Adaptive Observer-based Super Twisting Sliding Mode Control of Quadcopters for Trajectory Tracking and Low Altitude Grasping

Abstract

Aerial grasping is revolutionizing Industry 4.0 through robotics. However, achieving stable flight and precise grasping in the face of disturbances, especially during complex trajectory tracking and low altitude flights, remains a critical challenge. This thesis addresses position tracking and low altitude stabilization by designing, simulating, and experimentally demonstrating a control strategy to compensate for real-world unmodelled system disturbances. The method involves a robust Super-Twisting Sliding Mode Controller (STSMC) based on an Adaptive Higher-Order Sliding Mode Observer (AHOSMO), which handles system disturbances like payload variations, wind gusts, and ground effects. This approach ensures robust control insensitive to unmodelled dynamics, parametric uncertainties, and external disturbances while attenuating the chattering phenomena. The AHOSMO estimates disturbances using varying observer gains, enhancing disturbance rejection. The strategy accommodates sensor noise, and experimental validation on a COEX Clover drone with a rigid gripper mechanism showcases its effectiveness over standard control methods, highlighting its potential in real-time applications.