基于多传感器的空间遥操作机器人虚拟预测环境建模

虚拟预测环境技术是解决遥操作系统中时延问题的有力手段,但其有效性依赖于环境模型的精度。为了提高虚拟预测环境的建模精度,提出了基于多传感器信息融合的建模方法。首先根据CCD摄像机采集的图像建立虚拟环境模型的初始模型,再利用远地力觉和位置传感器信息对初始模型进行校验。为了观察命令在虚拟环境中的执行情况,采用一种简单有效的注册方法,将图形模型叠加在摄像机视频图像上,以便操作者进行对比分析。实验表明利用本文算法建立的虚拟环境模型是准确可靠的,有利于提高基于虚拟预测环境技术的遥操作系统的有效性和可操作性。

Predictive Virtual Environment Modeling of Space Tele-robot Based on Multi-Sensors

Virtual predictive display is a powerful tool to deal with time delay in teleoperation, but it depends on the accuracy of the environment model. This paper proposes a novel approach for environment modeling based on multi-|sensors to improve the precision of the virtual model. First, the model of the slave hand and its environment was initialized with images captured by a CCD camera, then the initialized model was modified by remote force and position information to ensure its accuracy. The virtual model is overlayed with the video images by a simple but effective register algorithm so that the operator can observe the effect of his order. Experimental results show that the model built by this method is precise and it is useful to improve the effectiveness and maneuverability of the teleoperation system based on virtual predictive environment.