机械故障特征信息提取的ICA信息融合方法

发布时间：2018-05-11 00:32

本文选题：独立成分分析 + 信息融合　；参考：《机械科学与技术》2016年07期

【摘要】：峭度和负熵是盲信号独立性的两个自然测度,可以被用来捕捉机械振动信号信息的动态变化特征,并提取机械设备的故障特征信息。峭度和负熵是从两个不同的角度和层面阐释机械设备的故障特征信息,信息量是互补的。若将峭度信息和负熵信息融合,则必然能够更全面、更深刻地来表征机械设备的状态。因此引入信息融合的思想,提出基于ICA信息融合的机械故障特征信息提取方法,综合峭度和负熵信息来提取机械设备的故障特征信息。液压齿轮泵模式识别试验表明,该方法可以应用于机械设备的故障特征信息提取。
[Abstract]:Kurtosis and negative entropy are two natural measures of blind signal independence, which can be used to capture the dynamic characteristics of mechanical vibration signal information and to extract the fault feature information of mechanical equipment. Kurtosis and negative entropy explain the fault feature information of mechanical equipment from two different angles and levels, and the amount of information is complementary. If the kurtosis information and the negative entropy information are fused, the state of mechanical equipment can be represented more comprehensively and profoundly. This paper introduces the idea of information fusion and proposes a method of extracting mechanical fault feature information based on ICA information fusion which combines kurtosis and negative entropy information to extract fault feature information of mechanical equipment. The pattern recognition test of hydraulic gear pump shows that this method can be used to extract fault feature information of mechanical equipment.
【作者单位】：武警警官学院;第二炮兵工程大学;武警工程大学;
【基金】：国家自然科学基金项目(61132008)资助
【分类号】：TP202;TH17

【相似文献】