This paper proposed a cooperative control algorithm for UAV (Unmanned Aerial Vehicle) formation control based on visual feedback and the formation kinematics model, enabling the formation to achieve coordinated obstacle avoidance and trajectory generation. Firstly, the UAV speed direction is adjusted through visual feedback information to achieve minimum deflection angle obstacle avoidance. Secondly, the semi-elastic formation kinematic model with linear correlation function of “repulsion velocity – safety radius” is constructed by separating and converging velocities, and the inter-vehicle distance in the formation and the distance between UAVs and obstacles are cooperatively controlled, so as to realise the cooperative obstacle avoidance and trajectory planning of the formation. Through simulation experiments, the results fully demonstrate the stability and robustness of the designed method, and the dynamic obstacle avoidance effect further illustrates the generalization of the algorithm.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Formation Cooperative Obstacle Avoidance Control Based on Visual Feedback Model

  • Junjie Wang,
  • Yongping Hao

摘要

This paper proposed a cooperative control algorithm for UAV (Unmanned Aerial Vehicle) formation control based on visual feedback and the formation kinematics model, enabling the formation to achieve coordinated obstacle avoidance and trajectory generation. Firstly, the UAV speed direction is adjusted through visual feedback information to achieve minimum deflection angle obstacle avoidance. Secondly, the semi-elastic formation kinematic model with linear correlation function of “repulsion velocity – safety radius” is constructed by separating and converging velocities, and the inter-vehicle distance in the formation and the distance between UAVs and obstacles are cooperatively controlled, so as to realise the cooperative obstacle avoidance and trajectory planning of the formation. Through simulation experiments, the results fully demonstrate the stability and robustness of the designed method, and the dynamic obstacle avoidance effect further illustrates the generalization of the algorithm.