CAJ | 학술논문

为充分利用旋转机械振动信号的多维图形信息,以提高故障诊断准确率,介绍了希尔伯特包络和双谱概念,由此引入振动信号的包络双谱图,提出了利用双谱图矩来表征轴承运行状态.对比分析了其他图像特征提取技术,实验验证了基于振动谱图矩特征识别故障诊断方法的可行性.
위충분이용선전궤계진동신호적다유도형신식,이제고고장진단준학솔,개소료희이백특포락화쌍보개념,유차인입진동신호적포락쌍보도,제출료이용쌍보도구래표정축승운행상태.대비분석료기타도상특정제취기술,실험험증료기우진동보도구특정식별고장진단방법적가행성.
In order to fully utilize multi-dimensions image information of vibration signals from rotating machinery,for improving fault diagnosis accuracy,the concept of Hilbert envelope and bi-spectrum were introduced. On this basis,the bi-spectrum of vibration signal envelope was made,and using the moments statistics of bi-spectrum to reveal the operating condition of rolling bearings was pro-posed. Comparative analysis of other image feature extraction techniques was made. The experiments validate that the diagnosis method of fault pattern recognition through moments features is feasible.