Altitude tracking controller design for a small unmanned aerial vehicle based on linear-quadratic-Gaussian-integral control

Abstract

This paper presents the design procedure of a digital altitude tracking controller based on optimal control theory. The controller governs the longitudinal dynamics of a small-scale fixed-wing unmanned aerial vehicle (UAV) via an inner-outer loop control structure. In particular, the inner feedback loop maintains the aircraft's longitudinal stability and controls the climb rate (vertical velocity) while the outer-loop tracks the desired altitude through barometric altitude feedback. The inner-loop controller is designed based on the linear-quadratic-Gaussian-integral (LQGI) approach which necessitates the aircraft's longitudinal model to be expressed by a state-space representation. In this work, the model is obtained via a system identification technique called the observer/Kalman filter identification (OKID) method. Actual flight test data is collected and used in the system identification algorithm. On the other hand, the outer-loop controller simply features a proportional-gain control which eliminates the error between a reference altitude and the aircraft's altitude measured in real-time. The proposed altitude tracking controller is implemented on the National Cheng Kung University's Swallow UAV system and has been validated through a series of successful flight tests.