Robust disturbance observer-based control of unmanned helicopter

Abstract This paper considers robust disturbance observer design for an autonomous helicopter. The disturbance observer is a well-established method originally developed for the mitigation of disturbance effects in minimum-phase single-input single-output systems. Recently, the disturbance observer design has been extended to non-minimum-phase multiple-input multiple-output systems using an $${\mathcal {H}}_{\infty }$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>H</mml:mi> <mml:mi>∞</mml:mi> </mml:msub> </mml:math> method. Due to the applicability to non-minimum-phase systems, the new disturbance observer design becomes also attractive for aerospace applications. However, there are no guaranteed stability margins that can be achieved with the $${\mathcal {H}}_{\infty }$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>H</mml:mi> <mml:mi>∞</mml:mi> </mml:msub> </mml:math> -based design approach. In fact, stability margins may easily be degraded when the disturbance observer is integrated into an existing controller-plant closed-loop. In this paper, a robust disturbance observer design is proposed. Therefore, complex uncertainties are introduced to model gain and phase variations. Robustification of the observer is achieved using design principals of $$\mu$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mi>μ</mml:mi> </mml:math> -synthesis. As consequence, the proposed method enhances the robust performance of the observer and leads to less degradation of stability margins in comparison to the original design procedure. An application example is provided which considers mitigation of wind effects for the inner-loop control of an unmanned helicopter.