摘要
通过对声发射传感器采集的刀具磨损状态信号进行分析,提取出反映刀具磨损状态的特征向量MFCC系数及差分系数,然后利用HMM进行信号处理。建立了检测镗刀刀具状态的监测系统。实验结果表明,该监测系统在刀具的正常磨损阶段,可以实现刀具大致磨损量的预报;在刀具破损或损坏情况下,能够及时监测和预报刀具损坏状态。这种监测方法能够进行实时在线监测,为刀具的磨损监测提供了一条切实可行的途径。
Signals of acoustic emission sensors about tool wear are analyzed,feature vectors reflecting tool wear MFCC coefficient and differential coefficient are extracted.Signals are processed by HMM.A test boring tool condition monitoring system is established.Experimental results show that the tool monitoring system can roughly forecast tool wear in the normal wear stage,and also can timely monitor and forecast tool damage to the state in the case of tool breakage or damage.This monitoring method can be real-time online monitoring of tool wear.It provides a practical way for tool condition monitoring.
出处
《制造技术与机床》
CSCD
北大核心
2011年第10期122-125,共4页
Manufacturing Technology & Machine Tool
基金
湖南省教育厅资助科研项目(项目编号09C1307)
关键词
镗刀磨损监测
梅尔系数
隐马尔可夫模型
Boring Tool Wear Monitoring
Mel Frequency Cepstrum Coefficient
Hidden Markov Model
