A Linear Extended State Observer With Scalable Bandwidth-Based Advanced Internal Model Attitude Control for a Quadrotor

Abstract

Quadrotors are popular unmanned aerial vehicles, but the unmodeled dynamics, parameter variations, and external disturbances may cause performance deterioration. Active disturbance rejection control has been shown as a feasible solution so long as the dynamic variations and disturbances are within its frequency range, outside which the closed-loop performance is difficult to maintain. To this end, a linear extended state observer with scalable bandwidth is proposed to reconstruct the total disturbance more accurately and timelier. This new observer is also less susceptible to sensor noises, since its bandwidth intelligently varies. Then, it is combined with an internal model control (IMC) to form the core of the proposed solution for the attitude control of quadrotors with a dual loop structure. Improved estimation of the total disturbance and hence its removal from the plant dynamics allows better control performance because the internal model on which the control design is based is now more authentic. The advantage of the proposed approach is shown in providing a novel and practical solution for the IMC design of complex nonlinear systems. Also given is the stability analysis of the closed-loop system.