Cascaded Extended-State-Observer-Based Synergetic Control for Quadcopter Translational Dynamics

Abstract

This paper presents an approach to quadcopter position control, utilizing a Cascaded Extended State Observer (CESO) integrated with synergetic control. The proposed control strategy enhances the quadcopter's stability and tracking accuracy by estimating and compensating for aerodynamic disturbances and drag forces to a significant extent, which are challenging to measure or model analytically. This extent increases as the levels of the cascaded structure grow, progressively enhancing both accuracy and compensation capability. An efficient tuning approach is introduced in the paper for tuning multiple ESOs in a cascaded structure that uses hierarchical gain reduction, ensuring distinct frequency ranges for each observer. This achieves a rapid initial estimation while reducing noise in later stages, enhancing stability and robustness. The CESO framework, combined with synergetic control, offers a robust solution, minimizing mean squared error and control effort while improving disturbance rejection. The PX4-ROS2 architecture was used to test our system in Gazebo and on a custom-built quadcopter experimentally, validating the efficacy of the proposed control scheme. This study contributes significantly to the development of advanced control techniques for unmanned aerial vehicles, emphasizing practical implementation and adaptability in real-world scenarios.