基于最大熵条件下机械噪声盲源分离方法的研究

Study on Mechanical Noise Separation based on Maximum Entropy of Blind Source

盲源分离方法是近年来出现的一种先进的信号处理方法。本文基于对最大熵(ME)的原理,推导了基于自然梯度法的盲信号处理的计算方法,并通过实例说明本方法对噪声源分离是可能的,为机械设备的故障诊断和监察提供了一个新的方法。

Method for Blind Separation of sources is the Modem Processing Method for Signal. Based on the principle of the Maximum Entropy natural gradient is proposed, A Simulation example proves the feasibility of this method for noise separation, provides a new method for Monitoring and Diagnosis of mechanical equipment.