The data processing of multi-wavelength pyrometry (MWP) is faced with the problem of solving N equations and N+1 unknown underdetermined equations. The traditional iterative optimization methods are difficult to meet the actual measurement requirements in terms of accuracy and efficiency. With the development of artificial intelligence technology in the field of data processing, it is expected to solve this problem. A generalized inverse matrix (GIM) is combined with a long short-term memory (LSTM) neural network algorithm for data processing of MWP is proposed, which emissivity influence is dispensed completely. Firstly, GIM is used for classification of the emissivity. Furthermore, inputting to the LSTM network not only ensures the accuracy of temperature measurement but also greatly improves the efficiency. The simulation results demonstrated that the accuracy of the GIM-LSTM algorithm was superior to that of the GIM-EPF and BP methods. After random noise was added, the relative error was still less than that for the GIM-EPF and BP methods, and the algorithm exhibited excellent anti-noise performance. Publicly available temperature data for the exhaust plume of a rocket engine were processed by the GIM-LSTM method, and the average relative error was less than the traditional method. Especially, in terms of inversion speed, the operational time of the GIM-LSTM algorithm was at the millisecond level, which is of great significance for the real-time monitoring of rocket exhaust plumes. The proposed GIM-LSTM data processing algorithm affords high accuracy and speed and is suitable for practical measurement of high-emissivity objects in real-time via MWP.

This paper presents and demonstrates a novel in-fiber integrated high sensitivity temperature sensor based on long Fabry-Perot (FP) resonator. In a quartz capillary, the FP resonator was composed of two single mode fibers (SMFs) whose end faces were coated by gold films. The temperature can be obtained by measuring the length variation of the FP cavity. A white light interference demodulation system was used to measure the length variation of the FP cavity. By the multiple reflections of light in the FP cavity, we achieved the sensitivity multiplication of the sensor. The proposed sensor measured the temperature up to 350°C for 2 hours, and the sensitivity of the sensor is six times that of the traditional interference temperature sensor. Due to the advantages of low cost, high sensitivity and simple fabrication, this temperature sensor can be widely used in high temperature applications.

Neural networks are the most promising tool to solve the problem that an assumed emissivity model is needed in the field of multispectral radiometric temperature measurement. Existing neural network multispectral radiometric temperature measurement algorithms have been investigating the problems of network selection, network porting, and parameter optimization. The inversion accuracy and adaptability of the algorithms have been unsatisfactory. In view of the great success of deep learning in the field of image processing, this Letter proposes the idea of converting one-dimensional multispectral radiometric temperature data into two-dimensional image data for data processing to improve the accuracy and adaptability of multispectral radiometric temperature measurement by deep learning algorithms. Simulation and experimental validation are carried out. In the simulation, the error is less than 0.71% without noise and 1.80% with 5% random noise, which improves the accuracy by more than 1.55% and 2.66% compared with the classical BP (backpropagation) algorithm, and 0.94% and 0.96% compared with the GIM-LSTM (generalized inverse matrix–long short-term memory) algorithm. In the experiment, the error is less than 0.83%. This indicates that the method has high research value and is expected to lead multispectral radiometric temperature measurement technology to a new level.

An all-fiber bend sensor based on Mach-Zehnder interferometer (MZI) is proposed and experimentally demonstrated. The fiber interferometer consists of two helical structures in single mode fiber (SMF) and is fabricated by continuous arc discharge technique. The first helical structure forms a twisted long-period fiber grating (LPFG) to excite the higher-order cladding modes and acts as a beam splitter. The LPFG arrangement is adopted to improve the bending sensitivity. The second helical structure only contains two turns and can recouple the light in the cladding back into the core. The bending characteristics of the MZI are discussed. The experimental results show that high curvature sensitivities of 24.06 nm/ $\text{m}^{-1}$ and −7.524 dB/ $\text{m}^{-1}$ are obtained in the curvature range of 0-0.422m ⁻¹ . Due to its low cost, small size, simple preparation, less time consuming, it could be a competitive candidate in many bending sensing applications.

A high sensitivity strain sensor based on micro-helix micro taper long period fiber grating (MH-MTLPFG) is proposed and experimentally investigated. The MH-MTLPFG is fabricated by periodically tapering and micro helix structure on a standard single-mode fiber with arc discharge technology. The arrangement of micro helix taper makes the sensor have high strain sensing sensitivity. The experiment results show that the sensitivity of strain sensing is up to −3.478 pm/ $\mu \varepsilon $ . Compared with the conventional MTLPFG sensor, the strain sensitivity is improved by about 10 times. The proposed high sensitivity strain sensor has a good potential for accurate strain measurement, due to its low cost, simple fabrication and good performance.

To address the problem of decreased recognition accuracy of event samples in practical phase-sensitive optical time-domain reflectometer (Φ-OTDR) monitoring scenarios due to external environmental interference, this paper proposes a feature correction algorithm based on sample feature weighting method. By establishing a correlation evaluation method and a weight allocation scheme based on sample feature correlation, combined with the back propagation (BP) algorithm, an average recognition rate of 99.50% for four types of events (climbing, strong wind, knocking and background, 6000 samples) in strong wind environments was achieved, which is 3% higher than the algorithm using BP classifier. The results demonstrate that the proposed algorithm can effectively enhance the performance of Φ-OTDR in complex environments.