Optimal coordinated planning strategy for space robots grasping targets

Abstract

The task of capturing a space target using a space robot carrying robotic manipulators can be divided into two phases: controlling the robot to reach a desired berth position near the floating target, and manipulating the manipulator to grasp the target. Most of the existing planning strategies only focus on a point-to-point trajectory planning of the manipulator to ensure the end-effector to reach a desired pose, while the berth position is usually set manually. In this paper, we propose an optimal coordinated planning strategy for space robots to grasp targets. Both the manipulator's joint trajectories and the robot's berth position are considered to autonomously accomplish the coordinated planning of the end-effector and the robot base. From the perspective of the task-space constraints, the manipulator is first planned to satisfy the attitude constraints of the end-effector and base, and then the berth position is planned to meet the end-effector's position. A Modified Generalized Jacobian Matrix (MGJM) is proposed to describe the velocity mapping from the joint space to the attitude of the end effector and the base. Based on the MGJM and Particle Swarm Optimization (PSO), the whole planning task is turned to optimize the joint trajectories and the berth position. Simulated experiments have verified that the proposed planning strategy can successfully automate the two phases of a space robot to grasp target in the space.