Accurate Detection Method Of Voltage Traveling Wave Based On EMD Decomposition Combined With Wiener Filtering

The physical structure of the traveling wave sensor itself and the interference factors such as high frequency noise,the transient information extracted in the distorted secondary traveling wave will deviate from the real quantity,and the traveling wave technology relies on the key transient information and a slight deviation will lead to the positioning or protection failure.In order to solve the problem of primary and seeondary traveling wave mismatch,the paper proposes the method of voltage traveling wave detection based on Empirical Mode Decomposition(EMD)combined with Wiener filtering,which is expected to improve the accuracy oftraveling wave detectiontechnology.The paper first analyzes the current development status of the existing traveling wave detection and extraction technology,and summarizes the problems existing in its existing methods,According to the existing problems,studying the characteristics of traveling wave signals and the transfer characteristics of transformers,puts forward a sub-band inversion voltage traveling wave detection method;Analyzes the transmission characteristics of the voltage traveling wave sensor based on Rogowski coil principle,compares and the high frequency transient transfer characteristics of the centralized model and the distributed model,Without affecting the accuracy,a centralized model more suitable for inversion method is selected as the high-frequency transient model of the traveling wave sensor,and its transfer function is obtained as the forward model of the voltage traveling wave sensor,the waveform distortion problem in the traveling wave detection process is verified by simulation and field experiment;In order to solve the waveform distortion problem,the paper uses the EMD decomposition method to decompose the detected secondary traveling wave signal into the Intrinsic Mode Function(IMF)component of different frequency bands,and denoising the noisy components,the real frequency domain information of each component is obtained by Fourier transform,and then the forward transfer function of the voltage traveling wave sensor is used as the convolution core of the Wiener filtering inversion algorithm,using the Wiener filtering algorithm to invert each IMF component separately,the inverted components are in the form of frequency domain,the time domain components are obtained by discrete inverse Fourier transform and the real part,the linear superposition and synthesis of the time domain components is the inversion of a traveling wave signal.The simulation results show that under different fault conditions,the method in this paper can truly restore the primary traveling wave signal,effectively solve the problem of transient data missing in the secondary traveling wave part,solve the problem of large positioning error and inactive protection under strong noise interference and weak traveling wave signal,realize accurate detection of time-frequency information of fault transient traveling wave,and improve the traveling wave detection technology.