CAJ | 학술논문

探究了一种基于奇异差分谱的信号升维途径，并将其和平稳子空间分析结合提出基于奇异差分谱和平稳子空间分析（Stationary Subspace Analysis，SSA）的滚动轴承故障诊断方法。该方法首先对滚动轴承原始故障振动信号进行奇异差分谱分析，根据奇异差分谱的峰值分布，确定不同的有用分量个数进行信号重构实现信号升维，然后利用平稳子空间分析将高维信号分解为平稳源信号和非平稳源信号，最后通过对峭度值最大的非平稳源信号进行包络谱分析得到滚动轴承故障特征频率。仿真信号和滚动轴承早期故障信号分析表明该方法可以实现滚动轴承欠定故障信号的盲分离，验证了该方法的有效性。
탐구료일충기우기이차분보적신호승유도경，병장기화평은자공간분석결합제출기우기이차분보화평은자공간분석（Stationary Subspace Analysis，SSA）적곤동축승고장진단방법。해방법수선대곤동축승원시고장진동신호진행기이차분보분석，근거기이차분보적봉치분포，학정불동적유용분량개수진행신호중구실현신호승유，연후이용평은자공간분석장고유신호분해위평은원신호화비평은원신호，최후통과대초도치최대적비평은원신호진행포락보분석득도곤동축승고장특정빈솔。방진신호화곤동축승조기고장신호분석표명해방법가이실현곤동축승흠정고장신호적맹분리，험증료해방법적유효성。
A thorough inquiry was made for an approach to increase signal dimensions based on difference spectrum of singular value and a new fault diagnosis method for rolling bearings based on difference spectrum of singular value and stationary subspace analysis (SSA)was proposed.The first step was to obtain the singular value difference spectrum of a bearing original vibration signal.According to the peak distribution of the singular value difference spectrum,different amounts useful component signals were used to reconstruct new signal matrices,so that signal dimensions increased.Then the multi-dimensional signals were taken as inputs of SSA to obtain stationary source signals and non-stationary source signals.Finally,the selected non-stationary source signal with the maximum kurtosis value was analyzed using the envelope spectrum to extract characteristic frequency of rolling bearing fault.The analysis of simulated signals and early fault signals of rolling bearings showed that the proposed method can realize the blind source separation of under-determined fault signals of rolling bearings,the effectiveness of the method was verified.