摘要
组合预测模型已经成为滑坡变形预测的一种有效方法.通过建立相关性优先的滑坡位移组合预测模型,同时顾及预测精度,以三峡库区店子湾滑坡为例,对GPS监测累积水平位移进行预测,同时构建平方和误差(SSE)、平均绝对误差(MAE)、均方误差(MSE)、平均绝对百分比误差(MAPE)、均方百分比误差(MSPE)等5个评价指标进行全方位评价,由计算结果可知,该模型预测的平均绝对百分比误差(MAPE)仅为5.5%,相关系数达到0.72,取得了较好的效果.
Combination forecasting model has become an effective method for displacement prediction of landslide.The combination model based on preferential correlation is established while taking into account the prediction accuracy to predict the GPS monitoring cumulative horizontal displacement,and take the Dianziwan landslide in Three Gorges reservoir area for example.Meanwhile,five evaluation indices,i.e.the sum squared error（SSE）,mean absolute error（MAE）,mean square error（MSE）,mean absolute percentage error（MAPE）and mean square error percentage（MSPE）,are built to make comprehensive evaluation.The calculation results show that the mean absolute percentage error（MSPE）of the prediction by the mode is only 5.5%;and the correlation coefficient reached 0.72,so as to achieve good effectiveness.
出处
《武汉大学学报（工学版）》
CAS
CSCD
北大核心
2016年第1期66-69,93,共5页
Engineering Journal of Wuhan University
基金
国家自然科学基金(编号:41172298
41302260)
湖北省自然科学基金创新群体项目(编号:2012FFA040)
关键词
组合预测
相关系数
滑坡位移
预测
combination prediction
correlation coefficient
landslide displacement
prediction
