Robustness investigation of a ducted-fan aerial vehicle control, using linear, adaptive, and model predictive controllers

Abstract

A comparison of three common controllers for stabilising a vertical take-off and landing air vehicle is presented. RMIT is a small sized tail-sitter ducted fan air vehicle with a particular configuration layout, multiple control surfaces, low weight, and high-speed flight capability. The main problem here is control effectiveness at low flight speeds and transition manoeuvres because of the inherent instability. In the current study, a comprehensive nonlinear model is firstly developed for RMIT, followed by a validation process. Subsequently, linear, adaptive and model predictive controllers are designed in vertical flight. Based on the simulation results, it is shown that the linear controller is not able to eliminate the inherent instability of the vehicle in hover, while the adaptive and model predictive controllers are able to tune for stabilisation. However, simulation results show that adaptive controller cannot eliminate the large external disturbances from the system, while the MPC has a robust behaviour against them and provide an acceptable closed-loop performance over a wide range of external disturbances.