Available at: https://digitalcommons.calpoly.edu/theses/3370
Date of Award
6-2026
Degree Name
MS in Mechanical Engineering
Department/Program
Mechanical Engineering
College
College of Engineering
Advisor
William R. Murray
Advisor Department
Mechanical Engineering
Advisor College
College of Engineering
Abstract
Ball Balancer or Ball on Plate systems are used for education and research in control theory. These consist of a plate which can be actively tilted, with a ball rolling on the surface. Students are challenged to model the dynamics of the system and implement a control loop which is capable of stabilizing the ball by actuating the plate. The mechatronics group at Cal Poly is developing hardware for one of these ball-balancer systems, for future use in the Mechanical Engineering curriculum.
The position of the ball relative to the plate must be measured in order to control and stabilize it. To more fully characterize the system, the velocity of the ball should also be measured or estimated, as well as the angular position and velocity of the plate. For our purposes, this data should be of reliable quality for direct input into the controllers developed by students, without further filtering.
In previous iterations of the design, the motion information for the ball came from a resistive touch panel. However, the quality of data which resulted from this limited the performance of the system as a whole. Specifically, the ball position data had a significant random noise component, and this was greatly amplified in the velocity estimation. Variation of the contact force with the plate as it moved caused variable and invalid position readings.
This thesis implements an alternative system for detecting the position and velocity of the ball. Using two line-scan cameras, the ball is located by triangulation without direct contact.