基于卡尔曼滤波的四轴飞行器成像、惯性和高度组合导航

Imaging,Inertial and Altitude Integrated Navigation for Quadrotor Based on Calman Filter

为了提高四轴飞行器在非结构化、室内和GPS失效环境下的导航精度,本文提出一种用于四轴飞行器的成像、惯性和高度组合导航。首先对四轴飞行器搭载的摄像机进行了标定,获取了摄像机的内外参数。其次设计了着陆标志为标准参照物,以及着陆标志识别方案。再利用卡尔曼滤波器融合视觉、惯性测量系统和超声波测距仪的数据,通过相似三角形定理估计飞行器的相对高度,利用视觉测程法估计飞行器位置和位移速度。最后验证组合导航的有效性,用成像、惯性与高度组合导航实验平台依次进行了相对高度估计实验、平移速度和相对位置估计实验,实验测试表明:在高度估计实验中,组合导航能够很好的描述四轴飞行器的飞行高度估计;在平移速度估计实验中,卡尔曼滤波器估计的速度估计值很平滑;相对位置估计实验中,四轴飞行器的x方向位置保持在原点附近,y方向位置保持在-1.0 m附近,高度位置的最大误差在0.3 m之内。

In order to improve the navigation accuracy of quadrotor in unstructured,indoor and GPS failure environments,this paper proposes an imaging,inertial and altitude integrated navigation for quadrotor.Firstly,the camera mounted on the quadrotor is calibrated,and the internal and external parameters of the camera are obtained.Secondly,we use the landing signs as standard reference objects and present a landing signs recognition scheme.Then,the Calman filter is used to fuse the data of the vision,inertial measurement system and ultrasonic range finder.And the relative height of the aircraft is estimated by the similar triangle theorem.The visual odometry is used to estimate the position and translation speed of the quadrotor.Finally,the effectiveness of the integrated navigation is verified.The experiment of the relative height estimation experiment,the translation speed and the relative position estimation are carried out successively with the imaging,inertial and altitude integrated navigation experimental platform.The experimental results show that combination navigation can accurately estimate the altitude of quadrotor in the relative height estimation experiment;In the translation velocity estimation experiment,the curve of velocity estimation by the Calman filter is smooth;In the relative position estimation experiment,the position of the X direction of the quadrotor is kept near the origin,the Y direction is kept near-1.0 m,and the maximum error of the altitude is within 0.3 m.