Optimal Predefined-time Trajectory Planning for a Free-floating Space Robot

TL;DR

This paper proposes a fast, collision-avoidance trajectory planning method for free-floating space robots that achieves micron-level accuracy and low joint velocity using a CDF algorithm and GA optimization.

Contribution

It introduces a novel predefined-time trajectory planning strategy with collision avoidance and joint velocity optimization for space robots.

Findings

Achieves micron-level end-effector tracking accuracy.

Effectively avoids collisions in complex space environments.

Reduces joint angular velocities through optimization.

Abstract

With the development of human space exploration, the space environment is gradually filled with abandoned satellite debris and unknown micrometeorites, which will seriously affect capture motion of space robot. Hence, a novel fast collision-avoidance trajectory planning strategy for a dual-arm free-floating space robot (FFSR) with predefined-time pose feedback will be mainly studied to achieve micron-level tracking accuracy of end-effector in this paper. However, similar to control, the exponential feedback results in larger initial joint angular velocity relative to proportional feedback. Firstly, a pose-error-based kinematic model of the FFSR will be derived from a control perspective. Then, a cumulative dangerous field (CDF) collision-avoidance algorithm is applied in predefined-time trajectory planning to achieve micron-level collision-avoidance trajectory tracking precision. In the end, a GA-based optimization algorithm is used to optimize the predefined-time parameter to obtain a motion trajectory of low joint angular velocity of robotic arms. The simulation results verify our conjecture and conclusion.