Deployment and shape change of a tensegrity structure using path-planning and feedback control

Abstract

Tensegrity structures are pin-jointed assemblies of struts and cables that are held together in a stable state of stress. Shape control is a combination of control-commands with measurements to achieve a desired form. Applying shape control to a near-full-scale deployable tensegrity structure presents a rare opportunity to analytically and experimentally study and control the effects of large shape changes on a closely coupled multi-element system. Simulated cable-length changes provide an initial activation plan to reach an effective sequence for self-stress. Controlling internal forces is more sensitive than controlling movements through cable-length changes; internal force-control is thus a better objective than movement-control for small adjustments to the structure. The deployment of a tensegrity structure in previous work was carried out using predetermined commands. In this paper, two deployment methods and a method for self-stress are presented. The first method uses feedback cycles to increase speed of deployment compared with implementation of empirically predetermined control-commands. The second method consists of three parts starting with a path-planning algorithm that generates search trees at the initial point and the target point using a greedy algorithm to create a deployment trajectory. Collision and overstress avoidance for the deployment trajectory involve checks of boundaries defined by positions of struts and cables. Even actuator deployment followed by commands obtained from a search algorithm results in the successful connection of the structure at midspan. Once deployment at midspan is achieved by either method, a self-stress algorithm is implemented to correct the position and element forces in the structure to the design configuration prior to in-service loading. Modification of deployment control-commands using the feedback method (with twenty cycles) compared with empirically predetermined control-commands successfully provides a more efficient deployment trajectory prior to midspan connection with up to 50% reduction in deployment time. The path-planning method successfully enables deployment and connection at midspan with a further time reduction of 68% compared with the feedback method (with twenty cycles). The feedback control, the path-planning method and the soft-constraint algorithm successfully lead to efficient deployment and preparation for service loading. Advanced computing algorithms have potential to improve the efficiency of complex deployment challenges.