基于FDM的机械故障信号诊断处理方法研究

Mechanical Fault Signal Diagnosis Method Based on Fourier Decomposition Method

傅里叶分解(FDM)是一种新的自适应时频分析方法,该方法首先定义了瞬时频率具有物理意义的傅里叶固有频带函数(FIBFs),然后在整个傅里叶域自适应地搜寻解析傅里叶固有频带函数(AFIBFs),从而将非平稳非线性信号自适应地分解为若干个傅里叶固有频带函数和一个残余分量。FDM方法基于傅里叶变换,且是一种完备的、正交的、局部的和自适应的信号处理方法,数学理论充分,并且可以直接依据解析傅里叶固有频带函数直接求得时频能量谱,时频分辨率高,优点非常明显。仿真结果表明,与EMD相比,该方法无端点效应,不存在模态混叠效应,且数学理论充分。最后成功地将该方法应用到实际的滚动轴承故障诊断中。

Fourier Decomposition Method(FDM)is a new self adaptive Time-Frequency Analysis method. This method firstly defines the Fourier intrinsic band functions(FIBFs),which has physical significance of the instantaneous frequency,then adaptively search the analytic Fourier intrinsic band functions in the Fourier frequency domain. Thus,the non-stationary and nonlinear signal?is decomposed to a number of Fourier intrinsic band functions and a residual component. Because of the method based on Fourier transform,and is a kind of perfect,orthogonal and partial,adaptive signal processing method,mathematical theory fully,and the time-frequency energy spectrum is obtained directly by analytic FIBFs,the time-frequency resolution is high and the advantage is obvious. The simulation results show that compared with the EMD method,this method has some distinct advantages,such as no endpoint effect,no mode mixing and mathematical theory fully,etc.Finally,the proposed method has been applied in the rolling bearing fault diagnosis.