基于源数估计的船舶振源盲识别研究  

Vibration Source Identification and Diagnosis Based on the Source Number Estimation for Ships

作　　者：王衍学[1] 何正嘉[1] 成玮[1] 陈略[1] 

机构地区：[1]西安交通大学机械工程学院．陕西西安710049

出　　处：《中国舰船研究》2010年第1期1-5,共5页Chinese Journal of Ship Research

基　　金：国家自然科学基金重点项目(50335030)

摘　　要：针对船舶等复杂设备的多传感器监测与诊断过程中振动源识别,提出一种基于源数估计的船舶振源盲识别方法。该方法首先采用盖尔圆理论确定系统振动源数目,运用信号盲源分离算法提取相对独立的主要源信号,进而运用所提出的谱相关系数分析方法确定所提取信号来自哪个设备。仿真分析与实际的某船舶的模型振动实验均证实该方法的有效性。Based on the source number estimation, a vibration source identification method was proposed for the surveillance and diagnosis of the ship. Using this method, source number is firstly estimated based on the Gerschgorin＇s disk theorem which can effectively distinguish the source signals from noise signal. Then ,independent sources are separated by the blind source separation technique ,and the separated signals can locate their individual vibration source by employing the spectrum correlation. Finally, the condition of the equipment in the ship can be identified using the available well-established diagnosis techniques. The validity of the proposed methodology was confirmed on a ship model in this paper.

关 键 词：振源 盲源分离 盖尔圆理论 故障诊断 船舶 

分 类 号：O327[理学—一般力学与力学基础]