Structural Health Monitoring of UAV Wing Platform Using FBG Sensors, Wavelet Transform, and Computational Intelligence

Abstract

The growing demand and interest in the use of Unmanned Aerial Vehicles (UAVs) for various applications has highlighted the need to develop more robust structural monitoring systems, particularly for aircraft constructed with composite materials. Although these materials offer advantages in terms of lightness and strength, they are susceptible to delaminations and microcracks, which can compromise the safety and operational efficiency of UAVs. Early detection and characterization of these defects are key to preventive maintenance strategies and structural design optimization. This study presents a structural monitoring architecture based on Fiber Bragg Grating (FBG) sensors, combined with signal processing methods and computational intelligence, to evaluate the structural integrity of UAV wings. Luna Innovations’ os1200 and os3200 sensors were selected for their high sensitivity, immunity to electromagnetic interference, and multiplexing capability on a single fiber. The os1200 sensors are positioned in critical areas of the wing profile, allowing for the mapping of stress distribution in high-tension regions, while the os3200 sensors are placed in hard-to-reach areas where the use of metallic sensors is not feasible. This arrangement facilitates detailed data acquisition on the strain distribution across the UAV structure. For data analysis, a methodology based on signal processing and machine learning was employed. Filtering and conditioning techniques were applied to reduce noise, followed by Fourier and wavelet transforms, which enabled the identification of subtle changes in the structural response typically associated with the presence of faults. Additionally, artificial neural networks and machine learning algorithms were implemented for defect classification and severity assessment, leveraging patterns extracted from the sensor signals. Hybrid models combining wavelet transforms with supervised learning were explored, optimizing the detection and prediction of structural damage. The initial validation of the architecture was conducted in a controlled laboratory environment, using UAV wing profile prototypes subjected to static and dynamic load tests to induce different types of failures. FBG sensor measurements were correlated with visual inspections and non-destructive evaluation (NDE) techniques. The results obtained are expected to lay the groundwork for the development of a real-time structural monitoring system that enhances the safety and reliability of UAVs, reduces the risk of catastrophic failures, and provides a key tool for intelligent maintenance management.