铣削颤振过程中的振动非线性特征试验  

Experimental Study of the Nonlinear Characteristics of Vibration in Milling Chatter

作　　者：吴石[1] 刘献礼[1] 肖飞[1] 

机构地区：[1]哈尔滨理工大学机械动力工程学院,哈尔滨150080

出　　处：《振动．测试与诊断》2012年第6期935-940,1035,共6页Journal of Vibration,Measurement & Diagnosis

基　　金：国家自然科学基金资助项目(编号:51275139);中国博士后科研基金资助项目(编号:20110491098);黑龙江省博士后科研基金资助项目(编号:LBH-Z11110)

摘　　要：为了分析铣削过程振动信号的非线性特征,使用双谱、李雅普诺夫系数、变形维数和近似熵分析变切深铣削过程中平稳铣削振动信号、颤振孕育振动信号和颤振振动信号。试验结果表明,振动信号在铣削颤振孕育和发生状态时具有强混沌特征,其双谱特征和混沌特征相结合可以作为识别颤振孕育、发生的有效手段,基于球形支持向量机分类器对平稳铣削、颤振孕育和颤振发生进行识别,识别准确率达98.0%。In order to analyze the milling process of nonlinear vibration signal characteristics,vibration signals between normal cutting,chatter gestation and outbreak is studied by the methods of bispectrum,Lyapunov exponent,fractal dimensions and approximate entropy.It is proved that,vibration signals have a strong state chaos characteristic in milling chatter gestation and outbreak,whose characteristics of bispectrum and chaos can be used as recognition.Based on the spherical support vector machine(SVM) classifier,the accuracy of recognition rate between normal cutting,chatter gestation and outbreak can reach at 98.0%.

关 键 词：铣削颤振 颤振预报 非线性特征 双谱 混沌 

分 类 号：TG506.1[金属学及工艺—金属切削加工及机床] O322[理学—一般力学与力学基础]