基于立体场景的人控交会对接仿真研究

Study on Manual Control Rendezvous and Docking Simulation Based on Stereo Scene

传统的人控交会对接主要通过观察摄像机产生的十字靶标图像来完成,容易造成人的视觉误差,从而导致手动对接失败。结合图形学理论和虚拟现实技术构建的人控交会对接仿真系统为航天员实时提供逼真的虚拟立体场景,通过各种任务想定对航天员进行飞行任务训练。在实验过程中,航天员通过立体眼镜观察虚拟场景中飞行器的飞行状态,同时操作控制手柄改变飞行器的位置和姿态进行手动对接训练。通过不同的对接实验参数设置,航天员可以反复进行不同条件下的对接训练,而不需要执行空间飞行任务。实验结果表明基于立体场景的人控交会对接训练有较高的成功率。

Astronauts perform manipulative RVD （rendezvous and docking） usually by means of cross drone image from vidicon, which can easily bring sight error and result in the failure of RVD. Using graphic theory and virtual reality technology, the simulation system of manipulative RVD provides a realistic stereo scene in real time for astronauts, who can be trained to fulfill many preformed task scripts in the system. In the process of experiment, astronauts wearing stereo glasses observe the flight status of spacecraft in virtual stereo scene, and at the same time, they can manipulate the stick to control the flight status of the spacecraft and carry on the training of manipulative RVD for real space flight. By setting different experimental parameter, astronauts can be trained in different conditions repeatedly without performing space flight task. Results of simulation experiment show that the success rate of manipulative RVD based on stereo scene is larger.