倒傳遞類神經網路法模擬崩塌滑移特性

Landslide Displacement Characteristic Predicted for Back Propagation Neural Network Method

黃文仁(Wen-Ren Huang) ； 徐弘明(Hon-Min Hsu) ； 許中立(Chung-Li Hsu)

《坡地防災學報》 9卷1期 (2010/10) Pp. 33-42

https://doi.org/10.29995/JSHR.201010.0004

摘要

崩塌地為山坡地最具破壞之災害型態之一，藉由實施現場之監測與分析評估，可以協助管理機關擬訂適當的防災策略。本研究以台東縣池上鄉山棕寮崩塌地為例，利用倒傳遞類神經網路分析法，建構具高複雜且非線性之關係模式，並用以預測坡面位移變化。網路輸入係以直接關係位移之物理因子為變數，建構最佳之4層網路，其中一層為輸入層，採用前期降雨量、入滲係數、降雨強度、地形坡度、地下水位、土壤凝聚力與土壤內摩擦角等7個因子為輸入變數，另有二層隱藏層及一層輸出層。試驗分析結果顯示所建構之倒傳遞類神經網路分析法其對於本崩塌地坡面滑移具有良好之預測精度，應可提供作為山坡地防災與保育管理之參考。

關鍵字

崩塌 ； 倒傳類神經網路 ； 預測

並列摘要

Landslides has become one disaster type of the most serious destroy on the slope lands. The way to monitoring and assessment for landslide area can help government agencies to select suitable management and plan mitigation in unstable landslide areas. This research presents a case study of landslide monitoring and assessment at Shanzongliao landslide area in Taitung County, attempt to predict slope movements using back propagation neural network (BPNN), as well as use powerful tools to model and investigate various complex and non-linear phenomena. The BPNN can perform calculation to use MATLAB program with the Levenberg-Marquardt algorithm. The data from the case study are used to train and test the developed model to enable prediction of the magnitude of the ground movements with the help of input variables that have direct physical significance. An infiltration coefficient is introduced in the network architecture apart from antecedent rainfall, rainfall intensity, slope profile, groundwater level and shear strength of soil. A four-layered back propagation neural network with an input layer, two hidden layers and one output layer is found optimal. The developed BPNN mode demonstrate a promising result, and have good potential accurately for predicting the slope movement, and can offer the reference of disaster prevention and utilizing management at the steep slope.

並列關鍵字

Landslide ； Back propagation neural network ； Prediction