基于自适应提升小波的电厂汽轮发电机组振动信号消噪

Denoising of vibration signals for turbo-generator units by using adaptive lifting wavelet

采用基于提升格式的自适应小波变换,并根据LMS(leastm ean square)自适应法使伯恩斯坦预测算子P(pred ictor)自适应匹配特定的数据序列。用此方法对电厂汽轮发电机组的振动信号进行消噪处理,并通过实例仿真证明了该方法的可行性和有效性,从而为汽轮发电机组振动信号的故障诊断提供有力依据。

Based on the Least Mean Square （LMS） adaptive criteria and by using the adaptive lifting wavelet transform, the Bernstein filter predictor was enabled to adaptively match the designed signal. The above method was applied to denoise the vibration signal for a turbo-generator unit and is proved practical and effective through practice, providing support for fault diagnosis for vibration signals of the turbo-generator unit.