Modeling and control of quadrotor MAV using vision-based measurement

In this paper we review the mathematics model of quadrotor using Lagrange's equation. We propose a vision-based measurement to stabilize this model. A dual camera method is used for estimating the pose of the quadrotor; positions and attitudes of the quadrotor MAV. One of these cameras is located on-board the quadrotor MAV, and the orther is located on the ground. The control system is developed in Matlab/Simulink. In this paper, we consider a linear controller for our purpose. Linear Quadratic tracking controller with integral action, and Optimal Linear Quadratic Gaussian (LQG) control with integral action are designed for stabilization of the attitude of the quadrotor MAV. Moreover, some measurement noises will be considered in the controller design, too. Finally, this paper will demonstrate how well this control works for a certain flight mission: quadrotor MAV on the ground, and then starting hover, sideway move and keeping the position with a pointing to a certain object fixed in space.