Model Predictive and Adaptive Controller Design of Tri-Tilt Rotor VTOL UAV by Using Multi-Body Modeling

This paper investigates the Tri-Tilt Rotor VTOL UAV. The aim of this study is to represent a comprehensive dynamic model, eleven degree of freedom at six flight phases (hover, descend, climb, forward, transient and cruise) and control the vehicle to reach best flight condition. For this purpose, the vehicle equations of motion are derived in tensor form and have been expanded in the coordinate systems, based on multi-body (vehicle and three electric motors) modeling approach in order to consideration of motors gyroscope effects on flight dynamic. Depending on vehicle flight phase, propulsion and aerodynamic forces and moments are determined separately. Blade Element Momentum Theory (BEMT) is used to obtain motors propulsion forces and moments at hover, descend, climb and forward phases. After that, with utilizing of controller for each channel flight, the trim condition is calculated and then for the sake of linearization using analytical method, dynamic and control matrixes are derived. This calculated model is qualified as linear model in order to design the model predictive and adaptive controller. For climb phase, as the nonlinear model receding from linear model, the linear model predictive controller performance was diminishing whereas the function of model reference adaptive control in spite of the uncertainties was better.