Abstract
We utilize singular spectrum analysis (SSA) for post-processing distance measurement data obtained by a dual-comb femtosecond laser rangefinder. Signal denoising technique based on SSA allows detection of low amplitude vibrations from time series of distance data with limited signal-to-noise ratio (SNR). Various one-dimensional vibrations, including harmonic, amplitude modulated harmonic, and sinc-function vibrations, with micrometer amplitude comparable with measurement uncertainty of laser rangefinder, have been imposed on the target mirror by a piezo positioner. These vibrations have been extracted by SSA and the obtained vibration amplitudes agree with that imposed on the target very well. A commercial interferometer is used to evaluate the accuracy of SSA. Standard deviations of residuals between the SSA extractions and interferometer measurements are smaller than 0.6 $\mu$m. This technique is promising for complex moving pattern detection in object tracking and remote sensing.
Highlights
Non-contact vibration measurement is crucial for a number of industrial applications, such as mechanical damage detection [1], [2], microsystem diagnosis [3] and biomedical imaging [4]
The noisy absolute distance measurements have been processed by singular spectrum analysis (SSA), and decomposed into components classified as oscillation and noise, allowing the extraction of target motion from measurement noise
For same window length and signal length, the errors introduced by each reconstructed components (RCs) are approximately same
Summary
Non-contact vibration measurement is crucial for a number of industrial applications, such as mechanical damage detection [1], [2], microsystem diagnosis [3] and biomedical imaging [4]. The high measurement precision allows dual-comb rangefinder to resolve small amplitude vibrations, and the upper limit for the detectable vibration frequency is determined by half of update rate. Extracted from measurement noise without loss of update rate, even for irregularly moving and vibrating targets. This method is only effective in the presence of Gaussian measurement noise with known intensity. We utilize a free running dual-comb laser rangefinder to measure a vibrational target. The noisy absolute distance measurements have been processed by SSA, and decomposed into components classified as oscillation and noise, allowing the extraction of target motion from measurement noise. The vibrations can be extracted out of the measurement noise even when the target motion amplitude is comparable with measurement noise. The retrieved vibrations are compared with the results of the commercial interferometer and agree with the vibrations imposed on the target very well
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
