Classification of non-natural seismic events is one of the daily tasks of the seis-mic monitoring business.This research is mainly aimed at the classification of earth-quakes,explosions and mining-induced earthquakes.On the basis of the research results of seismic wave data processing,feature extraction and artificial intelligence comprehensive classification,a seismic event recognition software(SERS)with good portability,expansi-bility and independent intellectual property rights is developed based on the Qt develop-ment framework and combined with Python,Matlab and other programming languages.The software can be deployed on different operating systems and consists of seven mod-ules:seismic data import module,data processing module,feature extraction module,comprehensive classification module,feature analysis module,yield estimation module,and result analysis module.The software integrates various time-frequency feature extrac-tion techniques and artificial intelligence classification methods,to form a comprehensive process for classifying the seismic events.The built-in classification models in the software have an accuracy rate exceeding 90％and a wide range of applications.It has been applied in a number of earthquake monitoring departments,achieved favorable outcomes and en-hanced the capability for rapid analysis of non-natural earthquakes.
Classification of non-natural seismic eventsQt development frameworkFeature extractionArtificial intelligence method
NETL
NSTL
万方数据
