基于ITD改进算法和关联维数的转子故障诊断方法

摘要

提出了基于ITD(Intrinsic Scale-Time Decomposition)改进算法和关联维数的转子故障诊断方法.首先用ITD改进算法对转子在不同状态下的振动信号进行分解,根据分量特征,把包含转子主要故障信息的前若干分量进行组合,并以此作为转子原始振动信号降噪后的信号；然后计算降噪后信号的关联维数.对转子系统故障振动信号的分析结果表明,相对于直接采用原始信号计算得到的关联维数,采用ITD改进算法降噪后得到的关联维数更具有区分性,可以用来有效地识别转子的工作状态和故障类型.%A rotor fault diagnosis method based on ITD (Intrinsic Scale-Time Decomposition, ITD) improved algorithm and correlation dimension was proposed here. Firstly, by using improved method based on ITD, the fault vibration signals of a rotor system in different states could be decomposed into a set of components and the first several components were taken as the denoised signals from the rotor original vibration signals according to the characters of the components. Secondly, the correlation dimension of the denoised signals was calculated. The analysis results of test data showed that the correlation dimension of the vibration signals denoised with ITD improved algorithm can be used to identify the working states and fault types of a rotor system more effectively, compared with that of the original signals.