Performance Evaluation of Two System Models for a MIMO System to Hover the Bicopter Unmanned Aerial Vehicle

Abstract

Building unmanned aerial vehicle (UAV) control system models are highly challenging due to multiple inputs and multiple outputs (MIMO). Not only does it have various angular position outputs such as roll, yaw, and pitch, but also flight control has more than one input; for instance, a bicopter has dual rotors. More rotors have more complex model. The hover condition has a zero resultant force which can be utilized to design a system model. On the other hand, an attractive identification system method is applied to develop the design. This research aims to evaluate the performance of two MIMO design on bicopter between methods based on the hover principle and identification technique. Experimental validation by employing bicopter simulator is an excellent strategy to fulfil this purpose. The results of the investigation of the experiment showed that the identification model was more accurate than the hover design, particularly regarding the overshot phenomenon and error. In addition, the hover principle tended to build ideal model because it did not include the dynamic, uncertainty and nonlinear conditions in aeroplane control design.  Although the identification system was complicated because it previously needed to measure the input and output values, it performed closer to the actual experiment. It performed more satisfactory overshoot values compared with the experimental validation than the hover model by 1°, 3°, and 8° in roll, pitch, and yaw angles, respectively.