Reduction of the Disturbance Effects in a Twin Rotor Multi-Input Multi-Output System Based on a Modified Smith Predictor Control Scheme

Three factors that complicate the position and orientation control of unmanned autonomous vehicles are the hardware-induced delays, delays caused by teleoperated modes of operation, and persistent disturbances such as wind. In this article, we study how to overcome these challenges by controlling a twin rotor multi-input, multi-output system. This nonlinear system has two degrees of freedom and significant cross-coupling, making its control particularly challenging. Time delay and external disturbances contribute to system instability and complicate flight control. The effects of time delay can be mitigated using the standard Smith predictor controller. However, this approach performs poorly when persistent disturbances are present. Our contribution is a modified Smith predictor that effectively mitigates disturbances. Additionally, a decoupler is employed to transform the system into two single-input, single-output models, enabling precise control of both the horizontal and vertical planes. The proposed method is implemented on a prototype and compared with the original Smith predictor-based control scheme. Experimental results validate the effectiveness of the proposed approach, showing that our approach reduces the steady-state tracking error to less than 2%, decreases settling time from 30 s to 20 s, and improves disturbance rejection under wind disturbance conditions by 35% compared to the standard Smith predictor. These results demonstrate the superior set-point tracking performance and enhanced disturbance rejection of the proposed approach.