采用关联维的溢流阀故障诊断  

Fault Diagnosis of Relief Valve by Using Correlation Dimension

作　　者：高宇[1] 黄宜坚[1] 

机构地区：[1]华侨大学机电及自动化学院,福建厦门361021

出　　处：《华侨大学学报（自然科学版）》2012年第3期241-246,共6页Journal of Huaqiao University(Natural Science)

基　　金：国家自然科学基金资助项目(50975098)

摘　　要：用Grassberger-Procaccia(G-P)算法合理选择嵌入维数、数据长度、延迟时间等重要参数,并且在对数曲线图中准确划定无标度区,以得到比较客观的关联维数.结果表明:关联维有定量描述非线性信息的特点,在一定的工作条件下,不同的故障类型对应不同的关联维数,通过计算溢流阀振动信号的关联维数来诊断出其工作状态.Grassberger-Procaccia（G-P） algorithm is used with reasonable selection of important parameters such as embedding dimension,length of data,delay time ect,and delimiting non-scale range accurately in logarithmic curve graph in order to get more objective correlation dimension.The result prove that correlative dimension can quantitatively describe non-linear information,and there are different correlative dimensions for different types of faults under certain conditions.So working state of relief valve can be diagnosed by computing correlation dimension of its vibration signal.

关 键 词：溢流阀 故障诊断 Grassberger-Procaccia算法 关联维 

分 类 号：TH165.3[机械工程—机械制造及自动化]