基于卷积型小波包变换的船用水位调节器故障辨识  

Fault Recognition of Water Level Regulator Based on the Convolution Type of Wavelet Packet

作　　者：刘东东[1] 陈兵[1] 王华明[2] 

机构地区：[1]海军工程大学船舶与动力学院,湖北武汉430033 [2]海装驻沈阳地区军事代表局,辽宁沈阳110034

出　　处：《中国舰船研究》2010年第1期68-71,共4页Chinese Journal of Ship Research

基　　金：武器装备预研基金项目资助(9140A27050106JB11)

摘　　要：以某型蒸汽船的主冷凝器水位调节器为研究对象,应用卷积型小波包变换提取各种故障信号分解后的相对小波包能量,建立水位调节器标准故障库,并引进Euclidean空间距离公式作为判据进行故障识别。实验结果表明,该方法可以对水位调节器的典型故障进行快速、正确分类,降低了其故障定位与维修的盲目性。Based on the research on water level regulator of steam-power ship＇s main condenser, this paper picked up the relative energy eigenvector by using the convolution type of wavelet packet decomposition, built the standard fault database, and took Euclidean space range formula as criterion to identify faults. The experimental results show that this method can diagnose the typical faults of the water level regulator quickly and correctly. And this method can also decrease the blindness of the diagnosis and the maintainance.

关 键 词：能量特征向量 故障辨识 实验 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置] TK223.7[自动化与计算机技术—控制科学与工程]