The use of neural networks for the prediction of the critical factor of safety of an artificial slope subjected to earthquake forces

In this paper, the factor of safety values of soil against liquefaction (FS) were investigated by mean of artificial neural network (ANN) and multiple regression (MR). To achieve this, two earthquake parameters, namely, earthquake magnitude (M w ) and horizontal peak ground acceleration (a max ), and six soil properties, namely, standard penetration test number (SPT-N), saturated unit weight (γ sat ), natural unit weight (γ n ), fines content (FC), the depth of ground water level (GWL), and the depth of the soil (d) varied in the liquefaction analysis and then the FS value was calculated from the simplified method for each case by using the Excel program developed and utilized in the simulation of the feed forward ANN model with back propagation algorithm and the MR model. The FS values predicted from both ANN and MR models were contrasted with those calculated from the simplified method.In additionally, five different performance indices were used to evaluate the predictabilities of the models developed. These performance indices indicated that the ANN models is superior to the MR model in predicting the FS value of the soil.

Section: Artificial Neural Network Modelmentioning

confidence: 99%

“…In addition to the performance indices, to examine the models' predictabilities, a scaled percent error (SPE) [30][31][32][33][34][35] …”

Section: Multiple Regression Modelmentioning

confidence: 99%

The Use of Neural Networks for the Prediction of the Factor of Safety of the Soil Against Liquefaction

Erzin

Tuskan

2018

“…Cheng and Lau [21] proposed the asymmetric slope stability models based on extensions of the Bishop [3], Janbu [5], and Morgenstern-Price methods [8] in three dimensions. Much progress has been attained in 2D and 3D methods of stability of slopes in recent years [22][23][24]. Zhou and Cheng [25] presented the rigorous limit equilibrium column method, in which intercolumn forces were taken into account based on six equilibrium conditions, which included three direction force equilibrium conditions along coordinate axes and three direction moment equilibrium conditions around three coordinate axes.…”

Section: Introductionmentioning

confidence: 99%

A New Method for Three Dimensional Stability Analysis of Earth Slopes

Hajiazizi

Kilanehei

2017

Abstract. In this research, a new three-dimensional limit equilibrium method was developed. In the proposed method, the slip surface was assumed to be spherical and slices were assumed along the slip surface radius converging to a point. Moreover, force and moment equilibrium equations were used. In order to calculate Factor of Safety (FS) with the proposed method, a code was developed for two-and three-dimensional states. The model used the iteration (trial and error) method to simultaneously satisfy force and moment equilibrium equations. In order to ensure convergence of the numerical model developed in the three-dimensional state, the under-relaxation method was used in the trial and error operations. In examples, the value of FS obtained by the proposed method was 2.5 to 12 percent higher than those obtained by the other methods. In order to validate the proposed method and assess the functionality of the three-dimensional numerical model, a few examples were solved under di erent circumstances; the results were properly compliant with the results of other methods.

“…Slope stability problems have been studied under static and seismic loads by a number of researchers [1][2][3][4][5]. However, less attention has been paid to the stability of slopes and trenches under impact loads induced by dynamic compaction process.…”

Section: Introductionmentioning

confidence: 99%

Stability Analysis of Dry Sandy Slopes Adjacent toDynamic Compaction Process

Ghanbari

Hamidi

2017

Abstract. Dynamic compaction is a useful economic method for improving di erent soil types, especially loose sandy lls. However, the method has been rarely used in the vicinity of slopes due to stability concerns. In this research, dynamic compaction method adjacent to slope edge was numerically simulated using 2D plain-strain nite element models. Stability of slope models under di erent compaction energies and slope geometries at the same initial static Factor of Safety (FS) was investigated considering di erent stability criteria. These criteria include Peak Particle Velocity (PPV) or Peak Particle Displacement (PPD) on the slope, rate of change in plastic volumetric strains, yield stress ratio on the induced slip surface, and the ratio of crater depths in at and sloped models. Safe compaction distances from slope heel were calculated for di erent criteria, and it was concluded that PPV criterion yields the most conservative distances, but PPD criterion almost shows the smallest safe distances. Based on comparison of di erent criteria, it was concluded that combination of yield stress ratio and rate of plastic volumetric strain achieves the most acceptable safe compaction distance values for consideration in slope stability analyses.