Application Research Of Wavelet Denoising Method In Fiber Current Transformer

With the rapid development of human society and economy,the power consumption level of urban residents and the pressure of distribution network have gradually increased.In order to ensure the safety of electrical equipment and residential power consumption,it is particularly important to effectively detect the current of transmission lines.Fiber-Optical Current Transformer(FOCT),based on Faraday magneto-optic effect,has gradually become an indispensable component of power grid for detecting and monitoring transmission line current in distribution network planning by transforming light energy into magnetic energy and then into light energy.However,because of the special working environment of FOCT and the complex noise contained in its output signal,the long-term stability and detection accuracy of FOCT cannot meet the requirements.In order to improve the signal-to-noise ratio of FOCT output signal,an improved threshold algorithm based on the existing wavelet threshold algorithm is proposed,and it is named D-Exp threshold algorithm according to its improved mechanism,According to the output characteristics of FOCT,an improved denoising scheme(Improved De-noising scheme)was proposed,based on D-Exp threshold denoising algorithm,which is called Imp-Dn denoising scheme.In addition,a wavelet threshold denoising system based on Labview is designed,which combines Matlab and Labview.Through this system,the performance of FOCT noisy output signal is compared with different threshold de-noising algorithms.This paper mainly carries out research work on the following aspects:(1)The mathematical models of the output signal of Sagnac reflective FOCT and the relationship between the noise variance of FOCT output signal and its power spectral density are analyzed respectively.The experimental simulation of these two mathematical models is realized,which verifies the noise complexity of FOCT.(2)In order to determine the selection of wavelet basis function and decomposition layer,the decomposition and reconstruction errors of sym and db wavelet bases of different order and the denoising effect of different decomposition layer are simulated and analyzed by using MATLAB.Through simulation and comparison,it is determined that the five-layer decomposition of sym5 wavelet basis is used to denoise the noisy signal of FOCT.(3)Through the wavelet threshold denoising system,D-Exp threshold algorithmand Imp-Dn denoising scheme are used to compare the denoising effect of FOCT noisy output signal.After analysis,when the initial SNR of FOCT is-3.7665,after de-noising by D-Exp threshold algorithm,the SNR increases 130.7%,RMS decreases 43.2%,SNR increases 312.4% and RMSE decreases 74.2%.It shows that the improved threshold algorithm and denoising scheme have strong anti-jamming ability under the condition of low signal-to-noise ratio of FOCT.(4)Finally,in order to verify the robustness of the D-Exp threshold algorithm,three groups of Sagnac ring signals are denoised and compared.The results show that the D-Exp thresholding algorithm has good robustness in practical application,no matter in time domain analysis,frequency domain analysis and comparing the distance between classes after denoising with different thresholding algorithms.