This article presents a three-axis high-sensitivity vibration sensor based on ultrafine and ultrashort fiber Bragg gratings (FBGs). A theoretical model for the sensor’s sensitivity and natural frequency is developed by deriving equations for steady-state torque balance. The effects of spring plate thickness, fiber diameter, and FBG packaging length on sensor performance are analyzed by this model. Results show that spring plate thickness has almost no impact, reducing fiber diameter improves sensitivity, and shortening FBG packaging length increases natural frequency. Therefore, based on the ultrafine and ultrashort FBGs, the sensor is 3-D modeled and fabricated. The performance of the sensor is verified through finite element analysis and experiments. The experimental results show that the natural frequencies of the sensor in X-, Y-, and Z-directions are 466, 477, and 445 Hz, respectively, with a working frequency band of 0–300 Hz, and sensitivities of 2258.4, 2240.8, and 2274.2 pm/g, respectively. The linearity in the three-axis directions reaches ${R}^{{2}}=0.9998$ , and the minimum acceleration resolution is about $95\mu $ g. The sensor’s performance, characterized by high sensitivity, wide bandwidth, and excellent three-axis consistency, offering valuable insights for high-sensitivity sensor designs.
In a serial wavelength division multiplexing (WDM) fiber Bragg grating (FBG) sensor network, it is well known that there are challenges in separating overlapping signals, which require high precision and low delays. And using an optical spectrum analyzer as a data source result in demodulation models that are impractical for use in engineering applications. Therefore, an overlapping spectral demodulation model based on transfer learning using a charge-coupled device (CCD) interrogator and light gated recurrent unit (Li-GRU) neural networks is proposed. This model can achieve a low signal demodulation error, even when applied to data collected using a CCD interrogator with low spectral resolution and a high signal-to-noise ratio. We describe the operation principle of the Li-GRU neural network and discuss the impact of transfer learning and CNN feature extraction layers on demodulation performance. The experimental results show that lowest root mean square error of our proposed model is 1.93 pm, and the single inference time of the model on the CPU is <45 ms. This serial WDM fiber grating demodulation method can be effectively applied in temperature and strain measurement demodulation.
Vibration sensing networks, which consist of numerous sensors, require high demodulation speeds. Therefore, optimizing the demodulation algorithm to reduce data processing volume or complexity is crucial for enhancing efficiency and performance. This paper proposes a vibration sensing demodulation method based on the light intensity difference (LID) of charge-coupled device (CCD) pixel points. The method requires calculating the center wavelength of all CCD pixel points only once. It then selects pairs of pixel points with opposite phases in the reflection spectrum to perform differential operations, significantly reducing data processing volume and computational complexity, thereby improving demodulation speed. Additionally, by increasing the number of calculation pairs, the method enhances demodulation resolution, suppresses noise, and improves robustness. The method is modeled through simulation and theoretical analysis, with results indicating reduced background noise influence. Tested with fiber Bragg grating (FBG) vibration sensors, the method achieves a resolution of 62 μg and a demodulation rate of 16 kHz, highlighting its promising application prospects in high-speed and large-capacity vibration sensing.