Signal processing of fiber-optic strain measurement system based on low-coherent interference

There's broad development prospect of applications of fiber-optic strain measurement technology used in structural health monitoring of materials and signal processing is one of the core issues. Signal processing methods of a low-coherent fiber-optic strain measurement system based on Michelson interferometer. SNR was compared after the wavelet threshold noise reduction using rules of rigrsure, sqtwolog, heursureg and minimaxi, respectively. The relative position of center fringes was located by the use of wavelet envelope extraction. Three-point bending was employed to test strain of the steel plate, Db10 wavelet was chosen to decompose signal and rigrsure threshold rule was selected to process and reconstruct signal, all of these were employed to realize a signal denoising. The fitting line showed linearity of the system was ±0.55%, which demonstrated that these methods could improve accuracy of the system.