A Review of Modeling and Control Techniques for Unmanned Aerial Vehicles

Abstract

Unmanned Aerial Vehicles (UAVs), particularly quadcopters, have found growing applications across diverse sectors such as surveillance, precision agriculture, and transport. However, their nonlinear dynamics, underactuated systems, and sensitivity to disturbances present persistent challenges in achieving robust and autonomous control. This review systematically examines advancements in UAV modeling and control techniques over the past five years. The study evaluates key modeling frameworks, Newton–Euler, Newton–Quaternion, and Geometry-Based Stochastic Models (GBSM), and analyzes a spectrum of control strategies, including observer-based, sliding mode, H-infinity, model predictive, and neural network-based controllers. Through a comparative assessment of their robustness, computational efficiency, and adaptability, the manuscript identifies critical limitations in handling uncertainties, scalability in UAV systems, and energy constraints. The findings highlight that hybrid control strategies incorporating adaptive mechanisms, learning-based algorithms, and quaternion-based modeling offer significant potential for enhancing autonomy and control. Therefore, this review provides a foundational roadmap for researchers and practitioners aiming to develop intelligent, efficient, and scalable UAV control systems capable of thriving in dynamic operational environments.