关键词: K-L信息量, 滚动轴承, 内蕴分量, 异音, IEC61373 机车车辆 随机振动 加速度功率谱密度 数据处理

Abstract: Aiming at the non-stationary time-varying characteristics of the vibration signal from rolling bearing, the method based on intrinsic mode weights modeling is presented to analyze noise from the rolling bearing. Fist tested signals are decomposed to multiple intrinsic mode weights by local-wave and AR(Autoregressive) modeling for K-L information using their intrinsic mode weights. The model based on K-L information will be modified by comparing with the result of manual experiment until the veracity meets given requirement. Three signals from the rolling bearing with different type of fault are analyzed and compared with normal signal from the good rolling bearing. The results prove that the approach based on intrinsic mode weights modeling is more accurate than other methods in detecting the presence of abnormal noise, and it can reach high accuracy rate set by user in advance, so it is highly practicable in engineering.

Key words: Abnormal noise, Intrinsic mode weights, K-L information, Rolling bearing, IEC61373 Railway vehicle Random vibration Acceleration spectral density Data processing