Detection of Gaussian-like vibration in fiber optical distributed sensing by 4th-order cumulants

Mechanical equipment failures, e.g., conveyor belt and engine failures, often behave as Gaussian-like distribution. This paper proposes a 4th-order cumulant algorithm based on the phase-sensitive optical time-domain reflectometer to detect and analyze vibrations with Gaussian-like probability distributions. The 4th-order cumulants of the Rayleigh backscattering along the sensing fiber are calculated to obtain the Gaussian-like vibration signal from the background white noises. The modeled results show that the 4th-order cumulant algorithm can efficiently recognize the Gaussian-like vibrations with enhanced signal-to-noise ratio and position accuracy. In experiments, a sawtooth signal with a uniform probability distribution is used as the driving signal of the vibration. Experimental results show that the signal-to-noise ratio of location signal reaches up to 15.5 dB, and the position accuracy increases to 2.7 m with pulse width of 100 ns.