Vibration signal processing for interferometric optical fiber sensor

A method for processing vibration signal for an interferometric optical fiber sensor is proposed in this article. A spatial correlation technique combined with a digital image approach based on the number and direction of fringe motion is suggested for random signal processing. A quantitative physical analysis is formulated, and a signal processing system including acquiring images, analyzing and displaying software is developed. The random signal analysis is carried out by determining the strain-time curve which is based on and is calibrated by the strain-time curve in a stationary vibration. The experiment results are in good agreement with the theoretical one with a maximum difference of 3.5%. The current method is suitable for measuring low frequency vibration. The low frequency is limited due to the limitation of correlation method and the image-sampling rate. For high frequency stationary vibration the method of spectra analysis can be used, but it cannot measure random vibration. High frequency random vibration response may be measured when a high speed photography system combined with a digital image processing setup and software developed in this article are used. This research provides an approach for monitoring vibration for the purpose of making structural engineering studies. (C) 2004 American Institute of Physics.