灰色线性回归组合模型在北京地面沉降分层预测中的应用
Grey-linear Regression Combination Model for Fitting and Predicting Layered Land Subsidence in Beijing
摘要 灰色系统理论GM(1,1)模型,应用于地面沉降模拟和预测中只能分析数据的指数变化规律。对于地面沉降发展过程中,存在的线性关系不能有效地反映。本文利用灰色组合模型中的第一类灰色组合模型即GM(1,1)与线性回归模型相融合。选取北京东部某地面沉降监测站2004~2012年的分层监测数据建立模型,计算出各监测层位沉降的数学模型,并以此预测各监测层位地面沉降量。结果表明:利用灰色线性回归组合模型在对地面沉降进行分层模拟和预测是可行的。在已有数据的基础上,利用数学模型进行沉降模拟时,两种模型的精度均很高,但通过模型预测未来一年沉降量时,灰色线性回归组合模型的精度,要远高于普通均值GM(1,1)模型。
Abstract:
GM (1, 1) model of grey system theory can only be applied to analyze monitoring data with exponent law, but can't effectively reflect a linear relationship in the simulation and prediction of land subsidence. A combination of GM (1, 1) grey model and linear regression model is proposed in this paper. Such combination model is established using data from 2004-2012 and used to make a prediction for each monitoring layer of the land subsidence monitoring station located in the east of Beijing. The results show that it is feasible to apply grey-linear regression combination model for iftting and predicting the layered land subsidence. The precision of iftting the existing data for both of GM (1, 1) model and combination model is high. However, the precision of the latter model is much higher than GM (1, 1) model when predicting the future land subsidence in a year.
Author: ZHOU Yi LUO Yun JIA Sanman TIAN Fang
作者单位: 北京大学地球与空间科学学院,北京 100871; 北京市水文地质工程地质大队,北京 100195 中国地质环境监测院,北京,100081 北京市水文地质工程地质大队,北京,100195
刊 名 城市地质
Journal Urban Geology
年,卷(期): 2014, (4)
分类号: P642.26
机标分类号: TB1 X70
在线出版日期: 2015年2月5日
基金项目: 北京地区地面沉降防控与地下水资源合理开发研究(Z131100005613022);基于北斗卫星的地面沉降监测信息实时传输技术研究