Control Design of Quadcopter using Linear Quadratic Gaussian (LQG)

Abstract

A quadcopter is an unmanned control aircraft with four rotors and is mounted in a square formation. Quadcopter moves with six degrees of freedom that are divided into two types of movements, there are rotational motion and translational motion. The Linear Quadratic Gaussian control method is used in this research, which combines the Linear Quadratic Regulator with Kalman Filter. LQG control was chosen because it can estimate state variables that are not measured, and it can save costs and time. The variables controlled and analyzed in this study are changes in the angle of rotational motion including roll, pitch, and yaw with a fixed linear position. The quadcopter mathematical model has been linearized and shows stability, controllability, and observability. Simulation results show that the system produces an overshoot less than 0.1 rad with a time stable less than 1 second.