CAJ | 학술논문

针对离散余弦变换（Discrete Cosine Transform，DCT）阈值降噪法存在的问题，提出一种自适应的余弦相邻系数（Cosine Neighboring Coefficients，CNC）降噪法，仿真信号分析结果表明CNC降噪法具有更好的降噪性能。将该方法与振动信号快速预白化（Fast Pre－Whitening，FastPW）技术相结合形成了一种液压泵振动信号预处理新方法：首先采用FastPW算法对液压泵振动信号预白化，去除信号中的谐波分量，得到仅包含冲击分量和白噪声的预白化信号；然后采用CNC降噪法对预白化信号进行自适应降噪。实测信号分析结果表明提出的方法能够很好地消除液压泵振动信号中的谐波分量和噪声成分，且能有效保留信号中的故障分量。
침대리산여현변환（Discrete Cosine Transform，DCT）역치강조법존재적문제，제출일충자괄응적여현상린계수（Cosine Neighboring Coefficients，CNC）강조법，방진신호분석결과표명CNC강조법구유경호적강조성능。장해방법여진동신호쾌속예백화（Fast Pre－Whitening，FastPW）기술상결합형성료일충액압빙진동신호예처리신방법：수선채용FastPW산법대액압빙진동신호예백화，거제신호중적해파분량，득도부포함충격분량화백조성적예백화신호；연후채용CNC강조법대예백화신호진행자괄응강조。실측신호분석결과표명제출적방법능구흔호지소제액압빙진동신호중적해파분량화조성성분，차능유효보류신호중적고장분량。
Aiming at existing problems of the discrete cosine transformation (DCT)threshold de-noising method, an adaptive de-noising method named cosine neighboring coefficients (CNC )was put forward.Some simulated signals were de-noised with CNC, the results demonstrated that the proposed method has a more excellent de-noising performance .A novel method for preprocessing of hydraulic pump vibration signals was proposed based on the combination of fast pre-whitening (FastPW)and cosine neighboring coefficients (CNC )de-noising method.Firstly,the hydraulic pump vibration signals were pre-whitened with Fast PW,the harmonic components of the original signals were removed, the pre-whitened signals only contain shock components and noise.Then,the pre-whitened signals were de-noised with CNC.The analysis results of practical signals showed that the propoed method can be used to remove harmonic components and noise of hydraulic pump vibration signals,and to keep their fault components effectively.