Weihao Lin

This paper proposes and experimentally demonstrates a fiber ring laser (FRL) salinity sensing system based on a Sagnac loop (SL) formed by a tapered polarization-maintaining fiber (TPMF). The operating principle is that salinity modulates the birefringence of the polarization-maintaining fiber (PMF), causing a shift in the interference wavelength of the SL transmission spectrum, while the FRL narrows the optical spectrum and enhances the signal-to-noise ratio (SNR). In the experiment, the SL consists of a 20-cm-long PMF with a tapered waist diameter of 10.86 μm. Over the salinity range of 0‰ to 30‰, the sensitivity of the laser-based sensing system is 97 pm/‰, which agrees well with the 93 pm/‰ sensitivity obtained using a broadband light source (BBS), and the salinity exhibits a good linear relationship with the wavelength shift, with a coefficient of determination (R²) of 0.997. Meanwhile, the ring laser cavity improves the SNR of the sensing system from 22 dB to approximately 54 dB, and compresses the 3-dB bandwidth from 1.75 nm to 0.06 nm. Further adopting the support vector regression (SVR) algorithm for linear regression modeling of the spectral data, the results show that the mean absolute error (MAE) decreases from 0.50‰ to 0.04‰, the root mean square error (RMSE) decreases from 0.54‰ to 0.11‰, and R² reaches as high as 0.99988. To the best of our knowledge, this is the first work that combines salinity laser sensing with an artificial intelligence algorithm. The proposed sensor leverages the narrow linewidth and high SNR advantages of the FRL together with the high-precision linear fitting capability of the SVR algorithm, achieving significantly improved accuracy for salinity measurement compared to conventional spectral demodulation.

This work presents a dual-point temperature sensing system capable of high-speed based on the photonic time-stretch dispersive Fourier transformation (TS-DFT) technique. The system aims to overcome the speed limitations of traditional optical spectrum interrogators. A sensing array consisting of two cascaded fiber Bragg gratings (FBGs) with a 5-meter separation is illuminated by a 50 MHz femtosecond laser source. The reflected spectrum, encoded with temperature information, is then stretched in time by propagating through a dispersion-compensating fiber (DCF) with a total dispersion of -360 ps/nm. This process enables the direct mapping of the optical spectrum into a measurable time-domain waveform at a 50 MHz repetition rate. Experimental results demonstrate the system’s capability to perform measurements over a broad temperature range from 40 °C to 250 °C. The temperature sensitivities of both FBGs were characterized to be 0.012 nm/°C in the optical wavelength domain. Correspondingly, the obtained time-domain sensitivities are -4.28 ps/°C and -4.31 ps/°C, with a temperature resolution of approximately 5.8 °C in the time domain. This study validates the feasibility of the TS-DFT scheme for high-speed dual-point sensing in demanding applications.

Temperature sensors are critical across diverse applications, yet their performance is inherently limited by measurement errors. Here, we proposed a hybrid-intelligent sensing paradigm combining a Sagnac interferometer (SI)-based fiber ring laser (FRL) with Backpropagation Neural Network (BPNN) optimization. The sensor, constructed with 30 cm polarization-maintaining fiber (PMF) with a high birefringence coefficient, achieves -0.548 nm/°C sensitivity at a range of 15 °C to 25 °C and maintains the signal-to-noise ratio (SNR) at 47.5 dB with a linewidth of 0.05 nm. Experimental results demonstrate a mean squared error (MSE) reduction from 0.01034 °C² in broadband light source (BBS) to 0.00691 °C² via FRL integration. To address residual errors impacting ultra-precision applications, BPNN-based spectral analysis further reduces MSE to 6.396×10^-6 °C² —a three-order-of-magnitude improvement over conventional methods. To the best of our knowledge, this work represents the first implementation of neural network-assisted error analysis in FRL system, demonstrating a hybrid-intelligent approach to enhance measurement accuracy. The designed sensing system offers transformative potential for monitoring temperature changes in quantum computing, superconducting electronics, and semiconductor fabrication.

Liquid level monitoring ensures precise control of liquid level changes, which is crucial for preventing accidents, ensuring production safety, and maintaining stable system operation. In this work, we theoretically proposed and experimentally demonstrated the integration of a Mach–Zehnder interferometer (MZI) based on cascaded peanut-shaped fiber structures with a random forest network for liquid level monitoring. Because of the phase difference arising during the transmission process between core mode and cladding modes, changes in the liquid level sensed by the sensor result in corresponding changes in the phase difference, subsequently causing a shift in the transmission spectrum. The experimental results show that within the range of 0 mm to 25 mm, the sensitivity of the sensor is 0.084 nm mm⁻¹, with a surprisingly high mean squared error (MSE) of 3.826. The significant error limits its practical application potential. To improve upon the high MSE, a random forest network was introduced for data optimization and training. By constructing multiple decision trees and using their predictions to obtain a final liquid level prediction through voting or averaging, the accuracy and robustness of the model were enhanced, improving the MSE to 0.000355, which represents a reduction in error by over 10 000 times. The designed sensing system holds promise for potential applications in soil moisture systems and liquid level detection in tank trucks. Additionally, the random forest network employed in this system demonstrates universal applicability within point-type optical fiber sensing systems which is possible to be widely utilized.

The measurement of salinity and temperature is vital for ocean observation. However, due to the cross-sensitivity phenomenon of fiber sensors, traditional fiber-based demodulation algorithms are difficult to measure the dual-parameter accurately and simultaneously. In this research, a cascaded structure of a tilted fiber Bragg grating (TFBG) and a Sagnac interferometer (SI) is employed to monitor salinity and temperature simultaneously. The transfer learning based ResNet network (TFRes) is proposed to demodulate two variables at the same time. A total of 1680 spectral samples were collected for training and testing when the sensor was immersed in sodium chloride solution with different salinities ranging from 0% to 25% with 5% intervals and temperatures ranging from 20°C to 32°C with 3°C intervals. The SI exhibits a temperature sensitivity of -1.458 nm/°C, accompanied by a mean absolute error (MAE) of 0.28 °C. The sensitivity of the TFBG core mode to temperature is 0.009 nm/°C, and the sensitivity of one of the cladding modes to salinity is 0.004 nm/%. After training through the TFRes, we successfully achieved precise demodulation of salinity and temperature variations by analyzing the composite reflection spectra of SI and TFBG. The MAE amounted to 0.07304 °C for temperature and 0.07285 % for salinity, outperforming traditional analyzer demodulation methods by a factor of four. The monitoring experiment of dual parameter simultaneous changes is conducted, with a temperature error of 0.18 °C and a salinity error of 0.1%. The designed sensing system is poised to play a significant role in marine physical quantity monitoring.

A novel demodulation scheme for a point-type fiber sensor is designed for salinity concentration monitoring based on a Sagnac interferometer (SI) composed of a tapered polarization-maintaining fiber (TPMF) and optical time stretching technology. The SI, constructed using a PMF with a taper region of 5.92 μm and an overall length of 30 cm, demonstrated a notable enhancement in the evanescent field, which intensifies the interaction between the light field and external salinity. This enhancement allows for a direct assessment of salinity concentration changes by analyzing the variations in the SI reflection spectra and the experimental results indicate that the sensitivity of the sensor is 0.151 nm/‰. In contrast to traditional fiber optic sensors that depend on spectral demodulation with slower response rates, this work introduces a new approach where the spectral shift is translated to the time domain, utilizing a dispersion compensation fiber (DCF) with the demodulation rate reaching up to 50 MHz. The experimental outcomes reveal that the sensor exhibits a sensitivity of −0.15 ns/‰ in the time domain. The designed sensor is anticipated to play a pivotal role in remote, real-time monitoring of ocean salinity.

High-resolution refractive index (RI) detection is of great significance in many fields such as biomedical detection. Traditional optical fiber RI sensors based on spectral wavelength analysis are limited by the device resolution of the optical spectrum analyzer (OSA). In this paper, a micro-D-shaped fiber laser sensor based on beat frequency analysis for RI sensing is proposed and demonstrated to be with high sensitivity and high resolution. Compared with ordinary side-polished fiber, the birefringence of micro-D-shaped fiber is more significant and has much higher response sensitivity. The beat frequency signal response to the external RI change is observed with a low-cost oscilloscope (range of 0-500 MHz). The experimental results show that the polarization mode beat frequency based on a micro-D-shaped fiber laser has a sensitivity of -1.35 GHz/RIU and a demodulation resolution of 7.40×10-6 RIU. The RI detection resolution of the proposed sensing system is superior to various optical fiber RI sensors based on spectral wavelength shift, light intensity resolution, and speckle demodulation reported in recent years, showing broad application prospects in the fields of high precision RI detection, such as biochemical analysis and tumor marker detection.

A new ultrafast interrogation is proposed based on the peanut shaped tapered fiber (PSTF) interference structure and time-stretching method. The single mode fiber (SMF) is tapered by tapering machine to achieve a minimum core diameter width of 10.86 μm. Thanks to the peanut structure as a beam expander and collector, multiple high order modes have been excited. The interference between core mode and cladding modes has a high response to changes in surrounding salinity, with a salinity sensitivity of 0.302 nm/‰ and a corresponding refractive index (RI) sensitivity of up to 1800.535 nm/RIU. The time-stretching method is put forward and experimentally demonstrated. Once the femtosecond light pulse is modulated by the PSTF interferometer, the pulse light carrying information is transmitted to a dispersion compensation fiber (DCF) to achieve wavelength to time mapping and the spectrum in the time domain can be directly collected on an oscilloscope (OSC). By monitoring the time shift of salinity modifications, a salinity sensitivity of -0.128 ns/‰ and a real-time acquisition speed of 50 MHz for the sensing system is realized. The proposed salinity monitoring system has the advantages of high sensitivity, fast response speed, low cost and high accuracy. It has significant potential for real-time monitoring and timely provision of salinity data, which can help improve water quality and achieve scientific aquaculture. The proposed interrogation technology has broad application prospects for any real-time monitoring based on fiber optic interference modulation.

Carcinoembryonic antigen (CEACAM5), as a broad-spectrum tumor biomarker, plays a crucial role in analyzing the therapeutic efficacy and progression of cancer. Herein, we propose a novel biosensor based on specklegrams of tapered multimode fiber (MMF) and two-dimensional convolutional neural networks (2D-CNNs) for the detection of CEACAM5. The microfiber is modified with CEA antibodies to specifically recognize antigens. The biosensor utilizes the interference effect of tapered MMF to generate highly sensitive specklegrams in response to different CEACAM5 concentrations. A zero mean normalized cross-correlation (ZNCC) function is explored to calculate the image matching degree of the specklegrams. Profiting from the extremely high detection limit of the speckle sensor, variations in the specklegrams of antibody concentrations from 1 to 1000 ng/mL are measured in the experiment. The surface sensitivity of the biosensor is 0.0012 (ng/mL)⁻¹ within a range of 1 to 50 ng/mL. Moreover, a 2D-CNN was introduced to solve the problem of nonlinear detection surface sensitivity variation in a large dynamic range, and in the search for image features to improve evaluation accuracy, achieving more accurate CEACAM5 monitoring, with a maximum detection error of 0.358%. The proposed fiber specklegram biosensing scheme is easy to implement and has great potential in analyzing the postoperative condition of patients.