基于VMD时频分析方法的颤振试飞应用研究

目的 针对颤振信号信噪比低、模态密集、非平稳的特点，发展一种能更好地应用于实际颤振试飞信号的时频分析方法。方法 采用近年来新发展的变分模态分解(Variational Mode Decomposition, VMD)方法把复杂信号分解为单一组分的本征模函数(Intrinsic Mode Function, IMF)，再对其进行希尔伯特变换，求得每一阶本征模函数的瞬时幅值和瞬时频率，从而进行信号的时频分析。结果 仿真算例验证了基于VMD的时频分析方法在展现非平稳信号时频特征方面的有效性。将该方法应用于实际颤振试飞的异常振动数据，可得到准确清晰的时频图，很好地展示出外挂物的极限坏振荡过程。结论 该方法可作为颤振试飞信号时频分析的有力工具，有助于分析异常数据，确定故障来源，为后续问题解决提供支撑。

Application of Flutter Flight Test Based on Time-frequency Analysis Method

The work aims to develop a time-frequency analysis method which can be better applied in practical flutter flight test in view of the flutter signal characteristics of low SNR, dense mode and non-stability. Variational mode decomposition (VMD) developed in recent years was used to decompose the complex signal into some intrinsic mode functions (IMF) of single component. The instantaneous frequency and amplitude were calculated by the Hilbert Transform of IMF, and then the time-frequency analysis for signals could be conducted. The effectiveness of the time-frequency analysis method based on VMD in expressing the non-stationary characteristics of signal was validated by simulation example. By applying the method to the abnormal vibration data of actual flutter flight test, accurate and clear time-frequency diagram could be obtained, which could well show the limit cycle oscillation process of the stores. The method can be used as a powerful tool for time-frequency analysis of flutter flight test signals, which is helpful to analyze abnormal data, identify fault sources, and provide support for solving subsequent problems.