The Research And Implementation Of System For Seismic Signals Interactive Processing Based On Feature Analysis

In seismic events,waveform data recorded by seismic observation instruments is the main basis of seismological research and the sources of itself character and the information. Seismic waveform data contains a lot of event features information,the time domain and frequency domain features of waveform signal can be extracted from the waveform date,based on the extracted features,can quickly and accurately identify the source type.It can provide a scientific basis for earthquake emergency response,earthquake relief and national defense construction. Institutions of different countries’ efforts for the construction of seismic data processing system, and different levels of progress has been made,but is still much room for improvement in practical application.Especially each extraction in seismic waveform feature,how to quickly extract the waveform time domain and frequency domain waveform through waveform interaction is the main purpose of building the system. To this end,the system design and implementation can greatly simplify the waveform analysis tasks in the research of seismic waveform.This paper comprehensively discusses the interaction which is based on the analysis of the features of seismic signal interactive processing system extract the platform design principle and development process.Based on the analysis of earthquake and explosion data,extract the event feature and the waveform feature that can identify the natural earthquake or artificial explosion, and developed the seismic signal interactive processing system,used for interactive extraction of natural earthquakes and artificial explosion,waveform display mode is analyzed in detail, and the waveform feature extraction algorithm is studied and implemented,fill the software defects and shortcomings of earthquake and explosion wave features.The platform is not only an reflect of seismic data processing that feature extraction technique used in the field of seismic source identification,but also provides user with a more efficient and diversified interactive extraction platform. Specific studies are as follows:(1) At first,the current situation of research background of seismic signal processing software system is analyzed, in the light of the feature of the seismic signal processing software system, considering the insufficient of current waveform processing software function and operation, the seismic signal interactive processing system is developed.The system is a software system that extract the time domain and frequency domain features from the seismic event waveform observation records:Main function modules are divided into the management of seismic events and waveform data, the earthquake wave interaction, waveform in time domain features interactive measurement and frequency domain、time-frequency features interactive extraction,etc. The system achieve analysis of the waveform date in seismic event, read the file event information, device information and observation waveform data,draw the picture of interactive display’s seismic waveform, completed the wave contrast visual function of the natural earthquakes and the artificial explosion.For the convenience of date maintaining and updating of the seismic event underlying data and waveform data, the design has realized the event directory management module and waveform date management module.(2) Research and analyze the feature information of seismic wave,It is mainly divided into the time domain features, the frequency domain features and the time frequency features information.On the basis of this, completed the waveform features of the interactive measuring function, thus the user can choose the appropriate band according to the experience,The time domain features of the waveform are extracted by the interactive extracted.Paper mainly extracted the P wave initial motion direction, Ap1/Asm, Apm/Asm and Tme, at the same time interactive extracted the waveform features of the complex、the waveform features of comprehensive ratio and the waveform features of autocorrelation coefficient, interactive selected wavebands to calculate waveform cross-correlation coefficient features and waveform energy ratio features etc,In this paper, the extraction algorithm of the above feature is analyzed and realized.(3) System realize the module of waveform in frequency domain feature interactive extract function, through the interface interactive operation, frequency domain features can be calculated, and based on the the interactive selection of P wave and S wave group,analyze and implement the wave spectrum ratio feature extraction algorithm. At the same time, in order to analyze the frequency domain of digital signal, the time domain sequence should be transformed into equivalent frequency domain, through the time frequency conversion module,the system is transformed from the time domain waveform to the frequency domain,using the linear convolution transform、wavelet transform and Hilbert-Huang transform for time-frequency conversion function implementation. Waveform features of these interactions extracted, can take use of personal experience of users of the system to effectively abandon the abnormal waveform and error waveform in automatic operation,can provide a more reliable and robust waveform features data for source type identification.Finally, the paper discusses the detailed test results of system functional modules and possible error conditions adopted by fault-tolerant design, and shows how to show the timely and reasonable countermeasures after the internal logic functions out of date.