无人直升机三维避障方法及仿真

3D obstacle avoidance method and simulation for unmanned helicopter

自主避障是无人直升机(UH)在复杂未知环境中执行飞行任务所必须的能力。针对未知环境下UH避障飞行问题,提出了一种新的三维实时避障方法,通过将三维激光雷达探测视野划分成若干不同半径的扇形柱区和等角度分布的直方柱区,结合雷达探测数据判断障碍物分布情况,UH根据避障方法执行相应的规避动作来实现避障飞行。同时提出一种基于CATIA三维建模软件二次开发的三维仿真方法,该方法可以对未知的飞行环境进行三维建模,结合避障方法实现UH在其中的真实避障飞行模拟,并利用该仿真方法验证了三维避障方法的可行性。

Autonomous obstacle avoidance is an indispensable ability for unmanned helicopter （UH） operating low-altitude flight. A new 3D real-time avoidance method was presented to solve UH obstacle avoidance problem in complicated environment. The vision space of sensor was divided into several unequal radii sector-shaped cylinders and cuboids distributed at the same angle, estimating the obstacle distribution based on sensor data, and UH can execute corresponding maneuver calculated by this method to avoid obstacles successfully. An originally 3D simulation method was also proposed using secondary development technology on CATIA, which can build 3D model for undiscovered environment. Combined with the obstacle avoidance method, real UH obstacle avoidance flight simulation was carried out in it. The feasibility of 3D obstacle avoidance method was validated using the proposed simulation method.