摘要
介绍小波方法和非线性回归估计的基本理论,研究利用小波变换得到非线性回归估计函数的方法,且将该方法的结果与仅用回归估计的结果进行对比研究,发现对监测数据列进行小波分解并对小波系数作用于阈值,然后通过对小波系数估计能得到回归参数。研究结果表明基于小波方法的非线性回归模型更接近变形体变形规律,是一种能够更好地进行预测的方法,该模型切实可行。
The elementary theory of wavelet method and nonlinear regression is introduced and nonlinear regression function obtained by making fully use of wavelet transform is given. The result is compared with result using only regression model. In addition, it is discovered that regression parameters are gained by estimating wavelet coefficient after wavelet coefficients are influenced by threshold during wavelet decomposition. The studied results show the nonlinear model based on wavelet method is more closed to the order of deformation and the method can predict efficaiently, and the model is effective and feasible.
出处
《测绘学报》
EI
CSCD
北大核心
2006年第4期337-341,共5页
Acta Geodaetica et Cartographica Sinica
基金
测绘基础研究基金项目(1469990322433-04-02)
关键词
噪声
阈值
预测
小波
回归
noise
threshold
prediction
wavelet
regression
