Filtern
Dokumenttyp
- Beitrag zu einem Tagungsband (15)
- Vortrag (14)
- Zeitschriftenartikel (12)
- Posterpräsentation (6)
Sprache
- Englisch (34)
- Deutsch (12)
- Mehrsprachig (1)
Schlagworte
Organisationseinheit der BAM
- 8 Zerstörungsfreie Prüfung (47) (entfernen)
Eingeladener Vortrag
- nein (14)
Faseroptische Sensorsysteme bieten heute die Möglichkeit des Online-Monitorings von sensiblen, für die Energieversorgung wichtigen Betriebsmitteln. Insbesondere Hochenergiekabel, wie diese sowohl im Offshore- als auch im Onshorebereich eingesetzt werden, sind prädestiniert für eine zustandsorientierte Überwachung. Gerade Hochspannungshochenergiekabel sind Betriebsmittel, die im Falle eines Ausfalls durch Schäden mit hohen Reparatur- und Folgekosten verbunden sind. Mit dem Einsatz von faseroptischen Sensoren sind mechanische Einflüsse wie Vibrationen z. B. verursacht durch Ankerfall, Temperaturhotspots oder auch Teilentladungsaktivitäten an dezidierten vulnerablen Stellen wie Kabelendverschlüssen oder –muffen über längere Kabelstrecken detektierbar. Der Einsatz von im Rahmen des Verbundprojekts Monalisa entwickelter faseroptischer Diagnosetechnik in Verbindung mit faseroptischen Sensoren wird hier aufgezeigt.
Die Verwendung von Methoden des Maschinellen Lernens (ML) und der Künstlichen Intelligenz (KI) im Fachbereich 8.6 Faseroptische Sensorik wird dargestellt. Die vielfältigen Möglichkeiten, Machine Learning auf Basis Künstlicher Neuronaler Netze (ANN) für eine schnelle und effiziente Datenverarbeitung eizusetzen werden demonstriert. Hierfür werden Beispiele für die Anwendungszwecke Messgrößenberechnung, Entrauschen, Interpolation, Bildverarbeitung und Messdatenauswertung aufgezeigt.
We present a preliminary investigation on distributed humidity monitoring during the drying process of concrete based on an embedded polymer optical fiber (POF). The water dissipated into the POF changes several properties of the fiber such as refractive index, scattering coefficient and attenuation factor, which eventually alters the Rayleigh backscattered light. The optical time Domain reflectometer (OTDR) technique is performed to acquire the backscattered signal at the wavelengths 650 nm and 500 nm, respectively. Experimental results show that the received signal increases at 650 nm while the fiber attenuation factor clearly increases at 500 nm, as the concrete dries out. In the hygroscopic range, the information retrieved from the signal change at 650 nm agrees well with the measurement result of the electrical humidity sensors also embedded in the concrete sample.
The feasibility of traffic monitoring along a major urban road using Distributed Acoustic Sensing (DAS) is demonstrated. We present measurement results of conventional intensity-based DAS along a dark fiber in a tubed fiber optic cable buried roadside. With only minimal postprocessing different classes of traffic participants can be distinguished (cars, bikes, pedestrians). Different methods for pre-processing raw data are compared, especially with regard to providing suitable inputs for pattern recognition algorithms. Furthermore, the various challenges for automatized vehicle detection and classification related to varying sensitivity and inhomogeneous signal propagation are discussed. Sensitivity fluctuations and variations are in part inherent to the measurement technology and originate in part in ground conditions. The measurement quality and usefulness for traffic monitoring of intensity-based DAS is compared to that of state-of-the-art phase-resolved DAS, allowing quantitative evaluation of vibration signals.
The feasibility of traffic monitoring along a major urban road using Distributed Acoustic Sensing (DAS) is demonstrated. We present measurement results of conventional intensity-based DAS along a dark fiber in a tubed fiber optic cable buried roadside. With only minimal postprocessing different classes of traffic participants can be distinguished (cars, bikes, pedestrians). Different methods for pre-processing raw data are compared, especially with regard to providing suitable inputs for pattern recognition algorithms. Furthermore, the various challenges for automatized vehicle detection and classification related to varying sensitivity and inhomogeneous signal propagation are discussed. Sensitivity fluctuations and variations are in part inherent to the measurement technology and originate in part in ground conditions. The measurement quality and usefulness for traffic monitoring of intensity-based DAS is compared to that of state-of-the-art phase-resolved DAS, allowing quantitative evaluation of vibration signals.
Division 8.6's competencies and work focus areas are shortly outlined and investigations related to pipeline and submarine power cable monitoring using different distributed fiber optic sensing techniques are presented with some results. Furthermore, BAM 8.6 expertise in the field of sensor application, embedding and integration is shown.
Fiber optic Distributed Acoustic Sensing (DAS) is an emerging method for many different monitoring purposes, enabling a spatially and temporally resolved collection of acoustic and vibration information over many kilometers. DAS, thus being a "dynamic" sensing technique, allows for online condition monitoring and the detection and localization of threats or hazards in real time via characteristic acoustic/vibration states and their changes or via occurring anomalous signals, respectively.
At BAM, we have employed this technology for a number of different applications of monitoring of large infrastructures, e.g., bridges, pipelines, submarine power cables or railway tracks. Currently, we are investigating the use of DAS for further innovative uses, which aim at facilitating the energy transition, enabling "smart" infrastructure and providing the basis for comprehensive hazard monitoring and warning systems, respectively. Our research fields include fiber optic borehole monitoring in the context of hydrogen storage caverns, traffic and road monitoring, using DAS for earthquake monitoring in urban areas for disaster management and long-term monitoring of large-scale subsidence caused by mining activities. Another new topic is structural health monitoring in concrete structures in the framework of the FSP Security.
In this talk, we will first briefly present the basics and capabilities of DAS. Then, we will portray our previous and current works related to this technology and show some attractive results. Finally, we will discuss our upcoming projects on exciting new applications of DAS.
Time-Efficient Convolutional Neural Network-Assisted Brillouin Optical Frequency Domain Analysis
(2021)
To our knowledge, this is the first report on a machine-learning-assisted Brillouin optical frequency domain analysis (BOFDA) for time-efficient temperature measurements. We propose a convolutional neural network (CNN)-based signal post-processing method that, compared to the conventional Lorentzian curve fitting approach, facilitates temperature extraction. Due to its robustness against noise, it can enhance the performance of the system. The CNN-assisted BOFDA is expected to shorten the measurement time by more than nine times and open the way for applications, where faster monitoring is essential.
In the last few years, the use of machine learning has emerged in the field of distributed fiber optic sensors as a promising approach to enhance their performance and provide new capabilities. In this study, we use machine learning for simultaneous measurements of temperature and humidity in polyimide (PI)-coated optical fibers based on Brillouin Brillouin optical frequency domain analysis (BOFDA). Different non-linear machine learning algorithms are employed, namely polynomial regression, decision trees and artificial neural networks (ANNs), and their discrimination performance is benchmarked against that of the conventional linear regression. The performance is evaluated using leave-one-out cross-validation to ensure that the models are reliable and able to generalize well on new data. We show that nonlinear machine learning algorithms outperform the conventional linear regression and thus could pave the way towards simultaneous cost-effective temperature and humidity distributed sensing, which has the potential to find attractive new applications in the field of civil and geotechnical engineering, from structural health monitoring of dikes and bridges to subsea cables and long pipelines corrosion detection.