论文部分内容阅读
设计并实现了一种基于计算机视觉的无人机(UAV,Unmanned Aerial Vehicle)自主空中加油仿真平台,采用点匹配与位姿估计算法对软管式加油锥发光二极管(LED,Light Emitting Diodes)标志进行识别.对大气扰动和尾涡流进行了建模,建立了加油锥模型,并设计了加油锥运动状态预估算法.在此基础上,设计了基于线性二次型调节器(LQR,Linear Quadratic Regulator)的飞行控制律.基于上述技术,在实验室环境下设计开发了基于单目机载摄像机的无人机自主空中加油半物理仿真平台.半物理试验结果表明:所设计的位姿匹配算法和飞行控制律能够满足复杂环境下无人机自主空中加油的性能要求,具有较强的实时性、准确性和鲁棒性.
A self-propelled aerial vehicle simulation platform based on computer vision (UAV) is designed and implemented. The spot matching and pose estimation algorithm is used to identify the characteristics of the hose-type LED (Light Emitting Diodes) The model of the oil coning was established, and the prediction algorithm of the oil coning kinematic state was designed.On the basis of this, a linear quadratic regulator (LQR) Regulator) based on the above-mentioned techniques.A semi-physical simulation platform for autonomous aerial refueling of UAV based on monocular airborne camera was designed and developed in laboratory environment.The results of the semi-physical experiments show that the proposed attitude-position matching algorithm And flight control law can meet the performance requirements of autonomous aerial refueling of UAV in complex environment, and has strong real-time, accuracy and robustness.