无人平台视觉导航算法验证仿真系统的设计与实现

Design and Implementation of Unmanned Platform Simulation System for Visual Navigation Algorithm Verification

近年来,基于视觉的导航和目标跟踪算法开始广泛应用于无人平台。针对各类视觉导航算法和综合任务方案在无人平台上实地测试风险性高、现场采集参数精度低、效率差等问题,设计并搭建了一套以高精度工控导轨平台、姿态控制云台、工业相机和模拟地形沙盘为主要部件的无人平台半物理仿真系统。引入OptiTrack光学运动捕捉设备对仿真系统真实的位姿控制精度进行了评价,结果显示系统空间三轴位置控制均方根误差在1 mm左右,横滚、俯仰、航向姿态角控制均方根误差小于0.1°,可以服务于视觉导航算法和综合任务方案在无人平台上的性能评估。

In recent years,visual-based navigation and object tracking algorithms have been widely used in unmanned platforms.To solve the problems of the high risk of field testing of various visual navigation algorithms and comprehensive task schemes on unmanned platforms,the low accuracy of field collection parameters and poor efficiency,an unmanned platform semi-physical simulation system is designed and built with high-precision industrial control guide platform,attitude control head,industrial camera and simulated terrain sand table as main components.The experiments are designed to evaluate the real pose control accuracy of the simulation system using OptiTrack optical motion capture equipment.The results show that the system space triaxial position control root mean square error is around 1 mm,attitude angle control root mean square error is less than 0.1°.So the system can be used to evaluate the performance of all kinds of visual navigation algorithms and integrated task schemes on the unmanned platform.