无人飞行器基于前视序列图像的地形跟随技术

Terrain following based on forward-looking image sequences for unmanned air vehicles

针对飞行器未知环境下自主地形跟随问题,提出了基于前视序列图像天地线的地形跟随方法,并利用适应角法设计了制导律,该方法利用被动成像实时探测飞行前方地形信息,降低了对数字地图保障及高精度导航的需求。六自由度非线性数学仿真结果表明:基于前视序列图像天地线的方法较传统基于数字地图的方法,地形跟随平均离地高度可降低35%以上,可大幅提高飞行器超低空突防性能。

According to the problem of autonomous terrain following in unknown environment for UAVs, a novel terrain following method was introduced based on the skyline of the forward-looking image sequences. And the guidance law of adaptive angle method was designed in detail, where the forward terrain information was detected by the passive images in real time, so the digital map and the high precision navigation requirement were reduced. The 6DOF simulation results demonstrate that the method is better than the traditional method based on digital map, the average terrain following height from the ground can be reduced by over thirty-five percent, so the method can improve the penetration performance significantly.