Quadrotor Motion Control using Adaptive Generalized Dynamic Inversion

Abstract

A novel two-loops structured control system based on Adaptive Generalized Dynamic Inversion (AGDI) is proposed for the quadrotors. The outer-loop is responsible to generate the desired pitch and roll attitude commands based on the positional errors to the inner-loop, which in turn produces the tilting angles that are required to control its position. The inner-loop is also responsible for tracking the desired altitude and yaw attitude commands. In AGDI control, the particular part enforces the constraint dynamics based on the quadrotor state deviation functions and are inverted using Moore-Penrose Generalized Inverse (MPGI) to realize the control law. The asymptotic stability of the linear and angular velocities is guaranteed by using the null control vector. The singularity problem is addressed by incorporating a dynamic scaling factor in the MPGI expression. To enhance robustness attributes, an additional sliding mode control with adaptive modulation gain is integrated, such that semi-global practically stable position and attitude tracking is guaranteed. Numerical simulations are conducted on a six degrees of freedom simulator of X-4 flyer quadrotor, to demonstrate the effectiveness of AGDI control under nominal and perturbed flight conditions.