中文摘要：

为了提高卫星自维护平台操作的自主性、安全性和鲁棒性，提出了一种融合阻抗控制的视觉自主操作方法．针对太空中变化强烈的光照环境，对物体识别采用合作目标亚像素精度定位的方法，提高了合作目标三维空间定位的精度．对合作目标的识别算法进行了特别的设计，以适用于卫星有限的计算能力，并将视觉伺服算法融入基于关节力矩的阻抗控制．当卫星自维护平台执行接触操作时，视觉伺服能提供与被操作物体的精确接触信息，提高了卫星自维护平台对工作环境的适应能力和顺从能力．实验结果证明了该方法的可行性．

英文摘要：

To improve the autonomy, safety and robustness of satellite self-servicing platform, an autonomous visual operation strategy implementing impedance control is proposed. The cooperative object method is adopted to recognize objects to adapt the extreme changed light conditions in space. To improve the locating accuracy of the cooperative object in 3D cartesian space, a locating method with sub-pixel accuracy is implemented. The cooperative object recognition algorithm is optimized depending on the limited computing capability in the satellite. The impedance control based on joint torque is implemented into the visual servoing algorithm. While executing contract task, the visual servoing can provide the tactile information with high accuracy, in the sequel, improve the satellite self-servicing platform＇s adaptability and compliance capability to environment. Finally, experimental results are provided to verify the effeciency of the proposed method.