Space station robot: design, mobility and manipulation.

空間站探索是世界熱點研究問題，空間站巨大因此往往在外太空在軌建造和組裝，因此需要由宇航員和機器人開展大量的艙外工作（ Extra-Vehicular Activities，EVAs）。目前，宇航員的艙外工作存在以下三個方面的問題：（1）宇航員在移動攀爬和任務操作相互衝突的問題， 宇航員在艙外工作時不能一邊沿著扶手攀爬，一邊搬運載荷；（2）當前的空間機器人工作空間有限，在空間站表面有許多地方無法達到開展艙體檢查等在軌服務；（3）當前的空間機器人在狹窄空間裡操作效率不高，不能很好地避開障礙物和做精細化靈巧操作。本論文針對以上問題提出了2 個機器人系統，四腿行走機器人（ Four-legged Robot Walker, FLRW ） 和圓弧軌道機器人（Circular Rail Robot System , CRRS）。四腿行走機器人擅長多退的扶手攀爬和操作，攀爬和操作可同時開展。四腿行走機器人還有一個可旋轉頭部（可旋轉的視覺系統）來增強攀爬過程中的視覺範圍。圓弧軌道機器人通過移動機器人在圓弧軌道的運動能完全覆蓋空間站的所有工作空間，該系統是在太空應用的第一個弧形軌道系統，同時也有最小的轉彎半徑。本論文對提出的2 個機器人系統的移動性進行了深入的研究，四腿機器人側重在雙臂攀爬的策略、攀爬步態，並開展了全艙攀爬的在軌任務模擬驗證。圓弧軌道機器人開展了艙體軌道系統、多艙體軌道切換器、移動基座平臺、移動平臺驅動和轉彎半徑的深入設計與分析，並且完成了移動軌道平臺的原型樣機試製驗證。本論文對提出的2 個機器人系統的操作臂開展了非球形腕部掛接、冗餘操作臂奇異點辨識研究，提出了一種雅克比初等變換（MJET）演算法進行操作臂奇異分析，該演算法可以將冗餘機械臂的6x7 奇異矩陣轉化到3x4 的子矩陣，大大提高了運算效率。論文還開展了在多移動物體環境下的避障研究，提出了一種即時的多移動物體障礙回避（MMOA）演算法，該演算法採用超曲面函數描述障礙物的包絡，採用偽距離即時計算與移動障礙物距離，取得了控制精度和即時性的平衡。本論文對提出的2 個機器人系統的操作臂開展了動力學建模和在軌裝配研究，採用拉格朗日建模方法對操作臂建模，並與商務軟體ADAMS 對比驗證建模準確度。同時，並運用阻抗控制演算法針對ORU 的在軌抓取、安裝和轉移等在軌任務的實現驗證。最後論文進行了總結和後續工作展望。Space station exploration is a global hot research topic. The space stations are usually large in scale so that they have to be fabricated and assembled in space, which involves a large number of Extra-Vehicular Activities (EVAs) by astronauts and robots. There are three main problems of EVA mission. (1) Astronauts experience a conflict between climbing and manipulation during EVA missions, as they cannot carry payloads while handrail climbing. (2) Current space robots have workspace limitations and cannot reach the whole exterior of a space station, making it challenging to carry out inspection and servicing. (3) It is also difficult for robots to avoid obstacles and perform fine manipulation tasks in a compact workspace.Two robotics systems, the Four-Legged Robot Walker (FLRW) and the Circular-Rail Robot System (CRRS), are proposed to address the above problems. The FLRW is good at handrail climbing as it has multiple, identical legs. It also has a rotatable vision system to enhance its field of view during climbing. The CRRS provides full coverage of the space station workspace, as it is a mobile robot that drives on a circular rail system around the space station. This system is the first design of robotic system with circular-rail in space and also has the smallest turning radius.The mobility of both robots is addressed. The FLRW analysis focuses on the climbing strategy and climbing gait analysis. The circular rail system, rail switch, mobile platform, driving force and turning radius of CRRS are carefully designed and analyzed. A prototype of the CRRS mobile platform is implemented for verification.The proposed manipulator is designed with redundant joint and non-spherical-wrists. A Modified Jacobian Elementary Transformation (MJET) approach is proposed to determine all of the singularity conditions. This approach has a singularity isolation feature to reduce the computational workload. A Multiple Moving Obstacle Avoidance (MMOA) approach is proposed for manipulator path planning in a compact workspace. A super-quadric surface function is used to describe the shape of an obstacle, and the pseudo-distance from the manipulator to the obstacle is measured and controlled in real time. This approach achieves a good balance between computational complexity and accuracy.The proposed manipulator is modeled using the Lagrangian dynamics formulation and the dynamics of the proposed manipulator is verified with the commercial software ADAMS (Automatic Dynamic Analysis of Mechanical Systems). The mathematics model has similar output in ADAMS under a constant torque input and a sine torque input. The Orbital Replacement Unit (ORU) assembly task is implemented using impedance control. Both simulation and hardware tests are completed for verification, and the experimental results show that the controller is good for on-orbit servicing tasks.The contributions of the thesis are summarized and future work is proposed.Detailed summary in vernacular field only.Detailed summary in vernacular field only.Detailed summary in vernacular field only.Detailed summary in vernacular field only.Detailed summary in vernacular field only.Detailed summary in vernacular field only.Chen, Yongquan.Thesis (Ph.D.) Chinese University of Hong Kong, 2014.Includes bibliographical references (leaves 131-148).Abstracts also in Chinese