基于光流的四旋翼飞行器控制技术研究

Research on Control Technology of Four-rotor Aircraft Based on Optical Flow

为了解决无人机在无GPS或GPS信号较弱情况下的稳定悬停问题,并考虑到成本与板载资源限制,将通过小型激光雷达获得的距离信息分别与通过单摄像头获得的视觉信息,以及通过IMU(惯性测量单元)获得的惯性信息利用互补滤波算法进行融合,以实现对四旋翼飞行器姿态与水平速度的运动估计。采用基于串级PID的多闭环控制策略,实现对四旋翼飞行器水平与垂直方向的控制。实验结果表明,所设计的基于光流和小型激光雷达的四旋翼飞行器控制策略与传统利用光流和超声波测距传感器方案相比,控制精度提高了10%左右,能够以最大±2°的姿态角误差,以及最大2.3cm/s的水平速度误差实现定点悬停功能。

In order to solve the problem of stable hovering of drones without GPS or GPS signals, and taking into account cost and onboard resource constraints, the distance information obtained by a small lidar is fused with the visual information obtained by a single camera and the inertial information obtained by the IMU (Inertial Measurement Unit) by a complementary filtering algorithm to realize the motion estimation of the attitude and horizontal velocity of the quadrotor. The horizontal and vertical control of the quadrotor is realized by a multi-closed loop control strategy based on cascade PID. The experimental results show that the control precision of the four-rotor aircraft based on optical flow and small lidar designed in this paper is about 10% higher than that of the conventional optical flow and ultrasonic ranging sensor, and can achieve a maximum ±2° attitude angular error and a horizontal speed error of up to 2.3 cm/s to achieve the function of fixed-point hover.