振动与冲击
振動與遲擊
진동여충격
Journal of Vibration and Shock
2015年
21期
76-81
,共6页
张云强%张培林%吴定海%李兵
張雲彊%張培林%吳定海%李兵
장운강%장배림%오정해%리병
齿轮箱%时频分析%多重分形%Q 阶矩结构分割函数法%分数阶 S 变换
齒輪箱%時頻分析%多重分形%Q 階矩結構分割函數法%分數階 S 變換
치륜상%시빈분석%다중분형%Q 계구결구분할함수법%분수계 S 변환
gearbox%time-frequency analysis%multifractal%Qth order weighted moment structure partition function method%fractional S transformation
针对齿轮箱振动信号的非线性和非平稳性,提出一种基于Q阶加权矩结构分割函数法的分数阶时频谱多重分形特征提取方法。首先构造时频分辨率较好的分数阶 S 变换技术获取齿轮箱振动信号的分数阶时频谱;然后针对分数阶时频谱的特点,设计出一种 Q 阶加权矩结构分割函数法,用于提取分数阶时频谱的多重分形特征参数。对5种状态的齿轮箱振动信号进行了分析和研究,结果表明齿轮箱振动信号的分数阶时频谱具有多重分形特性,Q 阶加权矩结构分割函数法提取的多重分形特征参数能有效描述分数阶时频谱的多重分形特征。
針對齒輪箱振動信號的非線性和非平穩性,提齣一種基于Q階加權矩結構分割函數法的分數階時頻譜多重分形特徵提取方法。首先構造時頻分辨率較好的分數階 S 變換技術穫取齒輪箱振動信號的分數階時頻譜;然後針對分數階時頻譜的特點,設計齣一種 Q 階加權矩結構分割函數法,用于提取分數階時頻譜的多重分形特徵參數。對5種狀態的齒輪箱振動信號進行瞭分析和研究,結果錶明齒輪箱振動信號的分數階時頻譜具有多重分形特性,Q 階加權矩結構分割函數法提取的多重分形特徵參數能有效描述分數階時頻譜的多重分形特徵。
침대치륜상진동신호적비선성화비평은성,제출일충기우Q계가권구결구분할함수법적분수계시빈보다중분형특정제취방법。수선구조시빈분변솔교호적분수계 S 변환기술획취치륜상진동신호적분수계시빈보;연후침대분수계시빈보적특점,설계출일충 Q 계가권구결구분할함수법,용우제취분수계시빈보적다중분형특정삼수。대5충상태적치륜상진동신호진행료분석화연구,결과표명치륜상진동신호적분수계시빈보구유다중분형특성,Q 계가권구결구분할함수법제취적다중분형특정삼수능유효묘술분수계시빈보적다중분형특정。
For the nonlinearity and nonstationarity of gearbox vibration signals,a multifractal feature extraction approach for their fractional time-frequency spectra based on a Qth order weighted moment structure partition function method was proposed.A fractional S transformation with better time-frequency resolution was firstly constructed to obtain fractional time-frequency spectra of gearbox vibration signals.Then,aiming at the characteristics of fractional time-frequency spectra,a Qth order weighted moment structure partition function method was designed to extract multifractal features of the fractional time-frequency spectra.The gearbox vibration signals under five states were analyzed.The results indicated that the fractional time-frequency spectra of gearbox vibration signals have the multifractal features,and the extracted multifractal features using the Qth order weighted moment structure partition function method can more effectively describe the multifractal characteristics of fractional time-frequency spectra.