后非线性马尔科夫算法应用于振动信号提取

Post Nonlinear Markov Algorithm Applied to Vibration Signal Extraction

实际环境中通过传感器检测到的设备状态信号往往是非线性混合信号;而设备状态信号是设备故障诊断的基础,因此从混合信号中分离出设备状态信号极其重要。现有线性独立分量分析方法分离效果并不理想,对此提出将后非线性马尔科夫盲源分离算法应用于设备状态信号提取。为验证算法有效性,将直升机齿轮箱振动信号的非线性混合信号进行分离实验。实验结果表明算法能有效分离出轴承故障振动信号,为进一步提高故障诊断准确性和方便性提供了帮助。

In the physical environment through sensors detect the device status signals often are non-linear mixed-signal, and the device status signals are the basis for fault diagnosis, therefore separate from the mixed-signal device status signals are extremely important. Existing linear independent component analysis separation effect are not ideal, them the post non-linear Markov blind source separation algorithm is applied to the device status signals extraction. In order to verify the effectiveness of the proposed algorithm helicopter gearbox vibration signals non-lin-ear mixed-signal is used for experiment separation. The experimental results show that the proposed algorithm can effectively separate the bearing fault vibration signals, and provide help to the fault diagnosis more accuracy and convenience in the further.