scholarly journals Assessment of shallow landslide susceptibility using an artificial neural network in Enshi region, China

2017 ◽  
Author(s):  
Bin Zeng ◽  
Wei Xiang ◽  
Joachim Rohn ◽  
Dominik Ehret ◽  
Xiaoxi Chen
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Spatial Distribution ◽  
Slope Stability ◽  
Land Use Planning ◽  
Landslide Susceptibility ◽  
Evaluation Index System ◽  
Mechanism Analysis ◽  
Unstable Slope ◽  
Artificial Neural

Abstract. Landslides are one of the most common and damaging natural hazards in mountainous areas. However, due to the complex mechanisms that influence the activation of landslides, it can often be very difficult to predict exactly when a landslide will occur. Therefore, research on landslide prevention and mitigation mainly focuses on the distribution forecasting of unstable slopes that are prone to landslides in specific regions and under multiple external forces. The prediction of the spatial distribution of these unstable slopes, termed Landslide Susceptibility Zonation, is important in helping with government land-use planning and in reducing unnecessary loss of life and property. Researching unstable slopes in the Silurian stratum in Enshi region, China, this investigation established a GIS and artificial neural network (ANN)-based method to predict the distribution of potential landslides in this area. Based on the failure mechanism analysis of typical landslides in Silurian stratum, development of evaluation index system which represents the most relevant factors that influence the slope stability, and establishment of intelligent slope stability susceptibility prediction model by artificial neural network, the spatial distribution of unstable slope zones that are prone to landslides were predicted in the study area. The results were further well supported from remote sensing data and field investigations. This research proves that the spatial unstable slope prediction method based on intelligence theory and GIS technology is accurate and reliable.

Freeway Slope Stability Evaluation Based on Multi-Level Fuzzy Neural Network

2012 ◽  
Vol 430-432 ◽  
pp. 1700-1703
Author(s):  
Yan Kai Wu ◽  
Xian Song Sang ◽  
Bin Niu
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Slope Stability ◽  
Basic Principle ◽  
Fuzzy Neural Network ◽  
Stability Assessment ◽  
High Slope ◽  
Fuzzy Neural ◽  
Multi Level ◽  
Artificial Neural

On the basis of introduced basic principle of fuzzy-artificial neural network, this article constructed a slope stability assessment index system with multi-level fuzzy neural network, and made detailed evaluation criterion according to the assessment characteristics of slope stability. Through introducing the basic principle of multi-level comprehensive assessment from fuzzy mathematics and artificial neural network theory, it overcomes the defect of difficult to be quantified in evaluation process of slope stability. Therefore, it can be better to deal with some uncertain problems occurred in the slope stability assessment process, and as much as possible to express all factors influencing slope stability really and objectively. We selected 20 single factor evaluation indexes to assess slope stability based on surveying the high slope stability in Mian county-Ningqiang county freeway section. It took "normal distribution model function" as a membership function to develop a program with the model of fuzzy neural network. Furthermore, we took 30 typical slope examples as training sample to conduct effectiveness test and feedback test for the program. After the precision requirement was met, we used the program to evaluate 21 high slope examples and compared the results with the ones solved by traditional mechanical methods. The coincidence degree by using two kinds of methods to assess the same slope stability is 76.2%.

NEURAL NETWORK AIDED EVALUATION OF LANDSLIDE SUSCEPTIBILITY IN SOUTHERN ITALY

2012 ◽  
Vol 23 (01) ◽  
pp. 1250002 ◽  
Author(s):  
SALVATORE RAMPONE ◽  
ALESSIO VALENTE
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Landslide Susceptibility ◽  
Slope Angle ◽  
Landslide Hazard ◽  
Hazard Mapping ◽  
The Neural Network ◽  
Network Output ◽  
Artificial Neural ◽  
Validation Set

Landslide hazard mapping is often performed through the identification and analysis of hillslope instability factors. In heuristic approaches, these factors are rated by the attribution of scores based on the assumed role played by each of them in controlling the development of a sliding process. The objective of this research is to forecast landslide susceptibility through the application of Artificial Neural Networks. In particular, given the availability of past events data, we mainly focused on the Calabria region (Italy). Vectors of eight hillslope factors (features) were considered for each considered event in this area (lithology, permeability, slope angle, vegetation cover in terms of type and density, land use, yearly rainfall and yearly temperature range). We collected 106 vectors and each one was labeled with its landslide susceptibility, which is assumed to be the output variable. Subsequently a set of these labeled vectors (examples) was used to train an artificial neural network belonging to the category of Multi-Layer Perceptron (MLP) to evaluate landslide susceptibility. Then the neural network predictions were verified on the vectors not used in the training (validation set), i.e. in previously unseen locations. The comparison between the expected output and the artificial neural network output showed satisfactory results, reporting a prediction discrepancy of less than 4.3%. This is an encouraging preliminary approach towards a systematic introduction of artificial neural network in landslide hazard assessment and mapping in the considered area.

Sign in / Sign up

Export Citation Format

Share Document