基于改进G-P算法的往复式压缩机故障诊断技术  

Improved technique of fault diagnosis for reciprocating compressor based on G-P algorithm

作　　者：刘岩[1] 

机构地区：[1]齐齐哈尔大学机械学院,黑龙江齐齐哈尔161006

出　　处：《齐齐哈尔大学学报（自然科学版）》2010年第2期33-36,共4页Journal of Qiqihar University(Natural Science Edition)

基　　金：齐齐哈尔大学青年教师科研启动支持计划项目

摘　　要：采用基于经验模态分解的G-P算法提取往复式压缩机振动信号关联维数,建立了小波降噪模型,利用自相关函数法和伪相图法结合起来确定延迟时间,并采用EMD法计算嵌入维数,通过洛仑兹系统的仿真研究,将此法应用在往复式压缩机模拟故障试验中,诊断效果较好。Extract correlation dimension of vibration signal of reciprocating compressor based on G-P Algorithm of Empirical Mode Decomposition（EMD）technology,with found mode of wavelet denoise,with the incorporation of method of self-interrelated function and method of pseudo-phase portrait ensure delay-time and calculate embedding dimension by EMD method.Through simulation analysis of Lorenz system Apply the method to identify valve faults of reciprocating compressor,the result indicate that the modeling method is effective.

关 键 词：往复式压缩机 故障诊断 关联维数 经验模态分解 自相关函数 伪相图 

分 类 号：TH457[机械工程—机械制造及自动化] TP277[自动化与计算机技术—检测技术与自动化装置]