中文摘要：

针对轮毂单元故障信号中的脉冲信号往往被背景噪声淹没的问题，将数学形态学滤波技术应用到一维振动信号的消噪中，并与3σ规则结合，提出了基于数学形态学运算和软阈值的振动信号消噪方法。首先，采用形态滤波对染噪的轮毂单元信号进行了过滤，并提取了峰谷信号；然后，采用3σ规则对峰谷信号进行了阈值处理，并将形态滤波结果与阈值处理后的峰谷信号相加，作为最终消噪结果；最后，对其进行了频谱分析以提取特征，并在Matlab中对该算法仿真试验进行了有效性评价以及轮毂单元振动信号的消噪试验。研究结果表明，该算法不仅运算简便，并且在最大限度抑制噪声的同时保留了绝大部分的有用信号，取得了较好的消噪效果，故障信号识别率提高了20％左右。

英文摘要：

Aiming at improving the efficiency of signal de-noising, the mathematical morphology was investigated and it was combined with 3σ rule. An algorithm of hub unit vibration signal de-noising based on morphological operations and soft threshold was put forward. Firstly, morphology filtering was used in hub unit vibration signal with noise and extract peak-valley signal. Secondly, for threshold processing, 3σ rule was applied to the peak-valley signal,then the processed peak-valley signals were combined with the results of the operation of morphology filtering, as the results of the de-noising. Finally, features were extracted from the results through anglicizing the spectrum. This algorithm was verified by using the simulation data and real signal from the hub unit test. The experimental results indicate that the algorithm is not simple but it can reserve most of the useful signals while de-noising to the maximum extent, and the rate of fault signal recognition increases by 20%.