基于变分贝叶斯理论的机械故障源盲分离方法研究

Blind separation of the mechanical fault sources based on variational bayesian theory

未知噪声环境下机械源信号盲分离方法由于忽略噪声影响往往得到很差的分离效果。针对此问题,提出了一种基于变分贝叶斯独立分量分析的机械故障源分离方法,该方法与传统的机械源分离方法相比,具有以下独特特点,即不需要将未知噪声看成一种独立源,也不需要进行消噪预处理,可直接对噪声干扰的机械源信号进行有效分离。仿真研究表明,提出的方法优于传统的机械源分离方法,分离误差大幅度降低。实验结果也验证了所提出方法的有效性。

Considering the deficiency of the traditional blind seperation methods under the condition of unknown noises, a new separation method for machine fault sources based on variational Bayesian independent component analysis was proposed. Compared with the traditional method, the proposed method has such unique characteristics: it is unnecessary to regard the unknown noise as an independent source, the denoising preprocessing can be cancelled, and the mechanical sources under the noisy environment can be directly separated. The simulation results show that the proposed method is superior to the traditional method, with the separation error greatly reduced. The experiment results also verify the validity of the proposed method.