摘要
为保证在铣刀的磨损监控中铣刀状态分类的可靠性 ,针对小铣刀磨损监控的特点 ,合理选择信号特征 ,给出了一种根据模式可分性测度大小进行特征优选的方法 .实验证明 ,经过本方法优选出的特征所组成的特征向量 。
Develops a method for feature selection based on the differences in the measurement of separability of features according to the peculiarity of tool wear monitoring for small sized milling cutters. Experiments showed that characteristic vectors comprising the acoustic emission features optimized by this method can be effectively employed in the pattern recognition for tool wear monitoring.
出处
《北京理工大学学报》
EI
CAS
CSCD
北大核心
2001年第2期180-184,共5页
Transactions of Beijing Institute of Technology
基金
部级基金资助项目
关键词
铣刀
磨损
声发射
特征优选
模式识别
可分性判据
状态监测
milling tool wear
acoustic emission
feature selection
pattern recognition
separability criterion
measurement
