A disturbance detection method with deep learning is proposed and experimentally demonstrated for OTDR. Compared with typical methods, better performance can be intensified by entirely using the multi-dimensional features of sensing data. The adaptability and anti-noise capacity are greatly advanced with the proposed method. In experiments, an ultra-high SNR of 53.98dB can be obtained in the best case. A spatial resolution of 1.06m and SNR of 38.49dB are achieved simultaneously when the pulse width is modulated to the narrowest. Moreover, additive noise and multiplicative noise with a level of 0.015 were applied, respectively. The SNR can reach 35.22dB and 42.39dB. With this approach, the stable high SNR is unprecedentedly improved. This method provides the potential for temporal-spatial disturbance detection in OTDR with strong background noise and extreme conditions.

A novel vibration location method with nested pulses for distributed optical fiber vibration sensing system is proposed. The interferometric structure for vibration detection and location with nested pulse method is firstly designed. The continuous-wave light is modulated into a series of periodic nested pulses in the linear Sagnac interferometer. Each nested pulse contains a certain number of sub pulses to locate the position of vibration signal with the help of envelope derivation and threshold judgment. Experimental results show the location feasibility for vibration signals at different positions of long-range optical fiber, with the positioning error of 10m at 0.5 km and the positioning error of 26 m at 45.314 km. This method provides one solution for the location of actual vibration without the limit of frequency band.

Synchronous intensification of spatial resolution and sensing range has attracted widespread concern for distributed acoustic sensor. The pulse compression approach as a representative technology is adopted for addressing this issue, while the performance is suffered from the Rayleigh side-lobe noise due to the insufficient side-lobe suppression ratio (SLSR). In this work, the poly-phase coding phase-sensitive optical time domain reflectometer (Φ-OTDR), possessing an ultra-high SLSR, is proposed to immensely alleviate the Rayleigh side-lobe noise influence. The spatial resolution is relevant to the single code width, as well as the sensing range and SLSR are relevant to the coding length. An instructive model is deduced to evaluate the influence of SLSR to Φ-OTDR system's SNR. According to the model, a theoretical SLSR value is derived, making the Rayleigh side-lobe noise could be neglected. In the proof-of-concept experiments, the modulated poly-phase coding with ultra-high 127.08 dB SLSR guarantees a 204.3 km sensing range and 2 m spatial resolution. Two sinusoidal perturbances with 20 Hz and 60 Hz are exerted at 173.3 km and 203.8 km on the fiber. As a result, the strain resolutions are achieved at 287.3 pϵ/√Hz and 142.1 pϵ/√Hz, respectively. This work demonstrated that ultra-high SLSR is the foundation for achieving excellent performance for compression pulse distributed acoustic sensors.

In order to enhance the signal-to-noise ratio (SNR) of a distributed optical fiber vibration sensor based on coherent optical time domain reflectometry (COTDR), a high extinction ratio cascade structure of an acousto-optic modulator and semiconductor optical amplifier is applied. The prior time-frequency analysis and least mean square error algorithm are adopted in the COTDR system for amplitude demodulation and phase demodulation, in order to improve the SNR by noise elimination. The experimental results show that the adaptive filter based on the least mean square error algorithm could realize the extraction of a three-order sinusoidal harmonic signal from strong background noise along the optical fiber and the SNR improvement from 10.4 dB to 42.2 dB. The proposed demodulation algorithm is suitable for the detection of vibration signals with characteristic frequencies in the application of acoustic fault diagnosis for electromechanical devices.

A crosstalk-free frequency division multiplexing distributed acoustic sensing (CFFDM-DAS) technology for partial discharge (PD) measurement in power cable is proposed in this paper. The multi-carrier crosstalk-free probe pulse, along with raised cosine roll-off processing, is constructed and implemented to avoid inter-channel crosstalk between the different frequency, significantly increasing multiplexing density. In this way, the CFFDM-DAS is modeled and established with an increased crosstalk suppression ratio of 61.39 dB compared to traditional multiplexing techniques. As a result, the frequency bandwidth of the proposed method is extended to 100 kHz using 10-channel multiplexing with a 5 km sensing range. Validation experiments confirm that the sensing system exhibits an excellent wideband response and reliable acoustic signal reconstruction capabilities. Discharge experiments on a 10 kV XLPE power cable with void insulation defect are conducted. The measurement of PD with the proposed system demonstrates its capacity for detecting discharges with a minimal apparent charge of approximately 31 pC and effectively capturing most PD events. The proposed CFFDM-DAS system presents a promising approach for the reliable and continuous monitoring of PD activity.

Laser-frequency-drift (LFD) restricts the low-frequency vibration sensing ability of phase-sensitive OTDR (Φ-OTDR). In this study, we propose a self-diagnosis structure combining Mach-Zehnder interference and Φ-OTDR is firstly designed to realize LFD measurement and vibration sensing simultaneously. A multi-scale variational mode decomposition (VMD) with the judgement at different amplitude scales, time scales and modal scales is proposed to suppress the influence of LFD at low-frequency vibration demodulation. Experimental results show a long-time continuous LFD measurement and vibration sensing along the sensing fiber of 10040 m. After the parameter optimization of iteration and mode decomposition with multi-scale VMD, infrasonic and sub-Hertz vibration signals can be detected with improved SNR. The ultra-low frequency vibration signal as low as 0.04 Hz is also demodulated under a long-time continuous detection of more than 40 min. The fitting coefficient between the recovered vibration frequency and the actual vibration frequency is 0.99966. This self-diagnostic structure, based on multi-scale VMD, effectively suppresses the influence of LFD on Φ-OTDR and reliably recovers ultra-low-frequency vibrations during long-term detection.