In-situ piezoelectric sensors for structural health monitoring with machine learning integration

This paper presents a novel approach to structural health monitoring (SHM) in aeronautical composite materials, leveraging embedded sensor data and advanced machine learning techniques for enhanced performance and simplified fault detection and identification. The study introduces an in-situ sensing system that integrates polymer-based piezoelectric sensors within the composite structure, enabling direct measurement and high-quality data acquisition. By employing a Gram angle field-based time-frequency transformation, the proposed method captures fault information from the in-situ measurements effectively. The study validates the effectiveness of the proposed approach by successfully completing diagnostic validation and identification of single and compound faults, such as scratches, holes, cuts, and other defects, using simple machine learning models. The findings of this study highlight the potential of combining in-situ sensing and advanced machine learning techniques for improved structural health monitoring in aeronautical composite materials. • Proposed in-situ SHM sensors fully embedded in composite laminate layups. • Reviewed direct and indirect sensing for damage assessment and ML performance. • End-to-end fault data capture using Gram angle field-based time-frequency transformation. • Complete diagnostic validation & identification of single/compound damage in composites.