Intelligent Suppression Algorithm For Horizontal Line Electromagnetic Disturbance Signal In Extremely Low-frequency Data

The extremely low frequency(ELF)data recorded by the seismic station of the China Seismological Bureau Network Center contains seismic signals and various interference signals,and one of the more common interference signals is represented as a series of horizontal lines in the time-frequency diagram(TFD).Unlike conventional power line harmonic radiation which has a fixed frequency range,horizontal line electromagnetic interference signals(HLED)exist randomly in different time and frequency domains.It is a great challenge to manually identify and suppress horizontal line electromagnetic interference signals from the huge amount of extremely low frequency data.To overcome these challenges,we propose an intelligent algorithm capable of automatically identifying and suppressing horizontal line electromagnetic disturbance signals by using computer vision techniques and signal processing methods.Firstly,a short-time Fourier transform is applied to the extremely low frequency data to obtain the time-frequency spectrogram(TFS);secondly,an image processing method is used to process the TFS to enhance the features of the horizontal line electromagnetic disturbance signals;again,a probabilistic Hough transform and a frequency range coordinate axis projection are used to identify the horizontal line electromagnetic disturbance signals(i.e.,the start and end times and upper and lower frequency limits of each horizontal line electromagnetic disturbance signal are obtained).Finally,the Fourier transform and Fourier inverse transform are used to process the experimental data and suppress the horizontal line electromagnetic disturbance signals in the experimental data.In order to verify the method proposed in this paper,experiments are conducted by simulated and measured data sets,where each TFS contains 15-25 horizontal line electromagnetic disturbance signals.In order to investigate the influence of the EM disturbance signal power on the horizontal line EM disturbance intelligent suppression algorithm,this paper constructs four types of simulation data sets with signal-to-noise ratios of 10 d B,20 d B,30 d B and 40 d B based on the simulation data sets,and the experimental results of the horizontal line EM disturbance intelligent suppression algorithm in these four types of simulation data sets are counted respectively.The experimental results show that the RMSE of the signal with the horizontal line electromagnetic disturbance removed manually and the measured signal processed by the intelligent suppression algorithm is 0.31.The algorithm can effectively remove the horizontal line electromagnetic disturbance signal and can greatly improve the seismic data quality.