摘要
本文选用磨屑热辐射流温度信号作为辨识信号源,提出了用自聚类神经网络辩识磨削烧伤的策略与方法。实验结果表明:自聚类神经网络可用于辨识磨削烧伤,辨识精度较高,且具有一个重要特点:网络训练是无师自学习的。
This paper selects signal of temperature of thermal radiation flow from ground chips as signal origin for recognition,puts forth tactics and methods for recognizing ginding burn by means of self-clustered nerve network.Result of experiment shows that nerve network of self-clustered type may be applied to recog- nize grinding burn.Precision of recognition is higher.Furthermore,it posseses important features:network training is selfstudied,teaching is unnecessary.
关键词
磨削烧伤
在线辨识
磨屑热辐射流
神经网络
Grinding burn
On-line recognition
Ground chip thermal radiation flow
Nerve network
