基于数据融合算法优化的GM(1,1)模型在矿区地表沉降中的应用  

Application of GM(1.1) Model in Mining Surface Subsidence Based on Optimization Algorithm of Data Fusion

作　　者：杨军[1] 马大喜[1] 

机构地区：[1]江西理工大学研究生院,江西赣州341000

出　　处：《温州大学学报（自然科学版）》2014年第2期51-57,共7页Journal of Wenzhou University(Natural Science Edition)

摘　　要：矿区地表沉降一直以来是矿山安全管理部门关注的重点,准确地预测矿区地表沉降可以给矿山安全带来指导性的意义.运用"幂函数-指数函数"的复合变换来提高监测原始数据的平滑度,然后对具有多个沉降监测数据的特定年份,运用GM(1,1)模型来预测地表沉降,利用数据融合算法对多次预测的结果进行优化分析,获得精度较高的预测结果.运用该方法对某矿区地表沉降数据进行预测,结果表明该模型具有良好的预测能力.Surface subsidence in mining areas is always the focus concerned by the safety department of mines. Thus, accurate prediction of surface subsidence in mines means significantly to mine safety. We can make use of the composite conversion of＆quot;power function-exponential function＆quot;to improve the evenness of monitoring primary data and then to predict the ground surface settlement by means of GM（1.1） Model regarding to a particular year with multiple settlement monitoring data. The optimized analysis to results of multiple predictions is made out of data fusion in order to obtain more accurate prediction results. The author ever made a mining area subsidence prediction in this method, which turn to prove that GM（1.1） Model possesses an ideal predictive power.

关 键 词：数据融合 复合变换 GM(1 1)模型 沉降预测 

分 类 号：TD173[矿业工程—矿山地质测量]