摘要
轴承加速度振动信号具有广谱性,利用广谱信号的经验模态分解性质,用本征模态函数从检测信号中分离周期、非周期和随机性的冲击成分,然后对其进行短时Fourier变换得到滚动轴承异常声时频域图形的两种表达形式。此方法的优点在于揭示了轴承在运行中瞬间冲击所发生的时刻、冲击的幅度和频率,时频域图形反映轴承异常声产生的本质,且直观易懂。
With a wide band spectrum in a bearing vibration acceleration signal,the IMF can be used to separate the impulse components with periodicity,aperiodicity as well as random from the measurement signal on the property of EMD.Through STFT processing on IMFs,both classes of the expression of time-frequency domain for the ball bearing abnormal noise are obtained.The processing has an advantage in describing the impact magnitude,frequency and time scale while bearing running.The corresponding diagram of time-frequency domain reflects the hypostasis of the bearing abnormal noise;moreover it is intuitionistic and easy to understand.
出处
《轴承》
北大核心
2010年第6期51-54,共4页
Bearing
