论文部分内容阅读
针对无人飞行器视觉定位结果存在较大时延而影响飞行器运动状态估计精度的问题,提出了一种基于多传感器数据融合的实时运动估计方法.首先,利用机载惯性测量元件(IMU)提供的姿态信息优化单目视觉定位算法,使得视觉定位结果的时延减小.然后,在利用卡尔曼滤波器估计飞行器运动状态的过程中,考虑了视觉定位结果的时延,利用加速度信息进行时延补偿.最终得到实时的高精度运动估计结果.在自主研制的四旋翼飞行器系统上对本文提出的方法进行了验证.通过与不考虑时延的方法的结果以及真实数据进行比较,证明了本方法的有效性.
Aiming at the problem that there is a big delay in the visual positioning results of UAV and affects the accuracy of the aircraft’s motion state estimation, a real-time motion estimation method based on multi-sensor data fusion is proposed.Firstly, using the IMU Attitude information to optimize the monocular vision positioning algorithm, the time delay of the visual positioning results is reduced.Furthermore, in the process of estimating the aircraft’s moving state by using Kalman filter, the time delay of the visual positioning result is taken into consideration, and the delay is made by using the acceleration information Compensation.Finally, the results of real-time high-precision motion estimation are obtained.The method proposed in this paper is validated on a self-developed quadrotor system.The results of the proposed method and the real data show that this method Effectiveness.