Abstract

Exploring the design space of tensegrity systems is the basis of the work presented in this thesis. The areas explored as part of this research include the optimization of tensegrity structures to minimize the size of a tensegrity structure given payload shock constraints, and the control and locomotion of an icosahedral tensegrity system using movable masses and using an accelerometer in conjunction with leveraging geometrical knowledge of an icosahedral tensegrity system to localize the system after the system moves. In the optimization design space, a simplified model was created to represent an icosahedral tensegrity structure. This was done by assuming that a system of springs could represent an icosahedral system with enough fidelity to be useful for optimization. These results were then validated and tested. The most extensive part of the research preformed was in regards to the control of a Tensegrity Icosahedron. This structure utilized novel locomotion techniques to allow the structure to move by changing its center of mass. Essentially, instead of actuating the system by changing the length of the strings that make up the system state, the system's center of mass is moved using movable masses. These masses make it so the system can rotate about one of the base pivot points. A controller was also created that allows for this system to go to a target point if the state of the system is known. Finally, work was done to attempt to localize a structure by combining a motion model based off the geometry of the structure and a measurement model based on accelerometer readings during the movement of the structure into an EKF. This EKF was then used to localize the structure based on the predicted motion model and the measurement model prescribed by the accelerometer. This allowed for the system's state to be estimated to within 3 standard deviations of the uncertainty of the motion and measurement models. Additional work on this system was also done to make a physical model of the system. This work includes making a bar so that movable masses can pass through it, creating an accelerometer model to roughly determine the system's state, and tracking the systemâ€™s displacement using some steady-state model assumptions.

Degree

College and Department

Ira A. Fulton College of Engineering; Mechanical Engineering

Rights

https://lib.byu.edu/about/copyright/