| White-light coherence sensor is a function of fiber-optic sensor which takes white-light scanning fiber interferometer as sensitive device. The practical sensors consistituted with it have a broad application field. When the sensor undergone the external parameters, such as temperature, strain and the location of the central stripe drift correspondingly. The change can be measured accurately, so it's widely used in highly accurate parameter measurement, all the developed countries pay great attention to it. And it becomes one of the most popular techonoloy in fiber sensoring field. Due to influnces of environment device etc. on it, coherence signal has a lot of noise with it, which is bad for the accuracy. So the key point is how to decrease the interference of noise effectively, and locate the central stripe's position exactely.Interference from different elements during the measurement, such as: light source, electonic parts, optical parts, have bad effects on white-light coherence signal, so denoising ahead of processing shows great importance. Denosing not only decreases the noise in the white-light coherence signal, but also makes the white-light coherence signal smooth. All these set a good base for the next step; peak value locating is the final aim of all these above, the accuracy of the peak value locating determines that of the whole systerm, and the peak value location is close with the parameter to be measured.This thesis mainly analyzes white-light signal, based on wavelet transform thresh algorithm and envelope fitting, containing signal de-noising, peak value locating and measuring optical path difference reaseach. Through experiment results comparing analysis, proves their each excellences and flaws. Wavelet thresh shrink algorithm can denoise the white-light sigal well. Algorithm based on envelope fitting has advantages of fast speed, large anti-interference capability in white-light signal processing field. |
PDF Full Text Request