非线性最小二乘法在金刚石研磨振动信号处理中的应用  

Application of Nonlinear Least Square Method on Vibration Signals Processing of Diamond Polishing

作　　者：黄仁宏[1] 傅惠南[1] 

机构地区：[1]广东工业大学机电工程学院,广东广州510006

出　　处：《机电工程技术》2009年第9期115-116,132,共3页Mechanical & Electrical Engineering Technology

摘　　要：将金刚石研磨时采集得到的振动数据拟合为假设多谐波叠加形式的问题转化为非线性最小二乘法计算问题,运用小波包分解重构和FFT变换来确定谐波初值,再通过Gauss—Newton非线性迭代算法进行求解计算,确定谐波参数模型。通过对金刚石研磨实际加工信号的处理,表明此方法能够较为准确的拟合出假设的谐波形式,为判断金刚石研磨振动与研磨质量之间的关系提供了有力的依据。This paper transformed the problem of vibration data acquired in diamond polishing fitting supposed multi-harmonic into nonlinear least square calculation.First,in order to get the initial value of harmonics,wavelet packet decomposition and reconstruction and FFT transformation were used.Second,Parameter Model of harmonics was given by Gauss-Newton nonlinear least iterative procedure with the initial value.This method was proved to give the supposed multi-harmonic exactly through processing actual vibration signal of diamond polishing. It also provided the judgment evidences of the relationship between the vibration of diamond polishing and polishing effect.

关 键 词：非线性最小二乘法 小波包分解重构 研磨振动 

分 类 号：TG7[金属学及工艺—刀具与模具]