Longitudinal UAV Motion with Plugged Error Model of Angular Velocity Sensor

Abstract

The Department of Avionics, Faculty of Aeronautics, Technical University of Košice uses a “Carbon Cab” equipped with a PixHawk PX-4 autopilot for its research in the field of UAV control. This autopilot allows a wide range of adjustments to the steering and the control loops themselves contained in the steering circuit. It is also possible to load the autopilot into its own control circuit and thus test its influence on the controlled quantities and the stability and controllability of the aircraft, which is the control object. The authors of the article discuss the simulation verification of the created model of a triaxial angular velocity sensor. The verification of the sensor model is performed by connecting it to the feedback loop of the mathematical model of the aircraft for longitudinal movement. The authors of the article discuss the changes that occur by connecting the sensor to the connection and their other possible solutions. Based on this connection, it is possible to analyze the interfering effect of raw data coming from the sensor and then create a filter that adjusts the data coming from the sensor to a suitable form, taking into account the increase in stability and controllability of the aircraft. The Allan variance method was used to analyze the sensor noise processes that had to be performed to create the sensor error model.