Sung, Hsin-Chuan (2015-05). Balancing Robot Control and Implementation. Master's Thesis.
This thesis presents several control methods on one wheeled and two wheeled balancing robot and implementation of a two wheeled robot. The purpose is to balance the robot at a standstill position and let it move smoothly. The first step is to model the robot by Lagrangian method. After getting the equation of motion of the robot, some linear and nonlinear controllers-PID, LQR, and Sliding Mode etc. are tested on the model in Matlab. Second, build a balancing robot by LEGO-Mindstorm EV3 package then test the simulation result on the LEGO-Mindstorm EV3 robot to see how the simulation results work and perform. The simulation results show that PID controller controls only the body's angle but no wheel's position. LQR controller controls all the states which are body's angle and wheel's position. Sliding Mode controller also can coverage both body's angel and wheel's position at the same time. By adding a PID controller on a LQR controlled two wheeled robot, the robot can move to the front and back by control the state which is the wheel's speed. The LQR and PID controllers are tested on the LEGO-Mindstorm EV3 robot. The experiments show the simulation is valid and physically achievable. The robot can be used in many situations to help people reach stuffs and pass lanes that human can't. Robots can overcome many human's limitation. The more researches goes on the balancing robot, more the robot can help people in different ways.