Variations of seawater salinity often cause ocean internal waves, water masses and stratification, which affect the stability of the ocean environment. Therefore, the study of seawater salinity is significant for the prediction of changes in the ocean environment. However, existing methods for measuring seawater salinity generally have the disadvantages of low sensitivity and low accuracy. In this work, we proposed a seawater salinity sensor based on long period fiber grating (LPFG) in the dispersion turning point (DTP), which has demonstrated the possibility to fabricate LPFG with a shorter grating period by CO₂ laser in a thin single mode fiber (SMF) of 80 μm cladding diameter without etching. For obtaining higher sensitivity that could meet the measurement requirement in practice, the proposed sensor was optimized by combining etching cladding and DTP. After the LPFG working near DTP was fabricated by a CO₂ laser, the cladding diameter was reduced to 57.14 μm for making cladding mode LP_1,7 work near DTP by hydrofluoric acid (HF) solutions. The experimental results have demonstrated that a sensitivity of 0.571 nm/‰ can be achieved when the salinity increases from 5.001‰ to 39.996‰, and the sensor shows good repeatability and stability. Based on its excellent performance, the optimized LPFG is a prospective sensor to monitor seawater salinity in real time. Meanwhile, a low-cost way was provided to make LPFG work near DTP instead of ultraviolet exposure and femtosecond laser writing.

In this work, a seawater salinity sensor based on dual resonance peaks long period fiber grating (LPFG) and back propagation neural network (BPNN) has been proposed. A LPFG working near dispersion turning point has been theoretically designed and actually fabricated, which has a refractive index (RI) sensitivity of 1304 nm/RIU in RI range of 1.33156 to 1.39947. Then, the sensor was investigated under different salinity and temperature conditions, in which the salinity changed from 5 to 40‰ and the temperature changed from 0 to 30°C. The highest seawater salinity sensitivity of 0.2662nm/‰ was achieved when the temperature was 30 °C, while the highest average temperature sensitivity was 0.482 nm/ ° C. The BPNN was employed for temperature compensation and salinity prediction because of nonlinear response to the variations of seawater salinity and temperature, whose mean absolute error and maximum absolute error are 2.1774 and 6.8901, respectively. Because the prediction accuracy is poor, the genetic algorithm has been employed for BPNN optimization. Then, BPNN based genetic algorithm method achieved the mean absolute error of 0.10139 and maximum absolute error of 0.19016, whose prediction accuracy was increased by an order of magnitude. This novel sensor based on optimized BPNN could be applied in the field of seawater salinity measurement.

Industrial equipment, such as wind turbine foundations and oil and gas pipelines in cold regions, may undergo extrusion/expansion deformation during freezing and thawing of frozen soil, which affects their power response and safe operation. Measuring the internal deformation of frozen soil can immediately reflect the strain situation of industrial equipment to reduce the risk of equipment operation. We designed a 6-dimensional strain sensor (6-S Sensor) based on Fiber Bragg Gratings (FBG) to obtain the spatial principal strain distribution. The strain range, linearity, and average error of the sensor were -4000~8000 με, 0.997, and 2.94%, respectively. The sensor accurately measured the frozen and thermal expansion of frozen soil at different temperatures in the laboratory. The maximum frozen expansion was 6471.38 με, which occurred in the X direction. The accuracy of spatial principal strain monitoring for the sensor was evaluated through uniaxial compression. The stability of the sensor was verified by the monitoring experiment under natural temperature for half a month. This study provided a pioneering method for monitoring the internal spatial principal strain of frozen soil.

Hydraulic engineering built in the cold region, such as reservoirs and hydropower stations, is often threatened by static ice pressure from nature. Therefore, it is of vital significance to research the pressure variation in the growth and melting processes of the ice layer for the design and protection of hydraulic structures in cold regions. This paper introduces an optical fiber sensor system based on the fiber loop ring-down spectroscopy technology and field-programmable gate array (FPGA) pulse modulation technology. An electro-optic modulation scheme that relied on FPGA to generate optical pulses with adjustable pulse width and period is proposed, which is more suitable for the in-situ observation. In addition, the temperature stability and repeatability of the system are also discussed. This system was applied to the real-time detection of static ice pressure on the sidewall and bottom of the polyvinyl chloride (PVC) pipe during the ice growth and melting processes. The results indicate that the system has favorable stability and sensitivity, and the relationship obtained between the static ice pressure and temperature could provide some references for the field application in the future.

To perceive spatial principal strain in frozen environments, the 6-axis strain sensor was proposed in previous research. When monitoring ice instability, it is necessary to demodulate the cascaded 6 FBGs of the sensor over a wide wavelength range. The edge filtering demodulation method based on dual-peak resonance LPFG spectral control is proposed, which achieves accurate demodulation of the center wavelength of cascaded 6 FBGs over a large wavelength range (each demodulation range is 15 nm, and the total range is 90 nm). We have etched a dual-peak resonance LPFG using CO2 laser and designed a demodulation system based on piezoelectric ceramics (PZT). By precisely controlling the bending degree of LPFG through PZT, the spectrum of the LPFG can be regularly changed. The linear mathematical model of the demodulation system has been established, which theoretically verifies the feasibility of the proposed demodulation method. To solve the nonlinear error caused by the linearity poor at the edge of the demodulation interval, the BP-neural network demodulation model was established. The average error is smaller, which can reach 0.014 nm, with an MSE of 0.0005. This work can provide useful technical support for the widespread application of FBG multi-dimensional strain sensors.

为了研制高灵敏海水盐度传感器，本文基于CO₂激光技术成功制备出一种工作在色散转折点（DTP）附近的长周期光纤光栅（LPFG）。首先，利用CO₂激光器在80 μm细单模光纤上制备出工作在DTP附近的LPFG，证明了采用CO₂激光微加工技术制备较短周期LPFG的可能性。其次，通过调控CO₂激光器的制备周期，使高阶包层模式LP_1,9工作在DTP附近，从而显著提高了LPFG的折射率灵敏度。在双峰谐振增敏效应的作用下，当海水盐度从5.001 ‰变化到39.996 ‰时，光栅周期为115.4 μm的双峰谐振LPFG平均灵敏度高达0.279 nm/‰。研究结果表明，本文制备的LPFG海水盐度传感器具有谐振损耗大和灵敏度高的优点，其在海水盐度监测领域具有较好的应用前景。