影响往复式压缩机关联维数提取的因素分析  

Influencing Factor Analysis of Correlation Dimension Extraction of Reciprocating Compressors

作　　者：刘岩[1] 王金东[2] 

机构地区：[1]齐齐哈尔大学,黑龙江齐齐哈尔161006 [2]大庆石油学院,黑龙江大庆163318

出　　处：《流体机械》2010年第3期33-36,共4页Fluid Machinery

摘　　要：论述了相空间重构方法影响振动信号关联维数准确性的因素。通过小波包改进算法对信号降噪,利用自相关函数法和伪相图法结合技术选择时间延迟和最近邻域发散度法确定嵌入维数,并将此法应用在较难识别的往复式压缩机轴瓦与十字头故障征兆中,获得较均一稳定的关联维数,验证了此法的准确性和有效性。The influencing factor of accuracy of phase space reconstruction method extract correlation dimensions was discussed. with meliorated wavelet packet analysis and reconstruction algorithm denoise, with incorporation technology of self-interrelated function and method of Pseudo-phase portrait ensure time delay , with method of Nearest Neighbor Divergence ensure embedding dimension ,Apply the method to identify undistinguishable reciprocating compressors fault syndrome of bearing shell and joint cross , we obtain uniform and steady correlation dimensions , Have proved effectivity and accuracy of the method.

关 键 词：往复式压缩机 故障诊断 小波包 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]