最优非负变权组合模型在滑坡变形预测中的应用  

Application of the Optimal Non-negative Variable Weight Combination Model in Landslide Deformation Prediction

作　　者：李飞达 LI Feida(Hubei Urban Construction Vocational and Technological College,Wuhan 430205,China)

机构地区：[1]湖北城市建设职业技术学院,湖北武汉430205

出　　处：《测绘与空间地理信息》2022年第8期210-212,219,共4页Geomatics & Spatial Information Technology

摘　　要：针对单一滑坡变形预测模型自身存在的不足,提出了基于灰色GM(1,1)、卡尔曼滤波和GA-BP神经网络的最优非负变权组合预测模型。通过与各单一模型对比分析,说明了最优非负变权组合预测模型能够综合各单一模型的有效信息,预测精度较高,在一定程度上降低了预测风险,提高了预测的可靠性。Aiming at the shortcomings of single landslide deformation prediction model,an optimal non-negative variable weight combination prediction model based on Grey GM(1,1),Kalman filter and GA-BP neural network is proposed.By comparing the different models,it is shown that the optimal non-negative variable weight combination prediction model can synthesize the effective information of each single model,and has higher prediction accuracy,which reduces the prediction risk to a certain extent and improves the reliability of prediction.

关 键 词：滑坡变形 单一模型 最优非负变权 精度评定 

分 类 号：P25[天文地球—测绘科学与技术] TB22[天文地球—大地测量学与测量工程]