Abstract:The paper extracts the ten acoustic emission characteristic parameters with the neural BP network of three layers. When the hidden neurons are thirteen decided by comparing the practice error and times of BP neural network, the approach effect of BP neural network is best and the network error is smallest. According to compute the crack signal sensitivity magnitude that the AE parameter represents, the paper cancels the AE parameter gradually to reduce the dimensions of input signal. Finally, the energy rate, mean signal level, amplitude, relative reach time, during time and rising counter are the acoustic emission characteristic parameters to identify the crack of drawing parts. The research has theoretical and practical significance to the online monitoring of cracks in the drawing parts.
骆志高;张保刚;何鑫 . 基于BP神经网络的金属拉深件裂纹在线监测[J]. , 2012, 31(10): 102-105.
LUO Zhi-gao;ZHANG Bao-gang;HE Xin. on-line monitoring of metal deep drawing parts crack based on the BP neural network. , 2012, 31(10): 102-105.