Noise character and filtering algorithm in fiber optic current transducer

Abstract

Compared with electromagnetic current transducer and electronic current transducer, fiber optic current transducer (FOCT) represents the developing direction of current transducer, which has a series of advantages, such as wide bandwidth, good insulation performance, high measuring accuracy, and so on. FOCT is one kind of new-type precise sensor based on the principle of Faraday magneto-optical phase circling, its' complicated structure results in shot noise, intensity noise, hot phase noise, and electromagnetic interference noise, which have great effect on FOCT's performance. Relativity of noise random process in time domain and stability of noise random process in frequency domain are analyzed, based on the analysis results, noise identification model and algorithm are proposed, and different type noise in FOCT showing main character is quantified by using noise identification model and algorithm. Aiming on decreasing the influence of noise to FOCT performance, filtering algorithm is applied in FOCT data processing. Experiments show that noise character is complex, noise dentification model and noise filtering algorithm are effective in FOCT noise processing system.