Open AccessLandslide Displacement Prediction Based on Extended Escendant Strategy PSO Neural Network
Author(s) -
Yuqiu Lin,
Jian Wen-xing,
Chang Xu,
Yuxiang Pan,
LI Lin-jun
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/750/1/012141
Subject(s) - landslide , artificial neural network , displacement (psychology) , three gorges , support vector machine , computer science , nonlinear system , artificial intelligence , series (stratigraphy) , engineering , geology , geotechnical engineering , quantum mechanics , psychotherapist , paleontology , psychology , physics
A new PSO algorithm based on extended escendant strategy was proposed. The dynamic change of the landslide displacement time series is nonlinear and uncertain. Combine extended escendant strategy PSO with the Elman neural network, and establish a landslide displacement prediction model based on EESPSO-ENN to realize the dynamic prediction of landslide displacement. Taking the Baishuihe landslide in the Three Gorges Reservoir area as an example, select the monitoring data of ZG93 from 2013 to 2016 as training samples and test samples for training and prediction. Comparing the prediction results of EESPSO-ENN with the BP neural network and SVM method, the results demonstrate that the EESPSO-ENN model has a small prediction error and its prediction effect applied in the Baishuihe landslide is better than BP neural network and SVM method. The validity of EESPSO-ENN was verified.