Research On Wavelength Feature Recognition Method Of Large-capacity Fiber Grating

In today’s society,the rapid development of Internet of Things technology has made great contributions to human society into an intelligent society,especially in the field of Internet of Things perception,fiber Bragg grating as a very important data acquisition means,can real-time online monitoring of strain,temperature,vibration and other parameters.It has the advantages of passive,corrosion resistance,anti-electromagnetic interference,fast measurement,long-distance transmission,etc.,which make fiber Bragg grating can be used in various extreme and complex environments.It is used to monitor applications such as structural health,fire alarm,perimeter security and other applications in power engineering,deep tunnel engineering,traffic tunneling and energy exploration.This paper mainly introduces the basic theory,denoising algorithm and wavelength feature recognition technology of FBG sensing technology,and uses these aspects as a theoretical basis to analyze and improve the existing optical fiber sensing demodulation technology.According to the actual engineering requirements and the sensing characteristics of the fiber grating,We propose a modified Gaussian nonlinear fitting algorithm based on Hilbert transform.Firstly,the sensing characteristics of the fiber grating sensing network are analyzed,and the signal is pre-processed by using the local supplementary overall mean empirical mode decomposition technique,and then the high-frequency signal is further refined by boosting the wavelet transform technique,so as to achieve the purpose of improving the signal quality,This method effectively solves the problems of insufficient accuracy and long signal processing time for high-frequency signals by traditional signal processing methods.Secondly,aiming at the poor accuracy of traditional unimodal detection algorithms in FBG demodulation,a Gaussian nonlinear fitting algorithm based on spectral symmetry correction is proposed for multi-peak detection.The input reflection spectrum is segmented and derived by the Hilbert transform to obtain multiple initial peaks,and then the exponentially corrected Gaussian nonlinear fitting algorithm is used to correct the peak position.Whether dealing with asymmetric or multimodal spectra,this method guarantees good measurement accuracy.Finally,in order to verify the feasibility of the signal processing method proposed in this paper in practical applications,an experimental platform for large-capacity FBG temperature sensing demodulation system is designed and built.The results of temperature sensing experiments show that the average temperature error of the demodulation system is about 0.2°C within the range of temperature measurement from-30°C to 50°C,and the linear fit of the demodulation can reach more than 99.9%.The method proposed in this paper has good application value,which can improve the signal processing and sensing performance of large-capacity FBG sensing demodulation system in practical engineering applications,and has certain theoretical reference significance.