TASK-BASED OPTIMIZATION OF MULTI-ARM SPACE ROBOTICS

Abstract

There are many benefits to using multi-arm systems over a single arm system including higher flexibility in planning, better payload handling capacity, and reduction of joint torques. However, multi-arm systems are inherently more complex. This complexity does not necessarily translate to ``bigger" and ``heavier". This research seeks to answer the question of whether or not a multi-arm system can have lower mass than a single arm system.

Using a task-based methodology, Independent single-arm and cooperative dual-arm manipulator systems are designed. A task defines the payload's motion and thus the manipulator's trajectory. Utilizing linear programming, a new method is developed in order to optimize the distribution of forces among the multiple arms in order to guarantee a minimum system mass. The mass of the motors and gears are estimated based on the required torque and speed, obtained from the trajectory and force-distribution.

This study shows that a well-designed multi-arm system can in fact have a lower mass than a single-arm system. Further optimization demonstrates that a multi-arm system, when designed as a complete system rather than individual parts, can significantly reduce the total system mass.