Nonlinear Model Predictive Control of Quadrotor Using Direct Multiple Shooting

Quadrotors have been applied to many areas. Control of the quadrotor for tracking the desired trajectory is challenging. Several traditional control approaches using PID controllers have been applied to control UAV. Such the approach requires the model to be linearized. This paper proposed to apply the nonlinear model predictive control to guide the quadrotor while getting collision free by using the concept of the hyperplane moving constraints. To address the NMPC, the multiple shooting method has been applied to discretize the optimal control problem to the nonlinear program (NLP) and it is efficiently solved by the standard NLP solver. Numerical computation has been shown that the application of the multiple shooting method gives the more efficient NLP with a less number of optimization variables.