PID AND LQR CONTROLLER ON FLYING INVERTED PENDULUM WITH QUADROTOR

• Main Author: Likafia , Aresti
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/21207
• Institution: Institut Teknologi Bandung

Growth of transportation technology is an inevitable factor for a country development. Nowadays, the technology of transportation has grown fast and it leads to Personal Aerial Vehicle (PAV) technology which can give an easiness for a person in mobilisation. One of PAV type is flying hoverboard. In this study, flying hoverboard was modelled by using flying inverted pendulum with quadrotor and analog joystick was used to measure the alteration angle of the pendulum. The purpose of this study is to create a control system of flying inverted pendulum with quadrotor which ables balancing inverted pendulum at 1 degree of freedom (DOF) and makes it stable at a constant height. The system was controlled by using Proportional-Integral-Derivative (PID) controller and Linear Quadratic Regulator (LQR) controller. A boundary condition of pendulum movement was set so its movement was limited to roll movement only to make sure the safety of flying experiments. <br /> <br /> <br /> The PID controller is model based designed. The model created by identifying the dynamics of inverted pendulum and the y-axis velocity of quadrotor versus the electrical pulse of Arduino Uno. The PID control system consists of two controllers which parameters are exchanged at a specific angle. According to the experiment’s results, the implementation of the PID controller design still requires fine tuning to obtain a satisfactory output response. The best response is given by PID controller with paramaters for small K_p= 20/°, big K_p = 50/°, K_i=5/°s, and K_d=25s/°, which provides zero steady state error and stable roll angle of inverted pendulum. On the other hand, the LQR controller is designed by converting Newton's model from the inverted pendulum dynamics into the state space. The addition of a digital low pass filter to accelerometer sensor with the the feedback gain K=[-7.05°/m-1.68°s/m-0.45°/m-0.52°s/m] gave better response performance. It can achieve the control objective to create zero stable and zero steady state error of roll angle inverted pendulum, and the quadrotor y-axis position. The stabilization cannot be carried out by the LQR controller without a low pass filter. <br />