130. dynamic model and control of a micro ball-balancing robot with high yaw rate

Department: Mechanical & Aerospace Engineering
Research Institute Affiliation: Center for Control Systems and Dynamics (CCSD)
Faculty Advisor(s): Thomas R. Bewley

Primary Student
Name: Eric Nauli Sihite
Email: esihite@ucsd.edu
Phone: 614-596-5240
Grad Year: 2018

Abstract
A ball-balancing robot (BBR) is a robot which balances on top of a ball, which is actuated by the motors through three omni-directional wheels. This robot can freely translate on a plane in any direction and can also change its heading while translating as well. The BBR has a complex 3-D nonlinear dynamics, so the controller used to balance this robot assumes the robot dynamic model as a decoupled mobile inverted pendulum problem in its roll and pitch directions. In addition to that, most BBRs are large (> 1 meter), heavy and have relatively slower dynamics. Our lab have developed a mini ball-balancing robot (MBBR), which is 25 cm tall and weighs 675 g, and successfully balanced it. When controlling the MBBR, we found out that the controller was significantly less stable when the robot is translating and rotating at the same time. This difficulty leads us to investigate for a better controller which works well under high yaw rate. Therefore, we developed a new dynamic model which allows for a significant yaw rate which can be represented in a linear time variant state space model. This model enabled us to develop a gain scheduled state feedback controller, which updates the gain depending on the current yaw rate of the robot.

Industry Application Area(s)
Control Systems | Robotics

« Back to Posters or Search Results