含噪振动信号中早期碰摩的故障检测研究

 研究了在有噪环境下发动机转子系统早期碰摩故障的检测问题。利用最优参数搜索法改进了独立分量分析( ICA) 算法, 用于求解转子系统振动信号与噪声的盲分离问题。在此基础上, 对分离后的振动信号利用小波包分解进行早期碰摩信号的检测。结果显示最优搜索的ICA 算法运行效率高, 信号分离纯度好, 对振动信号有高效的降噪作用, 并利用小波包分解准确地检测出振动信号中的碰摩信息, 其效果优于小波分解法。信噪分离与小波包分解相结合有望用于工程实践中的早期碰摩故障检测和诊断。

Early Rub-Impact Fault Detection in Noise Circumstances

The early rub impact fault detection of an engine rotor system in noise circumstances is researched. An optimum parameter searching method that can improve the ICA algorithm is proposed and applied to the separation between vibra signal and noise. After that, the separated vibra signal is analyzed on the wavelet packet decomposition to obtain the early rub impact fault features. Simulation results show that the modified ICA algorithm is effective on denoise of vibra signal, and wavelet packet decomposition may provide richer early rub impact fault information than wavelet decomposition. The proposed approach may be a desired solution to fault detection and diagnosis of early rub impact in engineering practice.