Feature extraction and classification of mine microseism and blast based on EMD-SVD
SHANG Xue-yi,LI Xi-bing,PENG Kang,DONG Long-jun,WANG Ze-wei
DOI: https://doi.org/10.11779/CJGE201610014
2016-01-01
Abstract:To solve the difficult problem in identifying rock mass microseism and blasting vibration signals, a method for feature extraction and classificationis proposed based ontheempirical mode decomposition(EMD) and singular value decomposition(SVD). Firstly, the mine signalsare decomposed by EMD, and the IMF1 to IMF6 selected by correlation coefficients and variance contribution ratiosare the main intrinsic mode functions(IMFs). Then the SVDis used to obtain singular valuessi(i=1,2,L,6) of feature vector matrix constructed of the main IMFs.Furthermore, thesupport vector machine(SVM)is adopted to train, classify and recognizethesignals of Yongshabamine. The results show that there arelarge differences of singular values1s,s2 and3s between microseisms and blasts, and the best pattern recognitionis obtained when1s is 7.5 with an accuracy rate of 88.25%. In addition, the SVM method with an accuracy rate of 93%is better than the BP neural network method, Bayes method and boundary value method. In conclusion, the proposed method provides a new way forthefeature extraction and classification of mine microseism and blast.