Nonlinear Predictive Control Design for Quadrotor Trajectory using Sequential Dynamic Operability Framework

Abstract

this paper demonstrates the computation of controller parameters to minimize the errors of a quadrotor trajectory despite the presence of some wind disturbances. The nonlinear quadrotor model depicts the dynamics generated by a system of four brushless DC-motors, gear boxes and propellers, and the generic 6 degree-of-freedom rigid-body equation derived with the Newton-Euler formalism. Wind disturbances are assumed affecting the speed in horizontal direction. A set of Proportional – Derivative (PD) control structure is used to achieve a desired quadrotor position based on earth’s xyz coordinates, and maintain small roll and pitch angles. The PD parameters is computed by utilizing the dynamic Nonlinear Programming (NLP) section of the Dynamic Operability Framework. The framework selects the control parameters for a receding prediction horizon. This method allows smooth changes of optimum control parameters throughout 11 trajectory segments. These segments represent movements in x, y, z and diagonal directions. The variation of control parameters is presented.