基于稀疏表示的数控机床主轴故障特征提取  

Fault Feature Extraction of CNC Machine Tool Spindle Based on Sparse Representation

作　　者：黄日进[1] 李成贵 邹才深 Huang Rijin;Li Chenggui;Zou Caishen

机构地区：[1]广西英华国际职业学院,广西钦州530213

出　　处：《机械制造》2023年第12期66-68,73,共4页Machinery

基　　金：广西壮族自治区中青年教师科研基础能力提升项目(编号:2023KY1956);广西英华国际职业学院校级中青年教师科研基础能力提升项目(编号:YHKY202313)。

摘　　要：数控机床主轴振动信号中包含丰富的设备运行状态信息,对故障监测分析而言具有重要价值。振动信号在时域呈现出非平稳特性,在噪声背景下难以有效提取故障信息。对此,提出一种基于稀疏表示的数控机床主轴故障特征提取方法。采用希尔伯特-黄变换对原始时域信号进行处理,在展开的瞬时频率谱上构建信号的过完备字典,并运用稀疏主成分分析对信号的字典系数矩阵进行求解,得到故障特征的稀疏表示。通过仿真,验证了这一方法的优点和有效性。The vibration signal of the spindle of CNC machine tool contains rich information on the operating status of equipment,which is of great value for fault monitoring and analysis.The vibration signal exhibits non-stationary characteristic in the time domain,and it is difficult to effectively extract fault information under the background of noise.In this regard,a fault feature extraction method of CNC machine tool spindle based on sparse representation was proposed.The Hilbert-Huang transform was used to process the original time domain signal,and the super complete dictionary of the signal was constructed on the unfolded instantaneous frequency spectrum,and the dictionary coefficient matrix of the signal was solved by sparse principal component analysis to obtain the sparse representation of fault feature.Through simulation,the advantage and effectiveness of this method were verified.

关 键 词：稀疏表示 数控机床 主轴 故障特征 提取 

分 类 号：TG502.31[金属学及工艺—金属切削加工及机床]