Model-free adaptive multivariable control of nonlinear twin-rotor dynamics in the presence of wind gust: A comparison of different approaches

Abstract: This study proposes a data-driven model-free adaptive controller (DDMFAC) for a Multi Input-Multi Output (MIMO) twin-rotor aerodynamic system. The primary objective is to achieve precise beam tracking and accurate positioning in two degrees of freedom (DOF) for azimuth and pitch control. To this end, Three MFAC algorithms— Compact Form Dynamic Linearization (CFDL), Partial Form Dynamic Linearization (PFDL), and Full Form Dynamic Linearization (FFDL)—are employed for azimuth and pitch tracking, evaluated under both wind-free and windy conditions. A comparative analysis showcases the performance and tracking capabilities of these three model-free methods and their efficacy in controlling the MIMO twin-rotor system under various conditions. Additionally, the adaptive laws are designed to enhance robustness against qind gust, ensuring riliable control under external disturbance