CAJ | 학술논문

为从复杂的轴承振动信号中获取故障信息，提出基于分量筛选奇异值分解的特征提取方法。阐述奇异值分解原理，构造轴承故障信号 Hankel 矩阵，利用互相关系数准则对奇异值分解处理的分量信号进行筛选，对筛选的分量信号进行重构以提取轴承故障特征频率。与传统方法相比，该方法的仿真与实际轴承信号处理结果均具优越性、有效性。
위종복잡적축승진동신호중획취고장신식，제출기우분량사선기이치분해적특정제취방법。천술기이치분해원리，구조축승고장신호 Hankel 구진，이용호상관계수준칙대기이치분해처리적분량신호진행사선，대사선적분량신호진행중구이제취축승고장특정빈솔。여전통방법상비，해방법적방진여실제축승신호처리결과균구우월성、유효성。
In order to extract fault informations from complicated bearing vibration signals,a fault diagnosis method based on component screening singular value decomposition (CSSVD)was proposed.The principle of SVD was explained and a Hankel matrix was constructed for the SVD of bearing vibration signals.To choose the component signals after SVD, the criterion of correlation coefficient was applied.Then the component signals were reconstructed and fault feature frequencies were extracted.Compared with the traditional method,the effectiveness and advantage of the proposed method were demonstrated by analyzing several simulated signals and actual bearing signals.