DİNLE

PhD Dissertation Defense:Hammad Zaki11-07-2019

Robust Hovering and Trajectory Tracking Control

of a Quadrotor Helicopter Using Acceleration Feedback

and a Novel Disturbance Observer

 

 

Hammad Zaki
Mechatronics, PhD Dissertation, 2019

 

Thesis Jury

Prof. Dr. Mustafa Ünel (Thesis Advisor), Asst. Prof. Dr. Meltem Elitaş,

Asst. Prof. Dr. Hüseyin Özkan, Prof. Dr. Şeref Naci Engin, Asst. Prof. Dr. Ertuğrul Çetinsoy

 

 

Date & Time: July 17th, 2019 – 10:00

Place: FENS G029

Keywords: Robust Control, Acceleration Feedback, Disturbance Observer, Quadrotor,

                 Hierarchical Control, Sliding Mode Control, Nonlinear Optimization

 

Abstract

 

Hovering and trajectory tracking control of rotary-wing aircrafts in the presence of uncertainties and external disturbances is a very challenging task. This thesis focuses on the development of the robust hovering and trajectory tracking control algorithms for a quadrotor helicopter subject to both periodic and aperiodic disturbances along with noise and parametric uncertainties. A hierarchical control structure is employed where high-level position controllers produce reference attitude angles for the low-level attitude controllers. Reference attitude angles are usually determined analytically from the position command signals that control the positional dynamics. However, such analytical formulas may produce large and non-smooth reference angles which must be saturated and low-pass filtered. In this thesis, desired attitude angles are determined numerically using constrained nonlinear optimization where certain magnitude and rate constraints are imposed. Furthermore, an acceleration based disturbance observer (AbDOB) is designed to estimate and suppress disturbances acting on the positional dynamics of the quadrotor. For the attitude control, a nested position, velocity, and inner acceleration feedback control structure consisting of PID and PI type controllers are developed to provide high stiffness against external disturbances. Reliable angular acceleration is estimated through an extended Kalman filter (EKF) cascaded with a classical Kalman filter (KF).

 

This thesis also proposes a novel disturbance observer which consists of a bank of band-pass filters connected parallel to the low-pass filter of a classical disturbance observer. Band-pass filters are centered at integer multiples of the fundamental frequency of the periodic disturbance. Number and bandwidth of the band-pass filters are two crucial parameters to be tuned in the implementation of the new structure. Proposed disturbance observer is integrated with a sliding mode controller to tackle the robust hovering and trajectory tracking control problem. The sensitivity of the proposed disturbance observer based control system to the number and bandwidth of the band-pass filters are thoroughly investigated via several simulations. Simulations are carried out on a high fidelity model where sensor biases and measurement noise are also considered. Results show that the proposed controllers are very effective in providing robust hovering and trajectory tracking performance when the quadrotor helicopter is subject to the wind gusts generated by the Dryden wind model along with plant uncertainties and measurement noise. A comparison with the classical disturbance observer-based control is also provided where better tracking performance with improved robustness is achieved in the presence of noise and external disturbances.