摘要
提出了一种利用刀具磨损状态隶属函数来训练BP网络的新方法，该方法较好地解决了在刀具磨损状态中对过渡样本识别率过低的问题。
To solve the low recognition accuracy problem of transient samples in tool wear condition recognition,a new method in which tool wear condition membership functions are used as neural network output target is presented in this paper to train BP network.
出处
《振动．测试与诊断》
EI
CSCD
1998年第2期103-107,共5页
Journal of Vibration,Measurement & Diagnosis
基金
高等学校博士基金!9824818
