Coordinated Motion Planning Study Of Dual-arm Space Robot For Capturing Spinning Target

With the development of space technology, space activities of human have become increasingly frequent. Space robot will play an increasingly important role in efficiency improvement and astronauts’ risk reduction. Target autonomous capture is the key technology of space robotic applications. This thesis was based on a project sponsored by the National High-Tech863Program in aerospace industry, and researches coordinated motion planning for spinning target capturing, which lays the foundation for the practical of future services technology.In fact, targets to be serviced in orbit are failure satellite, space debris, etc. These objects are in spin motion state that it is very difficult to capture them. Based on extensive research, we summed up the characteristics of the real movement of the spinning target and established its kinematic model which can be used to predict its movement by Unscented Kalman Filter (UKF), and then methods of autonomous capture and coordinated motion planning were proposed to capture the spin target successfully.Due to the existed of dynamic coupling, the motion of the arms will alter the attitude and position of the base, which can affect the implementation of the task. Firstly, established a “base centroid equivalent manipulator” model, then the singularity-free method was proposed to plan the coordinated trajectories of the balance arms. With solving inverse kinematics of the equivalent manipulator instead of solving the differential motion equations to determine the trajectory of the balance arm, the jacobian matrix singularity is avoided. Furthermore, there are not any specific constraints on the configuration, mass properties, geometry parameters and mounted position of the balance arm.The captured coupling system may be unstable due to the change of the quality characteristics and momentum, thus, two coordinated plan methods—joint damping and parameters of joint function configurable, were proposed to solve this problem. According to these methods, management and re-distribution of angular momentum were used to stabilize the coupling system by planning and controlling the velocity of the flywheel and joints. On the basis of them, PSO algorithm was used to optimize the parameters of joint function in order to ensure the minimum deflection of the base attitude. These methods adopted the actual flywheel as momentum exchange device, which can obtain great effectiveness in engineering. In order to verify the motion planning methods, a ground experiment system was developed based on the idea of dynamics simulation and the principle of kinematic equivalent. Then two methods of motion simulation, which were based on6-DOF industrial robot, were proposed: the conventional simulation method and shoulder singularity method. The second method owned a strong engineering feasibility, which took the advantage of the motion characteristics of the shoulder singular and considered actual joint constraint. The experimental system used commercial devices which was relatively low cost and easily to be implemented. In this system, the path plan method of spin target autonomous capture by space robot could be verified and evaluatedThe research in this dissertation is focused on the autonomous capture of spinning target which has to be settled urgently in servicing mission on high orbit in the future. It conforms to the trend on the development of on orbit service using space robot in our country, and has an important theoretical and practical significance in orbital maintenance, assembly, and rescue using space robot.