摘要
针对滑坡灾害预测问题,提出了一种基于二次指数平滑和Elman神经网络的滑坡位移预测方法。首先采用二次指数平滑算法将滑坡累积位移分解为趋势性位移和周期性位移,然后利用Elman神经网络建立影响因子和周期性位移之间的映射关系,最后将两个预测子位移叠加获得预测的累积位移。在三峡库区白水河滑坡上的验证结果表明,该方法是有效的,优于其他两种预测方法。
To solve the landslide disaster prediction problem,we proposed a landslide displacement prediction method based on double exponen-tial smoothing and Elman neural network.Firstly,we used double exponential smoothing algorithm to decompose cumulative landslide displace-ment into trend displacement and periodic displacement.Then,we used Elman neural network to establish the relationships between influencing factors and periodic displacements.Finally,we obtained predicted cumulative displacement by superposition of two predicted sub displacements.The verification results on Baishuihe landslide in the Three Gorges Reservoir area show that the method is effective and superior to the other two prediction methods.
作者
冯杭华
胡俊凯
FENG Hanghua;HU Junkai(Zhejiang Huadong Surveying and Engineering Safety Technology Co.,Ltd.,Hangzhou 310014,China)
出处
《地理空间信息》
2022年第3期103-105,112,共4页
Geospatial Information
关键词
滑坡位移预测
指数平滑算法
ELMAN神经网络
三峡库区
landslide displacement prediction
exponential smoothing algorithm
Elman neural network
the Three Gorges Reservoir
