Feedback Linearization for Model Agnostic Aircraft Simulation Control

Sammanfattning: This thesis presents a method, based on feedback linearization, for controlling a general 3-degree-of-freedom simulation model of an aircraft. Only aircraft models that utilize Taylor approximations of the aerodynamic coefficients are considered. The intent is to be able to simulate the same mission with different aircraft models without having to tune the control system specifically for each model. The method is based on feedback linearization and generates inverse models from the original aircraft model. These are then used to control different variables using a linear controller. Special care was taken to be able to switch the controlled variables during a flight and smooth out the transitions. Two test missions were successfully implemented based on the constant altitude cruise and cruise climb flight strategies respectively. After some tuning of the linearized controller, the desired trajectories were followed well. However, additional tuning and analysis of the linear controller could be performed to ensure more specific performance targets are met. Some work should also be done to make the method easier to use with regards to the mission specification the user provides. A better mission segment handler and state event based segments would reduce the requirements on the mission specification and thus make it easier to define missions that work for multiple aircraft.