Study On White Light Extrinsic Fabry-Perot Interferometric Optical Fiber Sensor And Its Applications

Extrinsic Fabry-Perot interferometric (EFPI) optical fiber sensor has won widely attention for its small size, compact configuration, good reliability and flexibility. It can be widely used in most harsh environments, such as oil well, electric power industry and chemical industry. In this dissertation, several key problems are studied systematically, which include the demodulation mechanism in spectral domain of the EFPI optical fiber sensor, and the performance evaluation of the sensing system. The main research works are outlined as foliowings:A white light optical fiber EFPI sensor system in spectral domain based on a miniature spectrometer is implemented. The normalized interference spectrum is obtained by storing the stable light source spectrum and the back scattering spectrum. The online normalization of the interference spectrum in frequency domain is implemented with Hilbert transform. Therefore the changes of the spectrum shape and the peak location have no effect on this normalizing course.A demodulating method of the EFPI sensor based on BP neural networks is presented. By tracing a special peak point in the interference spectrum, the gap length of the EFPI sensor can be demodulated. The order ambiguity problem of single peak tracing method is solved through the application of BP neural networks. Both high precision and large dynamic range are achieved with this method. The calibrating and measuring experiment of the EFPI strain sensor based on the BP neural networks method is also carried out. For a strain range more than 500με, a resolution less than 0.1με is achievedA demodulating method of the EFPI sensor based on cross correlation is presented. The discrete form of the cross correlation calculation is implemented. Fast algorithm of discrete cross correlation calculation is given and the real time capability of demodulating method is improved. The gap length range of cross correlation calculation is reduced by estimating the gap length of the EFPI sensor with Fast Fourier Transform algorithum.The effects of some facts as the drift of the light source, the distortion of the optical fiber and the reference drift of the spectrometer on measuring stability of the EFPI sensor system based on cross correlation demodulating method are evaluated. A resolution of the gap length less than 0.2nm is achieved in the EFPI sensor system with 1.5Km connecting optical fiber. The anneal effect of the EFPI sensor is evaluated. The residual stresses are released under...