Abstract:

Improving the capability of autonomous obstacle avoidance for unmanned aerial vehicle (UAV) in complex environment becomes the key problem to be solved urgently. Aiming at the problems that the existing algorithms of obstacle avoidance ignore “potential danger” obstacles′ influence and avoid multi-obstacles complicatedly, the threat classification model of obstacles is built and the velocity obstacle arc method is proposed, which quantifies the influence of threatening obstacles through velocity obstacle arc parameters. Meanwhile, the systematic calculation method of velocity obstacle arc parameters is given, and the differences of relative position and relative velocity between UAV and threatening obstacles are considered in complex environment. The method of collision avoidance based on the velocity obstacle arc method is given. Path planning is taken as an example to conduct simulation for the collision avoidance algorithm, and simulation results demonstrate the practicality and efficiency of the proposed method.