旋转机械故障信号的去噪源分析

旋转机械振动信号是其故障特征识别与诊断的重要信息来源，将应用统计特征来分解混合信号的去噪源分离(DSS)引入到旋转机械故障诊断中。研究DSS基本理论及其正切去噪函数，并进行模拟信号分离，其分离后的性能指标及与源信号相似系数均优于盲源分离；并将DSS应用于某燃气轮机的实测故障信号分析，诊断出转子发生不平衡及异频伪共振现象，表明该方法在旋转机械故障诊断中的有效性，为机械设备的状态监测和故障诊断提供新的思路和方法。

Fault Diagnosis of Rotating Machine Based on Denoising Source Separation

The rotating machinery vibration signal is the important information source of the fault characteristics identification and diagnosis. The denoising source separation (DSS) technology which separates the mixed signals by the statistical characteristics is used in the fault diagnosis. First the basic theory of DSS and its tangent function are studied, and then the analogous signals are separated. The results show that the performance index and correlation coefficients of DSS are better than those of blind source separation. Applying the DSS method to the gas turbine fault diagnosis, the unbalance and pseudo resonance phenomenon have been diagnosed through the measured fault signals in the rotor. This research work shows that the DSS method is efficient in analyzing the fault diagnosis and it provides new ideas and methods for condition monitoring and fault diagnosis of rotating machine.