Evaluation and Comparison of State and Disturbance Observers for a Terrestrial Hexacopter

Abstract

Unmanned Aerial Vehicles (UAVs) have been used for a wide range of civil and public domain applications, as well as for missions to Mars, during the last two decades. Multirotor UAV configurations, such as hexacopters, strike a balance between different design considerations, offering improved performance and reliability compared to other options. For autonomous exploration scenarios, Guidance, Navigation, and Control (GNC) subsystems play a critical role in ensuring hexacopter stability and autonomy, especially in GPS-denied environments. However, external disturbances and system uncertainties, such as sensor noise, directly impact both navigation accuracy and control integrity. To address this, two state observers, the Unknown Input Observer (UIO) and the Extended State Observer (ESO) are presented and analyzed. Performance evaluation and comparison of both observers, UIO and ESO, is conducted with a focus on estimating the states of a terrestrial flight hexacopter in the context of an autonomous navigation mission. The underlying hexacopter mathematical model includes external disturbances and system uncertainties, such as aerodynamic drag and sensor noise. The simulations are performed in a Matlab/Simulink environment. Both observers offer robust solutions for estimating system states and compensating for disturbances, and the superiority of the UIO when dealing with random sensor bias is demonstrated.