Research On Sampling-Based Motion Planning For Target Capture Of Freefloating Space Robots

In the future space robots will be an important direction of space technology.The on-orbit service mission carried out by space robots can significantly reduce the cost and risk compared with astronauts' extra-vehicular activities,and complete more complex operations.The free-floating space robot is one type that has an uncontrolled satellite base,which only controls joint rotation to complete various operations.In this paper,the application of sampling-based algorithm in the target capture motion planning of free-floating space robots is studied.This paper introduces a kinematic modeling method for free-floating space robots,and analyzes challenges to the motion planning problem from the system's coupling,nonintegrity and nonlinearity.On this basis,the modeling and simulation experiment of the kinematics of the six-DOF free-floating space robot with UR10 as the manipulator was completed,which lays a foundation for the next discuss.We improve the original RRT algorithm(Rapidly-Expanding Random Tree),and deal with the simplified motion planning problem with it.The simplified motion planning problem takes the joint speed as the control,and the limit of the joint angle,angular velocity and the disturbance of the base into consideration.The main improvements to RRT in this paper are the inclusion of the goal bias in work space and the propagation strategy of transpose jacobian matrix.Sampling-based algorithm is proved usable by simulation.A more advanced sampling motion planning algorithm,Stable Sparse-RRT(SST),is improved to solve the problem of capture motion planning for free-floating space robots in a more common situation.The SST algorithm adds the locally optimal concept and pruning operation on the basis of RRT-Extend,providing some more reliable properties.The capture motion planning problem takes joint angular acceleration as the control input,and on the basis of the above,the constraint of joint angular acceleration,collision avoidance and the desired(angular)velocity of the end effector are considered.In the framework of SST,this paper improves the expansion efficiency with goal bias and the guidance of task space metric,and introduces the rolling planning of task space sampling to speed up the algorithm to find a feasible solution,and realizes the high-dimensional and multi-constraint target capture motion planning of free-floating space robot.