STM32-Based Embedded Control of a Quadrotor UAV Using an Adaptive Neuro-Fuzzy Approach
Year: 2026
Type: conference-paper
Abstract: This paper presents the design and embedded implementation of an adaptive neuro-fuzzy control framework for a quadrotor Unmanned Aerial Vehicle (UAV) operating under nonlinear dynamics and parametric uncertainties. The proposed control architecture employs six Adaptive Neuro-Fuzzy Inference System (ANFIS)-based controllers with online parameter adaptation driven by a hybrid gradient descent and Extended Kalman Filter (EKF) learning mechanism, enabling stable flight and accurate trajectory tracking. The controller is deployed on a lowcost STM32 Nucleo embedded platform using a Processor-In-the-Loop (PIL) configuration, and evaluated under both floating-point and fixed-point arithmetic implementations. Experimental results demonstrate that the fixed-point realization achieves efficient CPU utilization while maintaining high numerical accuracy. Furthermore, the proposed approach exhibits robust tracking performance and superior disturbance rejection compared to conventional proportional-integral-derivative (PID) control, confirming its suitability for real-time embedded UAV applications.
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7203625503