计及负频率的极高频信号离散频谱校正新方法

对于机械振动与故障诊断等领域中常见的极高频(接近奈奎斯特频率)信号,传统的离散频谱校正方法存在着较大误差,负频率成分干涉严重是影响其频谱分析精度的重要因素。为提高极高频信号的频谱分析与校正精度,给出了一种计及负频率影响的离散频谱校正新方法。该方法基于Blackman窗,依据离散频谱的周期性并利用局部谱峰附近的3条谱线,建立包含正负频率贡献的离散频谱校正模型,通过对模型的求解获得频率、幅值和相位校正公式。采用频段内扫描的方式对频谱校正公式进行了仿真验证,结果表明所提方法可有效降低负频率成分对极高频信号频谱的干涉影响,提高其频率、幅值和相位校正精度。

New Discrete Spectrum Correction Method with Negative Frequency Contribution for Ultra High Frequency Signals

For the ultra high frequency(adjacent Nyquist frequency) signals that exist in vibration engineering and fault diagnosis domain,traditional discrete spectrum correction methods are not to be adapted to them exactly because of the serious interference caused by the negative frequency components.In order to reduce the influence,a discrete spectrum correction model with both positive and negative frequency component contribution is established for ultra high frequency signals by three frequency lines closing to the peak value according to the spectrum periodicity.The frequency,amplitude and phase correction formula is deduced based on the Blackman window.Simulated gradually increasing real sine signals are adapted to assessment the formulas.Simulation results show that the proposed method is expected to reduce the interference effectively,and can achieve higher accuracy correction results of spectrum parameters for ultra high frequency.