Abstract: This experiment implemented a full-state feedback controller on a Furuta Pendulum to keep the system stable in the upright, inverted pendulum case. The system was tested with its pendulum in the up and downward positions, while tracking a 30°, 0.1 Hz square wave motor impulse, for controller gains that were determined both through manual tuning and with Ackerman’s pole placement method. The results showed that for the up pendulum case, the manually tuned gains worked best, with a pendulum position RMS error of 6.22° and a motor effort level of 0.0729 %V to be stable at the 30° base link square wave, while the Ackerman’s gains caused system instability at a 20° square wave. The down-pendulum case saw agreement between both the manually-tuned and Ackerman’s controller gain methods, with, respectively, pendulum RMS errors of 6.12° and 4.70° and motor effort levels of 0.056 %V and 0.055 %V for tracking the 30° square wave and maintaining stability.

Full Report Found Here.

Note: Project conducted with Ryan Earl.

Abstract: This experiment was conducted to model and test the effectiveness of a full-state feedback controller applied to an inverted Furuta pendulum system. This pendulum system was intended to simulate a waiter carrying a tray with drinks. The controller was tested on a stationary system before being tested on a mobile system. Using Ackerman’s pole placement method, system gains were calculated for various payload conditions as well as various pendulum arm lengths. LQR method was also used to determine gains for various R-values. The RMS position errors (base link and pendulum arm) and command efforts were used to comparatively analyze the behavior of the various test conditions.

Full Report Found Here.

Note: Project conducted with Jeff Clark and Ryan Earl. Report reflects results from stationary system tests.