Analysis of methods and algorithms for quadrotor position control

The object of research is the system of position control of a quadrotor unmanned aerial vehicle (UAV) as a nonlinear multi-input multi-output (MIMO) system with strong cross-channel coupling and high sensitivity to parametric and structural uncertainty. The problem addressed is the lack of robust and computationally efficient control algorithms that can ensure stability under uncertainty and be implemented on embedded platforms with limited resources. This study presents an analytical review of modern control methods for quadrotor position stabilization. The methods analyzed include classical proportional-integral-derivative (PID) controllers, linear optimal, robust, adaptive, and intelligent systems (neural networks, fuzzy logic). The analysis focuses on the structure, sensitivity to uncertainty, computational complexity, and feasibility of implementation on STM32-based flight controllers. As a result of the review, it was established that classical PID controllers, while widely used, are highly sensitive to model variations and sensor noise. Intelligent systems show better adaptability but exceed the computational capacity of low-cost microcontrollers. The most promising direction is identified as energy-based control methods that minimize local functionals of instantaneous energy values. These methods allow generating closed-form control laws, avoid signal differentiation, and maintain robustness with minimal processor load. The comparative evaluation shows that the proposed algorithm has the potential to improve control quality by more than 7% and reduce the impact of parametric disturbances by an average of 10% compared to traditional PID-based systems. The results are recommended for UAV control systems operating under limited computational capacity, absence of GPS, or in disturbed environments, such as tactical drones, FPV platforms, and autonomous navigation systems.