HHT在爆裂噪声检测中的应用研究

Research on the application of HHT in burst noise detection

该文应用一种新型的非线性非稳态信号处理方法——希尔伯特黄变换(hibert huang tramsform,HHT)进行逆变器中低频噪声-爆裂噪声的检测与定位,该方法利用经验模态分解(empirical mode decomposition,EMD)将待测信号分解为各基本模态分量(intrinsic mode function,IMF),然后对所得IMF进行自适应阈值去噪。经希尔伯特变换(hilbert transform,HT)后,其瞬时振幅与瞬时频率即可清晰表现出爆裂噪声特点与准确突变位置与时长。通过与小波去噪和小波模极大值去噪检测进行对比分析可得,该方法可以同时从时频两方面对信号进行分析,能够实现对故障信号的准确检测与定位。

HHT（ hibert-huang tromsform） method is used in this paper for low frequency noise in the inverter circuit-burst noise detection and location, the method using empirical mode decomposition（ EMD） decomposed signal into the basic modal component under test the intrinsic mode function（ IMF）, and then use adaptive threshold de-noising to the IMF. By the hilbert transform（ HT）, the instantaneous amplitude and instantaneous frequency can clear performance burst noise characteristics and mutation position and length accurately. With the wavelet denoising and wavelet modulus maxima de-noising detection were analyzed. The simulation analysis show that, the method can simultaneously from two aspects of time and frequency analysis of signals,can realize the accurate detection and location of the fault signal.