摘要：The paper discusses the processing and analysis methods for the geoacoustic and electromagnetic emission pulse signals recorded for more than 20 years at the IKIR FEB RAS geodynamic proving ground (Kamchatka Peninsula). The methods for pulse detection, waveform reconstruction, pulse time-frequency analysis using adaptive sparse approximation, structural description of pulse waveforms and pulse classification are proposed. To detect pulses, the adaptive threshold scheme is used. It adjusts to the noise level of a processed signal. To analyze time-frequency structure of the pulses, the adaptive matching pursuit algorithm is used. To identify pulse waveform, the structural description method is proposed. It encodes pulses with special image matrices. The method of the identified pulses classification is considered. Since the methods for pulse structure analysis are sensitive to noise and distortions, the authors propose the method for pulse waveform reconstruction based on wavelet filtering. The geophysical signal information features determined during the analysis can be used to search for anomalies in the data, and then establish a relationship between these anomalies and deformation process dynamics, in particular, with earthquake development processes.
