The work address the relevant topic of improving the accuracy of determining the angular orientation of autonomous aircrafts using a dynamic motion model. The reserve of accuracy used is the application of a dynamic rotation model, that takes into account a priori information about control signals, generated by the on-board autopilot in the mode of parrying external influences. To analyze the effectiveness of the obtained algorithm, numerical modeling of estimating the angular orientation process of a multirotor unmanned aerial vehicle was carried out. The results obtained demonstrate an increase in the accuracy of estimating the angular orientation of an unmanned aerial vehicle by an average of 7–15% compared to a similar algorithm, that doesn’t take into account autopilot commands. The work is of the greatest practical interest in the development of small autonomous aircraft, the weight and size characteristics of which do not imply the installation of precision sensors and on-board computers with high computing power.