Ensuring stability at the stabilization stage of unmanned aircraft with a neuronetwork regulator

DOI: 10.31673/2518-7678.2023.020404

Abstract

This article addresses the complex problems associated with controlling multi-rotor unmanned aerial vehicles (UAVs) amidst external disturbances and internal system nonlinearities. Conventional PID (proportional-integral-derivative) controllers often fail to ensure the stability of these vehicles in real-world settings. By using artificial neural networks to dynamically tune the PID controller coefficients, this study presents a solution to mitigate the limitations of traditional methods while ensuring system stability under a variety of environmental conditions.
In addition, the paper emphasizes the distinct advantages of UAVs compared to manned aircraft. Their cost-effectiveness, stemming from the lack of requirements for pilot training and life support systems, positions UAVs as a highly effective tool for a variety of missions. This emphasis on cost-effectiveness demonstrates the potential of UAVs to outperform manned counterparts in a variety of tasks, thereby emphasizing the importance of promoting and improving this technology for widespread use in various industries. Ultimately, this study sheds light on the critical role that artificial neural networks play in enhancing the capabilities of PID controllers, and emphasizes the transformative impact of UAVs in the modern work landscape.

Keywords: unmanned aerial vehicles, artificial neural networks, nonlinear systems, system adaptation, system stability, stabilization stage.