| 摘要: |
| 无人机自主着陆过程中需要实时获得高精度的导航信息，对自主性、实时性的要求较高。现有的导航方式都存在各自的不足，且在室内等新型环境中不能使用。针对这一问题，提出了一种视觉/惯性组合导航算法。首先建立了世界坐标系下惯性导航的数学模型，随后通过Kalman滤波实现位置、姿态匹配，其中位置匹配完成速度误差、加表零偏的估计；姿态匹配完成安装误差角、陀螺漂移的估计，并利用估计得到的安装误差角和视觉导航系统输出的姿态信息对惯导姿态进行修正。仿真结果表明，该算法具有一定的工程应用价值。 |
| 关键词: 无人机 自主着陆 视觉/惯性组合导航 位置、姿态匹配 |
| DOI： |
|
| 基金项目: |
|
| Navigation Algorithm Based on Inertial/Vision Information Fusion of UAV Autonomous Landing |
| LIU Chang,YANG Liang-jun,LIU Feng,WANG Yi,XU Ce,LIU Chong-liang |
| (Beijing Institute of Automatic Control Equipment;Military Representative Office of Navy in Kunming Area) |
| Abstract: |
| High-precision navigation information which needs to be real-time and autonomic is essential during UAV autonomous landing. The existing navigation modes have their respective weaknesses and can’t be used indoors. An INS\Vision integrated navigation algorithm is raised to solve this problem.Mathematical model of inertial navigation in the world coordinate system is built firstly, then position matching and attitude matching by Kalman filtering are presented. Velocity error and accelerometer zero offset are estimated by position matching, installation error angle and gyroscope drift are estimated by attitude matching. The attitude information provided by INS can be corrected by the installation error angle and the attitude information provided by vision navigation system. The simulation results show that this algorithm is feasible |
| Key words: UAV Autonomous landing INS/Vision integrated navigation Position/Attitude matching |
