摘要
针对沉降数据序列的特征,研究了多项式回归方法、小波降噪方法、频谱分析法在沉降数据处理中的应用。多项式回归残差和小波降噪残差统计指标对比表明:在趋势项分离过程中,小波降噪方法优于多项式回归方法。小波降噪的残差频谱分析结果表明了沉降数据序列在2个月和4个月的周期性信息。
Probing to characters of subsidence series, we study on application of polynomial regression, wavelet denoising and frequency analysis on subsidence series processing. Contrast of statistic index of residuals shows priority of wavelet denosing. Frequency analysis on wavelet denoising residuals shows 2 months and 4 months periodical information.
出处
《测绘学院学报》
北大核心
2004年第2期90-92,共3页
Journal of Institute of Surveying and Mapping
关键词
沉降数据序列
多项式回归
小波降噪
频谱分析
subsidence series
polynomial regression
wavelet denoising
frequency analysis
