Aprimoramento de detecção acústica distribuída com os algoritmos de amostragem ilimitada e de diversidade de fase

Abstract

Distributed Acoustic Sensing (DAS) in optical fiber is an innovative technology that uses the phenomenon of light scattering to measure vibration over tens of kilometers. This capability transforms the optical fiber into a long, continuous sensor, allowing for a wide range of applications, from structural integrity monitoring to real-time analysis of flow behavior and conditions in pipelines. However, heterodyne DAS systems have limitations due to signal distortions under specific conditions such as backscattered light attenuation and phase unwrapping errors, significantly hindering data interpretation. This work presents signal processing techniques, based on Unlimited Sampling algorithm and Phase Diversity, with a special focus on optimizing algorithm parameters to enhance signal reconstruction. The detailed methodology of the research includes the experimental setup, emphasizing the use of a standard singlemode optical fiber and the use of a piezoelectric system to excite the fiber at different frequencies and strains. The experimental arrangement allows tests with good repeatability, aiding in the validation of the proposed algorithms. The algorithms composed of Phase Diversity and Unlimited Sampling unwrapping technique are compared with conventional methods like Itoh’s unwrapping, demonstrating a 20% improvement in accuracy and reliability. These methods prove to be promising in overcoming traditional limitations, especially in scenarios with high phase discontinuities, being crucial to mitigate the effects of distortions imposed on the signals. The results highlight the importance of this research in significantly improving the accuracy of heterodyne DAS systems, offering a solid foundation on which subsequent works can be built, aiming for further enhancements and broader applications in the field.