Abstract:This study processed the characteristic parameters extraction of acoustic emission signals through the acoustic emission test of the drawing parts molding state. The collected signals were processed the local wave decomposition to extract the energy value of each IMF (Intrinsic Mode Function) as the initial characteristic parameters, which were optimized by genetic algorithm to create the optimal characteristic parameters. The experimental data between the normal condition and the crack state were computed by the simple Mahalanobis distance method to compare the big or small of the Mathalanobis distance during the two states, and then the state corresponding to the minimum discriminated distance was the state type of the test sample. The research results states that this way can recognize effectively the crack acoustic emission signals of the drawing parts to judge the initial crack state of the drawing parts as well as achieve the characteristic parameters optimization of the acoustic emission and the molding quality state recognition of the metal drawing parts.
骆志高 范祥伟 陈强 . 金属拉深件拉深过程微裂纹AE信号特征参数的优化及状态识别[J]. , 2012, 31(17): 154-158.
LUO Zhi-gao;FAN Xiang-wei;CHEN Qiang. The AE signal characteristic parameters optimization of initial crack and status identification for drawing parts. , 2012, 31(17): 154-158.