Application of translation invariant wavelet de-nosing to axle of railway vehicles fault diagnosis online

The wavelet threshold method may produce fesndo-Gibbs phenomenon on the singularity points of signal. The transiation invariant wavelet is an improve method based on that algorithm. Comparing with threshold method, a translation invariance method can suppress Pesudo-Gibbs phenomenon and minish RMSE between the original signal and estimated one. At the same time, SNR of estimated signal can also be improved. This method is used to de-noise acoustic emission signals and compared with threshold method. The acoustic emission energy method is adopted to identify the fatigue cracks of axle of railway vehicle. The results demonstrate that this method can effectively eliminate noise, extract characteristic information of acoustic emission signals, is effective for online detectilon of the fault.