研究了基于独立分量分析的航空发动机气路静电监测信号处理方法,针对该方法在使用过程中存在的欠定问题,采用了中值滤波进行预处理,基于经验模态分解的方法构建了参考噪声,进而对信号进行降噪处理。提出了采用信号在低频段和高频段的能量比作为静电监测信号的信噪比,并以改善信噪比作为降噪效果的评价准则,对几种常用的降噪方法和本文的方法进行了对比实验,模拟信号和实测信号的降噪处理的结果表明了该方法的可行性和有效性。
A de-noise method based on ICA(Independent Component Analysis) is proposed.Aiming at the underdetermined problems in the process of using ICA,the median filtering method is employed to preprocess the signal,and then the referenced signal is constructed based on EMD(Empirical Mode Decomposition) for the ICA de-noising processing.The energy ratio in low frequency band and high frequency band is used to denote the SNR(Signal-to-noise Ration) of a signal,therefore,the improved SNR as an evaluation rule,is used to evaluate and compare the de-noising quality of common signal processing methods.The experimental results from simulated signal and experiment signals show the ICA de-noising method proposed in this paper is feasible and effective,also provides a useful reference to research the electrostatic monitoring technology for Aero-engines gas path.