Research Of The Noise Suppression For Fiber Optic Current Transformer

As a novel current transformer, Fiber Optic Current Transformer (FOCT) hasincomparable advantages, compared with traditional electromagnetic currenttransformer. No magnetic saturation, good insulating property, large dynamic rangeand good anti-interference performance make FOCT having a broader application ofspace and development prospects. In this dissertation the problems of FOCT inpractical development is put forward. The research work about the improvedperformance-related technology of FOCT is to be carried out. In this article the maincontent can be summarized as follows:The principle of FOCT with reflective structure is described and from thetheoretical analysis, closed-loop for phase modulation can obtain higher sensitivityand accuracy in comparison with open-loop. First, the impact of four main noises onFOCT’s output was analyzed as a whole. According to the characteristics of currentand noise signal in the frequency domain, a kind of real-time digital filtering systembased on Finite Impulse Response (FIR) is proposed. Choosing an appropriatewindow function and filter orders can effectively filter out the noise and the signalnoise ratio(SNR) of FOCT’s output is increased, then the measurement accuracy canbe reached to0.2level. Secondly, the influence of four main noises on FOCT’s outputwas discussed in part. In certain range, increasing optical power can effectivelyimprove SNR of FOCT’s output. When selecting a larger power, FOCT’s SNR isdetermined by the relative intensity noise of light source. In order to eliminate therelative intensity noise, two kinds of technology of inhibiting relative intensity noiseare introduced. Due to the defects in these two schemes, an adaptive noise subtractionsystem is proposed based on the two technical schemes. Since the adaptive algorithmhas a self learning process in the start, the larger forgetting factor is, the longerlearning process is. It will produces adverse effects on real time filtering. To solve thisproblem, a multi-algorithm fusion is applied. The simulation results show that the adaptive noise subtraction technology has a strong adaptive capacity of thetransmission delay and changing amplitude of relative intensity noise. The meansquare error of FOCT’s output signal is greatly reduced by the adaptive noisesubtraction; therefore, in theory this method has a certain significance to improve theperformance of FOCT.