摘要
以齿轮振动信号的时城特征作为神经网络输入、齿轮的主要故障形式作为神经网络输出，利用经BP算法训练后的该网络对齿轮故障进行诊断，取得了较好的效果．
In this paper, using characteristics value of vibration signal of gear in therealm time as input of neural networks and using main fault type of gear as outputof neural networks, faults of gear are diagnosed by using neural networks that hasbeen trained with algorism BP and achieve fairly good results.
出处
《南方冶金学院学报》
1998年第3期201-206,共6页
Journal of Southern Institute of Metallurgy
