of multi-unmanned aerial vehicles(Multi-UAVs)cooperating in high-rise fire fighting mission,path planning in indoor complex fire environment is one of the difficult problems to be solved.This paper proposes the RRT-forest algorithm which aims at the problems existing in traditional rapidly-exploring random tree(RRT)algorithm,such as repeated exploration in the same area,large time consumption and poor feasibility of the results.By adding and developing intermediate random trees,connecting and merging random trees,the RRT-forest algorithm can significantly improve the efficiency of cooperative path planning in complex indoor environment with multi-UAVs for high-rise fire fighting.Moreover,by utilizing dynamic programming(DP)for twice and a novel obstacle proximity detection approach,this method can further improve the feasibility of the result.Finally,simulation is conducted to prove the effectiveness of the proposed algorithm.