An experimental verification of ADRC robustness on a cross-coupled Aerodynamical System

This paper presents a robustness study of an Active Disturbance Rejection Control (ADRC) framework. An experimental verification is carried out on a Two Rotor Aerodynamical System (TRAS). The cross-couplings and highly nonlinear behavior of this complex system limit the performance of classical and linear controllers, widely-used in industrial practice. Additionally, the obtainment of a reliable mathematical model of TRAS is a challenging and tedious job. In ADRC, the unmeasured dynamics and external disturbance, associated with the 2-DOF motion of TRAS, are treated as an additional state variable. It is then estimated by a Nonlinear Extended State Observer (NESO) and compensated in the control signal. The obtained results showed the effectiveness of proposed ADRC-based controller in coping with dynamic and nonlinear variations, commonly seen in aerodynamical systems. It was noticed that due to disturbance estimation and its further cancellation a precise mathematical model of the plant is not required. The ADRC robustness was compared through experiment with other model-free method, i.e. the proportional-integral-derivative controller (PID). Better outcomes in terms of disturbance rejection have been reported for the ADRC approach.