Noise Signal Recognition and Noise Reduction Algorithm of Ships

Based on the study of sensing property fiber bragg grating (FBG) sensor network, this paper takes hybrid wavelength division/time division multiplexing network with high-capacity FBG as research object to noise source, category and traditional reduction algorithm. In accordance characteristic crosstalk accumulation in proposes LCEEMD-LWT signal processing method, which uses local complementary ensemble empirical mean decomposition (LCEEMD) method for preprocessing, then adopts LWT technology refine high-frequency solve related problems. Then put forwards detrended fluctuation analysis (DFA) evaluate FBG, preprocesses by LCEEWD-LWT conducts a temperature experiment. The average error demodulating system is reduced from 0.2923℃ 0.2357℃, not only overcomes shortcoming that meticulous processing, but also avoids problem time too long, exerting good effect pre-processing FBG. has certain theoretical technical reference value improvement engineering application.