In this paper, we propose a novel curved path following scheme for multiple fixed-wing unmanned aerial vehicles (UAVs) that can achieve coordinated curved path following and handle collision avoidance simultaneously. The proposed solution is a hybrid system that combines a path following strategy and a collision avoidance method to ensure collision-free maneuvering of a group of UAVs cooperative flight. A strategy based on a virtual structure and a kinematic model is proposed to derive the cooperative curved path following. Meanwhile, a possible collision will occur when multi-UAVs flight in the form of dense formations, or there are path intersections when the formation is changed. However, all fixed-wing vehicles need to maintain a minimum airspeed and cannot stop before the collision. Therefore, there is a need for a fast and efficient method that can be implemented online, taking into account the physical limitations of the vehicle (minimum speed, turning rate, etc.) to avoid possible collision. We employ a modified vector field histogram (VFH) method to provide real-time collision avoidance, and evasive maneuvers work only if the distance is within the conflict zone. In order to verify the proposed control scheme, dense formation flight tests of small fixed-wing UAVs were implemented.

在本文中，我们提出了一种适用于多架固定翼无人机的新型弯道跟踪方案，该方案可以实现协调的弯道跟踪并同时避免碰撞。所提出的解决方案是一种混合系统，该系统结合了路径跟踪策略和防撞方法，可确保无人无人机协同飞行的机动。提出了一种基于虚拟结构和运动学模型的策略来推导协同弯曲路径跟随。同时，当多UAV以密集地层的形式飞行时，或在改变地层时存在路径交叉点时，可能会发生碰撞。但是，所有固定翼车辆都需要保持最低空速，并且在碰撞前不能停车。因此，考虑到车辆的物理限制（最小速度，转弯速率等），以避免可能发生的碰撞，需要一种可在线实施的快速有效的方法。我们采用了改进的矢量场直方图（VFH）方法来提供实时的避碰功能，只有在距离在冲突区域内时，躲避回避行为才起作用。为了验证所提出的控制方案，对小型固定翼无人机进行了密集编队飞行试验。