Kane’s Method Based Dynamic Modeling and LQR Balancing Control of a Unicycle Mobile Robot

• Main Authors: LIN, SHIH-CHEN, 林昰辰
• Other Authors: LIN, NAN-ZHOU
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/64594h

碩士 === 逢甲大學 === 自動控制工程學系 === 106 === The purpose of this thesis aims to derive equations of motion and to perform analysis of motion for a single-wheel robot. In order to derive equations of motion, coordinate systems are defined and Kane's method is applied. Then, a numerical simulation is established based on the equations of motion and with MATLAB/Simulink. In addition, the robot is equipped with one driving wheel and one flywheel, which are used to control the pitch angle and the roll angle of the robot body, respectively. Finally, verification is conducted by using numerical simulation program. The influence and impact on the robot body due to driving torque and flywheel torque are studied and discussed. Feedback controller is designed by using LQR method. According to simulation results, motor, battery and components are selected to design robot. Moreover, algorithms are applied here to complete the design of the robot. The motor-related parameters are measured by experiments. Regression analysis is utilized to find the motor characteristic plane, by which the output torque command from feedback controller is converted into voltage signal to drive the motor. Finally, the human machine interface is designed, which allows user to adjust the parameters easily and to verify balance condition efficiently.