机载视觉和IMU融合的无人机自主着舰研究  

Autonomous landing of UAV with airborne vision and IMU fusion

作　　者：方斌 徐硕[1] 钟志伟 FANG Bin;XU Shuo;ZHONG Zhiwei(Department of Traffic Administration and Engineering,Hunan Police Academy,Changsha 410138,China;School of Systems Engineering,NUDT,Changsha 410073,China)

机构地区：[1]湖南警察学院交通管理系,湖南长沙410138 [2]国防科技大学系统工程学院,湖南长沙410073

出　　处：《飞行力学》2020年第3期41-45,共5页Flight Dynamics

基　　金：湖南省教育厅优秀青年资助项目(17B087);湖南警察学院博士专项基金资助(2016ZX03)。

摘　　要：针对无人机着舰过程中动-动平台视觉相对定位、定姿难的问题,搭建了一套可模拟不同海况下的无人机着舰物理仿真系统,并进行基于机载视觉和惯性测量仪(IMU)融合的无人机自主着舰研究。使用六自由度转台模拟不同级别下的海况,采用机载视觉检测和跟踪降落平台上的合作标志,计算出无人机相对于标志板的位姿,控制无人机在标志板上空悬停、自主降落到降落平台上。试验结果表明,该方法能为无人机在高海况条件下的自主着舰提供高精度的相对定位数据,并引导无人机实现安全地自主着舰。In this paper,a simulation about UAV landing system under different sea conditions is built to solve a dynamic moving platform of visual positioning and attitude determination problem,and UAV’s autonomous landing research based on airborne vision and inertial measurement unit(IMU)fusion is carried out.Six degree of freedom turntable is used to simulate different levels of sea conditions,while the airborne camera gets the visual information to detect and track landing marking plate on the motion platform,calculate the UAV pose relative to the marker board,control the UAV hovering over the marker board and autonomously landing on the platform.Experimental results show that this method can provide the UAV with high precise autonomous landing data under high sea condition,and guide UAVs to achieve safe autonomous landing.

关 键 词：自主着舰 机载视觉 神经网络 数据融合 

分 类 号：V279[航空宇航科学与技术—飞行器设计] V249