Search

Efficiently extracting long-term sequence features is the key for deep learning models to accurately predict the remaining useful life (RUL) of mechanical bearings. However, completely ignoring local features may result in losing a large amount of prediction information. In addition, the existing soft thresholding function loses a large number of feature values when denoising, which affects the prediction performance of the network. A new temporal convolutional neural network with an adaptive activation function and attention mechanism (TCNAAM) for RUL prediction is proposed in this article. The network consists of two branches; one branch is to be utilized to adaptively select key local information in the time series using an adaptive convolutional scale attention layer. The other branch is feature denoising by adding an improved activation function after the temporal convolutional neural network during training, to preserve most of the original features. Therefore, TCNAAM can simultaneously extract long-term and short-term information and constrain all eigenvalues within a certain interval. The bearing dataset on Xi'an Jiaotong University and IEEE PHM Challenge 2012 is used to validate the effectiveness of the proposed TCNAAM. Experimental results show that the proposed TCNAAM performs better than some existing prediction methods for bearings RUL prediction applications under different operating conditions.

We demonstrated an inflated long period fiber grating (I-LPFG) inscribed in a pure-silica photonic crystal fiber (PCF) for high-sensitivity gas pressure sensing applications. The I-LPFG was inscribed by use of the pressure-assisted CO_2 laser beam-scanning technique to inflate periodically air holes of a PCF along the fiber axis. Such an I-LPFG with periodic inflations exhibits a very high gas pressure sensitivity of 1.68 nm/MPa, which is one order of magnitude higher than that, i.e., 0.12 nm/Mpa, of the LPFG without periodic inflations. Moreover, the I-LPFG has a very low temperature sensitivity of 3.1 pm/°C due to the pure silica material in the PCF so that the pressure measurement error, resulting from the cross-sensitivity between temperature and gas pressure, is less than 1.8 Kpa/°C in the case of no temperature compensation. So the I-LPFG could be used to develop a promising gas pressure sensor, and the achieved pressure measurement range is up to 10 MPa.

We demonstrated a high-sensitivity strain sensor based on an inflated long period fiber grating (I-LPFG). The I-LPFG was inscribed, for the first time to the best of our knowledge, by use of the pressure-assisted CO_2 laser beam scanning technique to inflate periodically air holes of a photonic crystal fiber. Such periodic inflations enhanced the sensitivity of the LPFG-based strain sensor to −5.62  pm/μ ε. After high temperature annealing, the I-LPFG achieved a good repeatability and stability of temperature response with a sensitivity of 11.92 pm/°C.

We demonstrated a high-sensitivity strain sensor based on an in-fiber Fabry–Perot interferometer (FPI) with an air cavity, which was created by splicing together two sections of standard single-mode fibers. The sensitivity of this strain sensor was enhanced to 6.0  pm/με by improving the cavity length of the FPI by means of repeating arc discharges for reshaping the air cavity. Moreover, such a strain sensor has a very low temperature sensitivity of 1.1  pm/°C, which reduces the cross sensitivity between tensile strain and temperature.

We present a type of phase-shifted fiber Bragg gratings based on an in-grating bubble fabricated by femtosecond (fs) laser ablation together with a fusion-splicing technique. A microchannel vertically crossing the bubble is drilled by fs laser to allow liquid to flow in or out. By filling different refractive index (RI) liquid into the bubble, the phase-shift peak is found to experience a linear red shift with the increase of RI, while little contribution to the change of phase shift comes from the temperature and axial strain. Therefore, such a PS-FBG could be used to develop a promising tunable optical filter and sensor.

Photonic crystal fibers are usually divided into two different types of fibers: solid-core photonic crystal fibers (PCFs) and air-core photonic bandgaps fibers (PBFs). We presented the fabrication methods and applications of long period fiber gratings (LPFGs) written in these two types of photonic crystal fibers by use of a CO2laser. A stain sensor with a high sensitivity was demonstrated by use of an LPFG written in solid-core PCFs. An in-fiber polarizer based on an LPFG was fabricated by use of a focused CO2laser beam to notch periodically on a PCF. A novel LPFG was written in an air-core PBF by use of a CO2laser to collapse periodically air holes in the fiber cladding.