Study On The Hybrid Optical Current Transducers Based On Adaptive Theory

Conventional electromagnetic current transformers(CT) possess high accuracy of steady-state electrical current measurement, but produce problems caused by magnetic saturation. By comparison, optical current transducers(OCT)have many advantages over CT and can measure fault transient current. However, temperature or stress-induced linear birefringence in sensing materials cause error and instability. Based on adaptive signal processing and wavelet transform, hybrid optical current transducers are put forward in this paper, which can bring advantages of CT and OCT into play.Two schemes are advanced in this paper to measure transient current. After many tests on four model reference adaptive signal processing algorithms, improved upon the optimal QN algorithm , BQN algorithm is put forward which introduces the concept of block mean square error(BMSE). In optical lab this algorithm is tested to be satisfying.Two wavelet transform algorithms are advanced for the two schemes respectively: Mallat wavelet transform algorithm and realtime recursive wavelet transform algorithm. It is proved that the algorithms proposed are accurate to detect singularity of fault transient signals in power system by a mass of tests such as EMTP simulation and math model simulation.HOCT based on adaptive theory eliminates linear birefringence and overcomes restraints of optical materials and technics. So current measurement accuracy and stability of long-term running can be improved greatly. HOCT possesses stronger competitive power than other types of OCT.