For monitoring vibration, the signals acquired from sensors are the mixture of vibration signal of machines and the environmental noise. The blind source separation can extract the feature signals of each machine from the mixed signals. A blind source separation algorithm was proposed herein to extract the feature signal in the presence of the environmental noise. The experimental investigation of extraction of the feature signal from mechanical vibration signals mixed with environmental noise was carried out, and the results show that the algorithm can eliminate the influence of environmental noise and extract the weak signals of the machine. 

vibration and noise monitoring, blind source separation, weak feature signal;feature extraction

振动状态监测时，传感器采集的信号是各机械设备信号和环境噪声的混合信号。盲源分离技术可以有效去除环境噪声的干扰并提取出各设备的特征信号。提出噪声背景下机械振动弱特征信号提取的盲源分离算法，并对混有噪声的机械振动信号的特征进行试验研究，结果表明：该算法不仅可以去除环境噪声的干扰，而且可以实现对能量较弱的特征信号的提取。

振动噪声监测, 盲源分离, 弱特征信号, 特征提取