SYSTEM IDENTIFICATION AND CONTROLLER DESIGN OF PROPELLER DRIVEN PENDULUM (BI-COPTER)

Undergraduate control courses are usually significantly improved when in-class instruction is backed up with experiments aimed at providing the students an opportunity of experiential learning. The rapid development of computer multimedia technology has resulted in a tendency toward a hands-on laboratory that incorporates simulation tools and computer-controlled equipment. However, most of the open-source hardware prototyping platforms require programming environments, installation of additional software, and additional plug-in modules that hinders the large-scale deployment of such systems. Therefore, the goal of this project was to create a simple mechatronic system (the "Bi-Copter") that uses broadly available components along with MATLAB computing environment, in order to demonstrate an experiment that can be adopted at any educational institution with access to MATLAB. The main part of the setup consists of two small dc electric motors attached to each side of the beam with the propellers. By controlling the motors, the angular position of the beam can be changed to a desired one. A circuit board utilizing a breadboard and off-the-shelf plug-in microcontroller connected to a serialto-USB cable is used to control the motors. Unlike traditional controls experiments which start with a modelling task based on first principles, in this experiment a black-box model autoregressive model with exogenous input (ARX) is used. Model parameters are obtained from open-loop experiment. The extracted model is then used to design a feedback controller and evaluate its performance. The bicopter system was tested with over 50 students coming from the courses "Design of Mechatronic Systems"and "System Identification"of the academic years 2020-2022. The activity lasts eight weeks including 6 home assignments. During this period students work in groups of two and learn how to program the micro-controllers, carry out the system identificaiton experiments, design and implement feedback controllers. The laboratory allows students to build models of various complex- *Corresponding author: [email protected] Documentation for asmeconf.cls: Version 1.34, August 24, 2023. ity. To provide laboratory experience to students with various levels of expertise, this lab makes use of the Control Systems and System Identification toolboxes of MATLAB. Both graduate and undergraduate students utilized the apparatus and completed the design tasks. The bi-copter project was tested in a Design of Mechatronic Systems course at the University of Arizona. Since mechanical engineering students typically lack formal training in system identification methods, the course includes a brief introduction of this topic. One of the main outcomes from their experiments is appreciation of the importance of selecting proper input test sequence and the sensitivity of the resulting model to the bandwidth and amplitude of the input test sequence.