Font Size: a A A

Design And Implementation Of A Detection System Based On Infrasound Wave

Posted on:2022-07-23Degree:MasterType:Thesis
Country:ChinaCandidate:Y L YangFull Text:PDF
GTID:2518306341454254Subject:Software engineering
Abstract/Summary:PDF Full Text Request
Many natural phenomena and human activities can produce infrasound wave,such as lightning,typhoon,chemical explosion,etc.Infrasound wave has the characteristics of long spread and not easy to attenuate.And infrasound monitoring is one of the important and effective ways of the global nuclear prohibition organization's detection work.Therefore,it is of great significance to research the signal feature of infrasound wave and the event classification based on it.In this thesis,I research the abnormal infrasound wave by the way that combining signal processing technology and information entropy theory.The signal processing technologies used are the Fourier Transform,Wavelet Transform and Hilbert-Huang Transform.Then,calculating the time-domain feature,frequency-domain feature and time-frequency-domain feature of the infrasound signal.Meanwhile,in order to verify the effectiveness of different feature extraction algorithms in infrasound classification,this thesis uses SVM to to do a comparative experiment.In terms of infrasound classification,using LSTM and GRU to explore the feasibility of the recurrent neural network model.The results of these experiments show that in the SVM algorithm based on feature extraction,Hilbert-Huang Transform is the best;in the classification experiments based on neural network,LSTM is better than GRU.In order to make full use of data,this thesis explores the application of data fusion technology in this project.In this thesis,the characteristics of different levels of data fusion are compared,and the adaptive weighted fusion algorithm for multi-sensor infrasound data is used to realize the data-level fusion.In addition,this thesis makes a multi-sensor triangle array simulation orientation experiment based on the time delay location algorithm.Based on the research results,this thesis divides the system into four modules after demand analysis:system management module,event analysis module,data fusion module,and event display module.In the design of system,business process design,interface design,database design were given.Then the coding development of each function point is completed.Finally,the system realizes the functions of file upload,analysis and data storage.And at the same time,in order to ensure the safety of the system in use,it integrates the function of authority management.
Keywords/Search Tags:infrasound wave, feature extraction, event classification, data fusion
PDF Full Text Request
Related items