自适应时间尺度分解方法及其应用

针对齿轮故障振动信号的非线性、非平稳和多分量的特征,在定义了瞬时频率具有物理意义的本征时间尺度函数(intrinsic time-scale function,ITF)的基础上,结合固有时间尺度分解中基线信号的构造方法,提出自适应时间尺度分解(adaptive time-scale decomposition,ATD)的时频分析方法,该方法可以自适应地将一个复杂信号分解为若干个瞬时频率具有物理意义的本征时间尺度分量之和.仿真分析验证了ATD方法的有效性以及定义本征时间尺度函数方法的合理性.分别将ATD、经验模态分解(EMD)、局部均值分解(LMD)和固有时间尺度分解(ITD)与包络解调分析相结合应用于斜齿轮故障诊断中,实验结果表明:自适应时间尺度分解方法在保证分解结果正确性的前提下,计算效率方面具有明显优势,将该方法与包络解调相结合能够有效提取到齿轮的故障特征.

Adaptive time-scale decomposition and its application to gear fault diagnosis

Aiming at the problem of nonlinear, non-stationary and multi-components of fault gear vibration signal, and with defining the intrinsic time-scale function(ITF) whose instantaneous frequency has real physical significance, a new time-frequency analysis method named adaptive time-scale decomposition(ATD) is proposed based on the construction method of baseline signal of the intrinsic time scale decomposition (ITD). By the ATD method, a complex multi-components signal can be adaptively decomposed into the summation of a number of ITFs whose instantaneous frequencies own physical sense. The simulation results verify the effectiveness of the ATD method and the feasibility of the definition of ITF. Combined with envelop demodulation analysis, the ATD, the empirical mode decomposition(EMD), the local mean decomposition(LMD) and the intrinsic time scale decomposition(ITD) are respectively applied to the gear fault diagnosis. The experimental results show that the ATD method has obvious advantage in computational efficiency with the guarantee of decomposition correctness, and the method can effectively extract gear fault features combining with the envelope demodulation analysis.