Zhengrong Tong

When measuring pressure using a fiber Fabry-Perot (F-P) sensor, cross-sensitivity between pressure and temperature poses a significant challenge to accurate sensing. To address this issue, a genetic algorithm-backpropagation (GA-BP) neural network is employed to construct a pressure-temperature decoupling model for an all-silicon dual-cavity fiber-optic pressure sensor, thereby enabling effective temperature compensation. To alleviate the computational burden and enhance model efficiency, the collected spectral data are first processed through feature extraction to generate a representative feature dataset. This dataset is subsequently fed into the BP neural network as input, where the strong nonlinear mapping capability of the BP model enables accurate regression prediction of pressure values. Meanwhile, to avoid the BP model converging to a local optimum during training, a genetic algorithm (GA) is introduced to optimize the parameters of the BP neural network. Experimental results verify the effectiveness of the proposed approach. Within a pressure range of 20 kPa to 280 kPa and a temperature range of -20 °C to 60 °C, the deviation from the pressure setpoint is maintained between -0.002 kPa and 0.0016 kPa, with the pressure-temperature cross-sensitivity lowered to 0.045 Pa/°C.

To address the intrinsic bottleneck coupling high sensitivity with a broad measurement range in optical fiber interferometric sensors caused by free spectral range (FSR) limitations, this paper proposes a novel intelligent demodulation method. This approach combines the optical Vernier effect with an attention-integrated deep convolutional neural network (DCNN-Attention). By connecting two parallel open-cavity MZIs with slightly disparate lengths, the Vernier effect yields an ultra-high refractive index (RI) responsivity reaching –152702.38 nm/RIU. However, conventional demodulation schemes constrain the detectable span to 0–9 mg/mL, yielding a sensitivity and R2 of –20.91 nm/(mg/mL) and 0.999151, respectively. To overcome this bottleneck, the DCNN-Attention model is introduced to perform end-to-end intelligent demodulation directly from raw interference spectra. This approach entirely eliminates the need for manual spectral preprocessing and Vernier envelope extraction, thereby avoiding the extraction errors and system noise impact associated with these conventional steps. Consequently, the sensing detection span is dramatically stretched to 0–200 mg/mL, representing a substantial 22-fold improvement. Experimental results demonstrate that within this broad range, the proposed model achieves a remarkable R2 of 0.999997, exhibiting superior demodulation accuracy. By synergistically combining physical sensing mechanisms with intelligent algorithms, this work effectively breaks the physical limitations imposed by the FSR. Ultimately, this method offers a new paradigm for high-performance optical sensing, showing broad application prospects in health monitoring and food safety analysis.

A Mach-Zehnder interferometer (MZI) fiber optic sensor based on a spindle-shaped structure is designed and fabricated, and with the assistance of a neural network, it enables accurate micro-displacement measurement over a wide range. The single-mode fiber (SMF) is bent by baking with a flame to form a spindle-shaped fiber optic sensor. Due to the curved structure of the optical fiber, some light leaks from the core into the cladding and excites higher-order cladding modes. The sensing characteristics of micro-displacement and temperature are analyzed by comparing sensors with three different bending diameters. Experimental results show that when the bending diameter is 2.5 mm, the sensor exhibits a sensitivity of –725.49 pm/μm over the micro-displacement range of 0–50 μm, with an R² of 0.99935. To expand the measurement range of the sensor and improve its practicality, a Long Short-Term Memory (LSTM) neural network is introduced to analyze the measurement signal. LSTM, with its recurrent mechanism and gating technology, can efficiently process sequence data, thereby overcoming the dynamic range limitation of traditional trough tracking methods. While maintaining measurement accuracy, the effective micro-displacement measurement range is expanded to 0-500 μm, which is 10 times that of traditional trough tracking methods. Furthermore, the sensor is insensitive to common temperature variations of 20-70℃, effectively avoiding temperature interference with micro-displacement measurement. Combining a spindle-shaped fiber optic sensor with neural network analysis offers a new technical approach for expanding the application of fiber optic sensing technology in high-precision measurement and industrial inspection.

A blood glucose concentration and temperature sensor with a balloon-shaped single-mode fiber (SMF) based on a core-offset structure is proposed and experimentally demonstrated. The balloon-shaped SMF is created by offset-fusing a straight-line SMF between two other SMFs, thereby forming a Mach–Zehnder interferometer (MZI). The core-offset structure can effectively excite higher-order cladding modes. The experimental results showed that the maximum sensitivity of blood glucose concentration was 0.331 nm/(mmol/l) and the maximum sensitivity of temperature was 0.216 nm/°C when the offset distance was 10 μm. Dual-parameter measurement was achieved through a dual-parameter matrix. In addition, the sensor has characteristics such as simple structure, low cost, good stability, and electromagnetic interference resistance, making it potentially valuable for diagnosing high blood glucose and related conditions.

In the fields of structural health monitoring, civil engineering safety monitoring, aerospace and other fields, the key challenge is to increase the range of strain measurements to solve with the information monitoring of large strains. Reducing the strain sensitivity of the sensor is an important way to solve the problem of small sensor ranges. In this paper, the inclined transfer structure is proposed to achieve large strain distributed optical fiber sensing. When the fiber is axially stretched directly to fracture, the inclined transfer structure is used to increase the strain measurement range of the system. The theoretical relationship between the strain monitored by single mode fiber sensors and the matrix strain is established. It is experimentally verified that the inclined transfer structure using polypropylene material as the matrix achieves the strain measurement from 36667[ ][ ] at sensing spatial resolution of 8[ ]mm, 10[ ]nm wavelength tuning range of the laser source and [ ] inclined angle, obtaining a good linear fit of more than 0.99. One hour stability of less than 11[ ]pm is obtained, corresponding to the strain of 9[ ][ ].