Resumen de Contribuciones al desarrollo de sensores distribuidos basados en la dispersión estimulada de Brillouin
- español
El objetivo principal de esta tesis es contribuir al desarrollo y la mejora del rendimiento de los sensores distribuidos basados en la dispersión Brillouin. Durante el desarrollo de este trabajo se han considerado diferentes áreas de mejora. En primer lugar, se han propuesto diversas configuraciones experimentales para superar algunas de las limitaciones típicas que tienen estos sensores, como son los efectos no locales en los sensores BOTDA o la aparición de sub-picos en el espectro de ganancia de Brillouin en sistemas basados en el dominio de frecuencia. Otro objetivo principal de este trabajo es aplicar diferentes enfoques de procesado para resolver problemáticas aún no resueltas, como la discriminación entre las medidas de temperatura y las de deformación obtenidas con los sensores Brillouin. Además, también se han estudiado algunos métodos alternativos al método tradicional basado en la aplicación de ajustes Lorentzianos para estimar el cambio de la frecuencia Brillouin. Finalmente, este trabajo también ha tratado de contribuir a la validación de los conocimientos adquiridos mediante la validación en escenarios real
- English
Nowadays, society demands more and more services in a wide range of different sectors: transport, energy, health or communications, among others. To have all these services working property, many infrastructures are required. Roads, bridges, railways, tunnels, dams, industrial factories or pipelines are examples of these indispensable infrastructures to satisfy our needs. However, they are not perfect and exhibit deterioration with time, with the appearance of cracks, corrosion processes or other problems that can give rise to their collapse. Accordingly, monitoring the integrity of structures has become a task that concerns society due to the potential risks that they imply on the population. In this way, there is a growing number of monitored structures using all kinds of technological approaches.
In this regard, the development of fiber optic sensors has enabled a wide range of solutions to be employed in structural health monitoring due to their unique properties. Optical fiber sensors present some advantages such as long distance monitoring, electromagnetic immunity, the possibility of multiplexing more than one sensor in a fiber, the capability of being embedded on structures or not requiring electricity near the measurement point. They are based on different physical phenomena, such as diffraction, interferometry, polarization or scattering, among others. Based on the latter effect, there is a kind of optical fiber sensor that has very interesting properties to be employed with monitoring purposes, especially when the structure is large. They are known as distributed fiber sensors and are based on different scattering phenomena: Rayleigh, Brillouin or Raman. They enable to perform distributed measurements along large optical fibers with a defined spatial resolution, obtaining thousands of measurement points. Moreover, these sensors present the advantage that the optical fiber itself is the sensing element, not requiring any modifications to measure temperature and strain variations surrounding the optical fiber. Specifically, these sensors based on stimulated Brillouin scattering typically employ 1 m spatial resolution measuring along tenths of kilometers long fibers, although it is possible to increase their performance by applying some proposed improvements, reaching more than 100 km or sub-meter spatial resolutions, among others.
This thesis aims to contribute to the development of distributed sensors based on Brillouin scattering. With this purpose, first of all, a theoretical study of Brillouin scattering is carried out, focusing on the employment of this effect with sensing objectives. Especially, a deep understanding of Brillouin optical time domain sensors is considered, given that most of the proposed contributions are based on their employment. Once the main limitations of these systems are identified, such as non-local effects, high measurement time, limited spatial resolution or pump depletion, the proposed contributions of the thesis are explained.
The contributions are divided into three main groups depending on the proposed approaches: new experimental configurations, processing techniques applied to BOTDA sensors and laboratory demonstrations of field applications. In this way, proposals aiming to overcome non-local effects in BOTDA, to increase the performance of Brillouin distributed sensors based on the frequency domain and to use other kinds of laser sources in BOTDA are included in the first group. The second group contains proposals based on using artificial neural networks to discriminate temperature and strain measurements and the use of subpixel algorithms to improve the BOTDA performance when estimating the Brillouin frequency shift. Finally, the laboratory demonstrations are focused on providing solutions for high-temperature environments and water leakage detection, and also to analyse the modal shapes of a composite panel to allow the location of defects.
In summary, the solutions presented in this thesis seek to contribute to the field of Brillouin distributed sensors through the proposal of new approaches. They try to overcome some limitations detected in the literature such as non-local effects, the detrimental effect of sub-peaks in BOFDA sensors, the temperature and strain discrimination or their employment in harsh environments.
