摘要
在灰色系统全数据GM(1,1)模型基础上,经过新陈代谢过程、分阶残差修正和权重修正,依次建立了多个不同修正方式的模型,建模过程考虑了实测信息的非等间隔性。结合富阳来龙山滑坡监控信息,对位于滑坡体主滑段,历时4个典型滑移阶段的B14测孔滑带土位移作预测分析,经过各模型精度检验比较表明,所提出的二阶残差权重修正GM(1,1)模型大幅提高了预测精度和适应性。
Based on the all data GM（1,1） model of grey system, different modified models are set up in their given order through out the course of metabolism, residual error modified of different rank, and weighted modified. In setting up the model, unequal intervals in the data of the actual measure are considered; and equal intervals of grey system forecasting model are extended to unequal ones. Together with the information of landslide monitored in Lailong Mountain in Fuyang City, the inclinometer B14, which lies in the active slide area and lasts four typical sliding phases, is studied in the course of where the displacement in sliding zone is forecasted and analyzed comparatively by the computational programs of various models. The result of model precision test shows that the residual error of two rank and weighted modified GM（1, 1） model can greatly improve the forecasting precision and adjustability.
出处
《岩石力学与工程学报》
EI
CAS
CSCD
北大核心
2005年第22期4099-4105,共7页
Chinese Journal of Rock Mechanics and Engineering
关键词
工程地质
滑坡
灰色系统预测模型
修正GM(1
1)模型
残差
权重
engineering geology
landslide
grey system forecasting model
modified GM（1, 1） model
residualerror
weight
