摘要
在大型回转机械状态监测与诊断中,常利用两路相互垂直的振动信号合成轴心轨迹来监测其运行状态和故障类型,不易获得清晰的特征。本文利用数学形态学的一些基本功能还原出干净的轴心轨迹,为研究轴心轨迹图像的特征提取和识别提供可靠的数据,使计算机能自动进行旋转机械故障的诊断。
In state monitoring and fault diagnosis of large rotary machinery, shaft orbit is often obtained from two mutual vertical singles to monitor its running condition and fault types, but it is difficult to get its clear characteristics. In order to study shaft orbits, some basic functions of mathematical morphology are used to re- move noise and provide reliable data of feature extraction and recognition. Then computers can diagnose faults for rotary machinery automatically.
出处
《苏州大学学报（工科版）》
CAS
2007年第4期39-41,共3页
Journal of Soochow University Engineering Science Edition (Bimonthly)
基金
江苏省高校自然科学基金资助项目(编号03KJD46019399)
关键词
数学形态学
滤波
轴心轨迹
mathematical morphology
filter
shaft orbit
