State space identification and implementation of H∞ control design for small‐scale helicopter

Abstract

PurposeThe purpose of this paper is to design a model‐based robust controller for autonomous hovering of a small‐scale helicopter.Design/methodology/approachThe model is developed using prediction error minimization (PEM) system identification method implemented to flight data. Based on the extracted linear model, an H∞ controller is synthesized for robustness against parametric uncertainties and disturbances.FindingsThe proposed techniques for modelling provide a linear state‐space model which correlates well with the recorded flight data. The synthesized H∞ controller demonstrates an effective performance which rejects both sinusoidal and step input disturbances. The controller enables the attitude angle follow the reference target while keeping the attitude rate constant about zero for hover flight condition.Research limitations/implicationsThe synthesized controller is effective for hovering and low‐speed flight condition.Practical implicationsThis work provides an efficient hovering/low‐speed autonomous helicopter flight control required in many civilian UAV applications such as aerial surveillance and photography.Originality/valueThe paper addresses the challenges of controlling a small‐scale helicopter during hover with inherent modelling uncertainties and disturbances.