【作者】 刘国光； 程青蟾； 李燮里； 张月兰；

【Author】 Liu Guoguang Cheng Qingchan Li Xieli Zhang Yuelan (East China University of Science and Technology, Shanghai 200237, China)

【机构】 华东理工大学；

【摘要】 在声发射BP神经网络的基础上,通过添加模糊输入层和模糊输出层,构筑成声发射模糊神经网络。对原先的声发射信号进行模式识别,既提高了神经网络训练的收敛精度,又改善了收敛速度和稳定性。从而进一步提高了声发射信号模式识别的实用价值。更多还原

【Abstract】 On the basis of acoustic emission BP neural network, by means of adding fuzzy input layer and fuzzy output layer, an acoustic emission fuzzy neural network was constructed. With this fuzzy neural network, the pattern recognition of acoustic emission was completed. The training convergence precision of neural network was greatly raised, the convergence speed and its stability were also improved. As the result of this model, the practice value of pattern recognition of acoustic emission signals was greatly increased.更多还原