Trajectory planning and control of manipulators based on tensegrity structures with manipulability and modal properties optimization

Abstract

This paper presents a control methodology for robots based on tensegrity structures with redundant actuators. This methodology is divided into two tasks: trajectory planning and shape control. The basis of the trajectory planning is the workspace exploration using the Dynamic Relaxation method, followed by the interpolation and key manipulator properties optimization of the rod angles and the cable forces with respect to the desired end-effector trajectory. The Computed Torque Control method extended with Singular Value Decomposition is used for the shape control. The methodology is demonstrated on a virtual prototype of planar manipulator. The values of the tracking errors of the end-effector trajectory confirm its functionality.