Fiber-Optic Sensor Spectrum Noise Reduction Based on a Generative Adversarial Network.

Fiber-optic sensors have experienced great interest in research and development since their invention. Dedicated measurement systems are essential for utilization and development of these sensors. Instruments based on tunable lasers are established devices for the demodulation of such sensors. We introduce and demonstrate a tunable laser measurement system based on a widely tunable monolithic laser diode (TML). It is capable of demodulating different types of fiber-optic sensors, although we focus on the interrogation of fiber Bragg grating (FBG) sensors. The rapid tuning of the laser's wavelength is critical for achieving high measurement rates. However, the high tuning rate is demanding and requires careful characterization. We present a method for this and show, to our knowledge for the first time, data on the rapid tuning through the whole spectrum of this type of laser diode. We propose a modified centroid algorithm that can cope with nonequidistantly sampled spectra caused by the laser's functional principle. Finally, we demonstrate the dynamic performance of the new measurement system in an application: An FBG-based acceleration sensor is demodulated at a 5-kHz sample rate during a vibrational test showing the potential of the approach.

Due to the advantages of small size, high sensitivity and strong anti-interference ability, optical fiber sensors are widely used in gas detection. Meanwhile, concentration and humidity detection based on optical fiber sensing are paramount technologies in the field of gas sensing. However, how to solve the issue of cross sensitivity of concentration and humidity of optical fiber gas sensors and realize synchronous detection of two parameters are current common challenges. In this study, taking carbon dioxide as an example, a detection method based on double fiber Bragg grating (FBG) string is proposed to measure carbon dioxide concentration and air humidity simultaneously. Here, a pair of FBGs are welded in series on an optical fiber after being filled with special materials. Using the control variable method, the performance parameters of the sensor can be determined through calculation. The experimental results demonstrate that the gas concentration sensitivity is 0.5952 pm/ppm in the range of 0 ppm~2250 ppm. In the range of 25%~65%, the air humidity sensitivity is 28.62 pm/RH. Moreover, it is less affected by temperature and has satisfactory repeatability and reproducibility. The sensor realizes high sensitivity and low crosstalk measurement of concentration and humidity and can be applied to the simultaneous detection of carbon dioxide concentration and air humidity in specific confined space.

In this paper, a novel arc-shape FBG (Fiber Bragg Grating) load sensor is developed with a pair of FBGs embedded symmetrically in an elastic bending beam. The FBG pair is produced with peak wavelength separation of about 0.4 nm. The FBG with shorter wavelength is located above the neutral layer and another with longer wavelength below the neutral layer. As a result, the two peak wavelengths of the FBG pair move apart with lateral force applied but move together in the same direction when surrounding temperature varies, both in a linear manner. A very low temperature-dependency of only 0.04 pm/°C across the temperature range from 35°C to 82°C has been achieved, which makes it suitable for dynamic lateral force measurement even when the environmental temperature varies. Due to the fact that the sensor module consists of only the grating fiber with two FBGs and some elastic material, which are purely passive components, hence, the temperature effect of this sensor can be self-compensated without a need of any active temperature controller or other temperature-isolation mechanism, thus in principle, this sensor module is passively temperature-independent.

This thesis presents two designs for high power linearly polarized all-fiber linear cavity lasers, continuous wave (CW) and mode-locked. The cavity designs use Polarization Maintaining (PM) fibers for both gain medium and Fiber Bragg Gratings (FBGs). The FBG pairs select lasing wavelength and polarization. The fiber lasers incorporating specialty designed FBGs achieve an extinction ratio larger than 23 dB. Firstly, an all-fiber linear cavity design of a high power picoseconds mode-locked laser is introduced. The proposed configuration is based on Non-Linear Polarization (NPR) using PM Yb-doped active fiber and two matching FBGs to form the laser cavity. The combination of nonlinearity, gain and birefringence in cavity made the laser generate mode-locked pulses in picoseconds range and with high average output power. The output mode-locked pulses amplitude is modulated with an envelope whose mechanism is also investigated in this thesis project. Experimental data and numerical simulations of the self mode-locking fiber laser are presented. Main parameters affecting mode-locked pulses and its envelope are identified. In addition, a new theoretical model based on Nonlinear Schrödinger Equation (NLSE) is developed and implemented on the MATLAB platform. The model explains the self mode-locking mechanism and the source of the pulse envelope. In this model, it is proven that self phase modulation (SPM) plays an essential role in pulse formation and shaping. The theoretical model and experimental results are in a very good agreement at different pumping levels. A method of regulating the mode-locked pulses is presented. This is achieved by applying a pulsed current to pump diode. This method successfully stabilizes the mode-locked pulses underneath a Q-switched pulse envelope. Further scale-up of average power and pulse energy is realized by adding an amplifier stage. Secondly, a CW dual-wavelength all-fiber laser is presented. The laser consists of two pairs of FBGs and a PM Er/Yb co-doped fiber as a gain medium. The laser emits at both 1 μm and 1.5 μm wavelengths simultaneously with a stable output. This laser provides a compact fiber-based pumping source that is suitable for mid-IR generation.

The purpose of this paper is to review the application of various fiber-optic and optical sensor technologies in structural health monitoring (SHM) for detecting and measuring mechanical strains and stresses. As is known, fiber optic sensors have low operating costs and small dimensions compared with traditional sensors based on electrical and thermal effects, therefore, they are gradually becoming one of the most promising areas of development of sensor technologies for monitoring structural systems. This paper presents the basic operating principles of several widely used fiber optic sensor types (e.g., based on the Fabry-Perot interferometer, Bragg diffraction, reflectometry, etc.), and describes the experience of using fiber optic sensors in monitoring various physical parameters and demonstrates the possibilities of using fiber optic sensors in structural systems such as bridges, overpasses, tunnels, buildings, etc. It is shown that two types of sensors have found the greatest application for measuring and detecting mechanical deformations in building structures: sensors based on fiber Bragg gratings, as well as optical reflectometry methods (primarily Brillouin reflectometry). Thus, the specified types of sensors are the most promising and effective in the context of SHM technologies.

In this paper, we propose a simple fiber-optic sensing system to measure the concentration of the methane based on a special catalyst and fiber Bragg gratings (FBGs). The temperature around the catalyst rises rapidly as the concentration of methane increases, and hence can be detected by an FBG-type sensor through monitoring the wavelength shift of the reflection spectrum. By employing a well-matched FBG as a reference, the wavelength shift of the sensing FBG can be transformed effectively into the variation in the measured optical intensity. An experiment is demonstrated to show the good performances of our system, such as the high resolution and good stability to the environmental temperature.

In this paper, we fabricate a pair of fiber Bragg gratings (FBGs) in large mode area double cladding fiber (LMA-DCF, core/inner cladding diameter: 20/400 μm) by visible (515 nm) femtosecond laser and phase mask for the first time to the best of our knowledge. Then this pair of FBGs are utilized to construct a high power fiber oscillator operated near 1080 nm. The maximum output laser power is more than 3.2 kW with a slope efficiency of ∼77.9%, and the beam quality factor M <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> is about 1.28. The FBGs are simply fixed on water-cooling plate without any special package, and the temperature of the FBGs is about 40 °C at the maximum output power. Our research work confirms the reliability of visible femtosecond laser in fabrication of FBGs intended to be applied in high power fiber oscillators.

We demonstrate a dual-wavelength linearly polarized all-fiber laser emitting 1- μm and 1.5- μm wavelengths with a total output power of 1.3 W. The laser consists of two pairs of fiber Bragg gratings (FBGs) fabricated on PM fibers to form a cavity and an Er/Yb co-doped fiber as a gain medium. A high extinction ratio of 23 dB is achieved using specially designed PM FBGs. The laser output has a narrow bandwidth of 26 pm at both 1057 and 1554 nm. When the dual-wavelength all-fiber laser is used to pump a PPLN crystal, the generation of mid-infrared emission at 3.3 μm is successfully demonstrated.

A novel kind of fiber optic ultrasonic sensor based on matching fiber Bragg gratings (FBGs) is proposed and demonstrated. The sensors consist of a pair of matching FBGs fixed to a special bracket. The bracket plays a role in stretching and squeezing the FBGs, with the push–pull effect efficiently coupling the ultrasonic signal to the sensor, thus, improving the sensor’s sensitivity. Side-band filtering technology-based intensity interrogation was used to detect ultrasounds in water. With the synergic effect of the matching FBGs, the sensor performed with a high signal-to-noise ratio (56.9 dB at 300 KHz, 53 dB at 1 MHz and 31.8 dB at 5 MHz) and the observed ultrasonic sinusoidal signals were undistorted and distinguishable in the time domain.

: This report describes the work performed with Production Products Manufacturing & Sales (PPMS), Inc., under the Liquid Molded Composite Armor Smart Structures Using Embedded Sensors Small Business Innovative Research (SBlR) Program sponsored by the U.S. Army Research Laboratory (ARL). In the Phase I effort, fiber optic sensor systems were investigated for in-process cure monitoring and in-service health and dynamic response monitoring of monocoque and hybrid liquid molded composite armor structural parts. Sensor embedding techniques during resin infusion molding were developed, thick panels with varying residual stress characteristics were fabricated, two fiber optic sensor types were utilized and compared, several demodulation techniques were studied, and information processing programs were written for converting the sensor signal to engineering data. Bragg grating and fluorescence optrode fiber optic sensors were selected and embedded in armor panels to monitor the cure of the systems and health of the panels during impact and four-point bend tests. During panel curing, the Bragg gratings detected the change in strain in the material at the onset of cross-linking as well as the strain changes (residual stress) resulting from panel processing. The Bragg gratings, serving a dual purpose, measured the resulting strain from the external loads applied due to impact and bending. The fluorescence optrode was used during the cure process only to measure the changes in the material as it cured (degree of cure). Stitching of thick-section fiberglass preforms was also investigated.

In the last decades, Fiber Optic (FO) sensor has gained increasing acceptance. Among the different FO sensor types, Fiber Bragg Grating is most widely used due to its commercial availability and the unique multiplexing potential. The latter feature enables the development of large sensor array and/or sensor network for multi-parameter sensing to measure temperature, strain, pressure, vibration, flow, ... . Other advantages of FO sensors are intrinsically safe, unsusceptible to EMI, potentially lightweight, small dimensions and can be embedded and integrated in composite material. The main application of FBG sensor is Structural Health Monitoring (SHM) of civil engineering constructions. FBG sensor systems for other applications such as medical, oil & gas, public transport, aerospace, defense and nuclear fusion are also presented or under development. For special applications, corresponding interrogation systems with dedicated features (detection bandwidth >;MHz, wavelength noise <;fm/√(Hz), ... ) are developed.

Deep learning models have been used to improve seismic trace interpolation in recent years. Encode-Decode models, such as U-Net, have been implemented to solve interpolation problems. The success of U-Net in seismic interpolation has inspired us to test the U-net also as an image generator for further Generative Adversarial Network (GAN) interpolation models. The objective of the present paper is to compare the performance of U-Net inside the GAN models for seismic interpolation: the U-Net alone and two GAN models, the conditional GAN (cGAN) and the Cycle Consistent GAN (CycleGAN), both using the U-Net as a generator inside their workflow. We test the methodologies for two scenarios: regular and irregular interpolation. All tests were performed in a real dataset from the Tupi Field which belongs to the Brazilian pre-salt region. A comparison of the statistical metrics shows that cGAN performs better than CycleGAN and the U-Net alone in most cases. The computational training time of the cGAN model, for all interpolation scenarios, is better than the CycleGAN. Finally, the cGAN training time is comparable to the training of the U-Net alone.

We propose and experimentally demonstrate the feasibility of an integrated hybrid optical fiber sensing interrogation technique that efficiently combines distributed Raman-based temperature sensing with fiber Bragg grating (FBG)-based dynamic strain measurements. The proposed sensing system is highly integrated, making use of a common optical source/receiver block and exploiting the advantages of both (distributed and point) sensing technologies simultaneously. A multimode fiber is used for distributed temperature sensing, and a pair of FBGs in each discrete sensing point, partially overlapped in the spectral domain, allows for temperature-independent discrete strain measurements. Experimental results report a dynamic strain resolution of 7.8 nε/√Hz within a full range of 1700 με and a distributed temperature resolution of 1°C at 20 km distance with 2.7 m spatial resolution.

High sensitivity optical fiber pressure sensor based on thin-walled oval cylinder

We propose and demonstrate a relative humidity (RH) sensor based on a polarimetric fiber laser and beat frequency (BF) demodulation. The laser cavity of the proposed fiber laser sensor consists of a pair of fiber Bragg gratings (FBGs) and a high-birefringence erbium-doped fiber (Hi-Bi EDF) coated by a thin polyvinyl alcohol (PVA) film layer. This thin-film layer responds to humidity changes in the environment by swelling or contracting, inducing stress effects on the EDF. The stress effect, in turn, results in changes to the polarization mode beat frequency (PMBF) detected by the photodetector (PD) in the sensing system. The inherent high-birefringence property of the Hi-Bi EDF significantly enhances the sensitivity of the PMBF to humidity-induced stress. Humidity sensing experiments are conducted utilizing the proposed sensor. Experiment results show that the sensor has a high average humidity sensitivity of 13.9 MHz/%RH within the humidity range of 20.4%RH–36.0%RH. A stability test is also conducted, demonstrating that the proposed sensor achieves an accuracy of 0.65%RH.