Optimal Control Aided by the ELECTRE III Multicriteria Decision Method Applied to Quadcopter Autopilot

Abstract

Unmanned aerial vehicles, such as quadcopters, can perform several activities. To enable such activities, it is necessary to use control techniques to guarantee the completion of the task amid environmental adversities. In this work, multivariable control was used, known as Linear Quadratic Regulator, which features optimal control based on weighting matrices, which were tuned based on the Elimination and Choice Expressing Reality III multicriteria decision method, where the criteria are the desired performance and robustness characteristics for the task the drone was assigned to. This application is used as a basis for an estimated model based on experimental flights of Parrot’s AR. Drone 2.0 quadcopter, thus introducing actual flight conditions into the proposed control project. The simulated results show that the controller tuning, selected with the aid of the Elimination and Choice Expressing Reality III method, obtained performance indices (Integral Squared Error and Integral Quadratic Control effort indices) and robustness (gain margin and phase margin) consistent with the objectives of the control project, thus establishing itself as a practical alternative in helping the designer in choosing controller parameters.