Falkowski Krzysztof, Duda Michał
Avionics Department, Institute of Aviation Technology, Faculty of Mechatronics, Armament and Aerospace, Military University of Technology, 00-908 Warsaw, Poland.
Doctoral School, Military University of Technology, 00-908 Warsaw, Poland.
Sensors (Basel). 2023 Nov 24;23(23):9378. doi: 10.3390/s23239378.
This article presents the method of identifying dynamic models for different flight states of a rotary-wing UAV for simulations. Experimental flights with real-life UAVs were conducted to obtain data necessary for identification. Dynamic models were identified with time series methods performed using Matlab R2022b software. Such models can later be implemented in simulations to represent the behavior of real-life objects. Simulation is the first stage of developing a real-life UAV system, where prototyping with physical models is problematic. Therefore, obtaining accurate models is crucial for the simulation process to be reliable. Presented methods do not require knowledge of UAV construction, and complex mathematical equations do not need to be derived. Also, verification of obtained models was performed to make sure that they were identified correctly. In particular, the presented method was proven effective and successfully used in some applications.
本文介绍了用于模拟的旋翼无人机不同飞行状态动态模型的识别方法。使用实际无人机进行了实验飞行，以获取识别所需的数据。使用Matlab R2022b软件通过时间序列方法识别动态模型。此类模型随后可在模拟中实现，以代表实际物体的行为。模拟是开发实际无人机系统的第一阶段，而使用物理模型进行原型制作存在问题。因此，获得准确的模型对于使模拟过程可靠至关重要。所提出的方法不需要了解无人机结构，也无需推导复杂的数学方程。此外，还对获得的模型进行了验证，以确保它们被正确识别。特别是，所提出的方法已被证明是有效的，并在一些应用中成功使用。
