基于经验模式分解与关联维数的柴油机故障诊断  

Fault Diagnosis of Diesel Engine Based on EMD and Correlation Dimension

作　　者：陆金铭[1,2] 王醇涛[2] 陈永杰[2] 马捷[1] 

机构地区：[1]上海交通大学船舶海洋与建筑工程学院,上海200030 [2]江苏科技大学船舶与海洋工程学院,江苏镇江212003

出　　处：《船海工程》2009年第3期95-98,共4页Ship & Ocean Engineering

摘　　要：提出一种基于经验模式分解与关联维数的柴油机故障诊断方法。对3110型柴油机几种故障工况及正常情况下的缸盖振动信号进行了测试分析,采用经验模式分解经验模式分解方法对振动信号进行分解,得到固有模态函数IMF,结合G-P算法对主IMF分量分别求其关联维数。该算法用关联积分法(C-C方法)和收敛试算方法确定重构相空间的两个重要参数:时间延迟τ和嵌入维数m。通过对缸盖振动信号的关联维数分析,柴油机在不同工况状态下具有不同的关联维数特征。A new method of detecting valve fault using vihration signals of cylinder head based on EMD and correlation dimension is put forward in this paper. Surface vibration signals have been obtained under the fault condition and the normal condition on 3110 diesel. The EMD method is used to decompose the vibration signal into a number of intrinsic mode function components and then calculate the correlation dimension of each IMF component using G_P algorithm. C_ C method and the convergence of direct trial-and-error method are used to solve two important parameters in the phase reconstruction space： delay time and embedding dimension. The analyzed results show that correlation dimensions are different in the fault condition and the normal condition.

关 键 词：柴油机 故障诊断 经验模式分解EMD 关联维数 G-P算法 

分 类 号：U664.121[交通运输工程—船舶及航道工程] TH133[交通运输工程—船舶与海洋工程]