Time series analysis and long short-term memory neural network to predict landslide displacement

Abstract

A good prediction of landslide displacement is an essential component for implementing an early warning system. In the Three Gorges Reservoir Area (TGRA), many landslides deform distinctly and in steps from April to September each year under the influence of seasonal rainfall and periodic fluctuation in reservoir water level. The sliding becomes more uniform again from October to April. This landslide deformation pattern leads to accumulated displacement versus time showing a step-wise curve. Most of the existing predictive models express static relationships only. However, the evolution of a landslide is a complex nonlinear dynamic process. This paper proposes a dynamic model to predict landslide displacement, based on time series analysis and long short-term memory neural network (LSTM). The accumulated displacement was decomposed into a trend term and a periodic term in the time series analysis. A cubic polynomial function was selected to predict the trend displacement. By analyzing the relationships between landslide deformation, rainfall and reservoir water level, a LSTM model was used to predict the periodic displacement. The LSTM approach was found to more properly model the dynamic characteristics of landslides than static models, and make full use of the historical information. The performance of the model was validated with the observations of two step-wise landslides in the TGRA, the Baishuihe landslide and Bazimen landslide. The application of the model to those two landslides demonstrates that the LSTM model provides a good representation of the measured displacements and gives a more reliable prediction of landslide displacement than the static support vector machine (SVM) model. It is concluded that the proposed model can be used to effectively predict the displacement of step-wise landslides in the TGRA.

An AI‐based model for describing cyclic characteristics of granular materials

“…LSTM, as a variation of RNNs, has been first proposed to solve complex, long–time span tasks that have never been solved by traditional RNNs. Based on this, LSTM and its variants such as Gated Recurrent Unit (GRU) and bidirectional LSTM have been extensively used in engineering domain . Herein, LSTM, as the standard algorithm of this class of algorithms, has been adopted in this study to investigate its feasibility in modelling soil cyclic behaviour.…”

“…Compared with general neural networks, connections between hidden units can be established in recurrent neural networks (RNNs), allowing the latter to retain memories of recent events. Conventional RNNs exist gradients vanishing and exploding, a long short‐term memory (LSTM) neural network as a variation of RNNs was thus developed to overcome this problem The LSTM algorithm has recently been used in practical engineering with time‐series characteristics such as the prediction of long‐term settlement of structures, hydro‐mechanical responses of multi‐permeability porous media and structural seismic response . Because soil behaviour under cyclic loading is a continuous process, the current stress‐strain status depends on the soil behaviour at previous steps and also affects the soil behaviour at the later steps.…”

A Brief Report of Pingdi Landslide (23 July 2019) in Guizhou Province, China

“…Zhu et al [29] applied the load-unload response ratio theory to establish a landslide prediction model. Yang et al [30] applied a time series analysis and long short-term memory neural network to predict landslide displacement. The RS and permanent scatter interferometry synthetic aperture radar (PSInSAR) techniques are widely used to monitor the development of geohazards [31][32][33][34][35][36][37].…”

The Prediction of liquid holdup in horizontal pipe with BP neural network

“…A Radial basis function neural network (RBF‐NN) model was presented to determine the pressure gradient . Although deep learning algorithms such as LSTM and ESN have gain great popularity, as both a long‐ and short‐term memory network, LSTM is a kind of temporal recursive neural network, suitable for processing and predicting important events with relatively long interval and delay in time series . ESN requires a scale much larger than the node size of the general neural network, and the most mature application of ESN is still focused on the learning of time series .…”