摘要
将指数平滑法应用于基坑深层位移预测中,并利用累计方差和最小准则选取平滑系数α。运用指数平滑法对某基坑工程3#测斜孔20 m深处监测数据进行预测分析,得到最大绝对误差为0.41 mm,远小于规定值,表明预测效果良好。通过比较3#测斜孔15 m深处与20 m深处的同期预测数据的平均绝对误差与均方差,得出15 m深处预测误差较20 m深处大,表明20 m处预报较准确。
Exponential smoothing was adopted for the predication of deep displacement in foundation pit. The smoothing coefficient a was selected based on the minimum value of cumulative sum of variance. The monitoring data obtained from 3# hole with clinometer at the depth of 20 m was predicted and analyzed by using exponential smoothing. The maximum error obtained is 0.41 far less than specified value, showing a good prediction effect. Based on the comparison between the mean absolute error and mean square error for the predicted results obtained at the depth of 15 m and 20 m of 3# hole, it is found that the prediction for the depth of 15 m has a bigger error than the depth of 20 m, so the prediction for the depth of 20 m is more precise.
出处
《矿冶工程》
CAS
CSCD
北大核心
2013年第2期5-7,共3页
Mining and Metallurgical Engineering
基金
湖南省重点学科建设项目资助
关键词
基坑
深层位移
指数平滑
预报
foundation pit
deep displacement
exponential smoothing
prediction
