数学形态学在旋转机械振动信号处理中的应用  

Mathematical Morphology and Its Application in Processing Rotating Machinery Vibration Signals

作　　者：唐贵基[1] 王维珍[1] 胡爱军[1] 张文斌[1] 

机构地区：[1]华北电力大学机械工程学院,保定071003

出　　处：《汽轮机技术》2005年第4期271-272,318,共3页Turbine Technology

基　　金：华北电力大学校内博士学位教师基金资助项目;项目编号:92104392

摘　　要：信号处理技术是旋转机械故障诊断的核心。数学形态学是一种信号处理方法,介绍了数学形态学滤波的原理,提出一种基于数学形态学的方法来处理现场采集的旋转机械振动信号,以消除噪声污染和直流偏移等问题。该方法通过移动结构元素对采样信号进行形态处理。仿真实验证明该方法切实可行。该方法计算速度快,易于实现,具有较好的适用价值。The core of rotating machinery fault diagnosis technology is signal processing. In this paper, the principle of mathematical morphology is briefly introduced. A new signal processing method based on mathematical filter is presented.The method can deal with the vibration signal of rotating machinery, in order to remove noise and DC offset. With moving structure element, the signal can be detected and smoothed. Simulation results show that this method is practicable. The method possesses following advantages： high calculation speed, easy implementation and better use value.

关 键 词：数学形态学 振动信号 旋转机械 滤波器 

分 类 号：TH17[机械工程—机械制造及自动化]