Adaptive Neural Network Controller for Quadrotor with Two Degree of Freedom Robotic Arm

Abstract

Modeling of the quadrotor with a two degree-of-freedom robotic arm is performed, and baseline controller and adaptive controller are designed for the attitude control of the quadrotor. Newton-Euler dynamics and Articulated body dynamics are used to model the quadrotor with a robotic arm. For a case that a robotic arm in the quadrotor does not hold payload, a baseline controller is designed using sliding mode control scheme. If the robotic arm holds any unknown payload, then some moment is generated from the robot arm. The generated moment can be considered as uncertainty in the quadrotor system. To compensate the effects of the uncertainty, adaptive neural network controller is designed. Numerical simulation is performed to verify the performance of the proposed controller for a quadrotor system with uncertainty.