应用盲源分离法消除旋转机械振动信号噪音与干扰

工业过程状态检测和故障诊断的信号,会受到相邻设备振动信号的干扰及噪音信号污染,影响对机器运行状态的准确判断和诊断.提出扩展干扰或噪音为一个源向量方法,利用盲源分离技术,研究机械源为线性卷积混叠模型,建立盲源分离技术消除振动干扰或噪音的算法.应用在风机轴振动信号试验表明,经盲源分离处理风机转子失衡引起轴振动信号的功率谱,优于未经处理的信号功率谱,证明了盲源分离技术消除机械设备间振动干扰和噪音信号的有效性.

Application Blind Source Separation Approach for Removal Interference and De-noise within Vibration Signal of Rotor

In the industry process,the signal of condition monitor system and fault diagnosis was corrupted by the noise from neighbor machines,which is influenced the accurately of state estimation and fault diagnosis for machine.The method was proposed that the noise and interference was extended as a source vector.The blind source separation(BSS) technique was employed.An algorithm was researched to removal interference and de-noise based on blind source separation,and is used in experiment of fault diagnosis for vibration signal of shaft.The result shown: the power spectrum that the signal of vibration was transacted is better than that of which isn't done with.The validity of BSS,in interference removal and de-noise,is verified by computer computing.