Liquefaction Analysis in 3D based on Neural Network Algorithm

Simplified techniques based on in situ testing methods are commonly used to assess seismic liquefaction potential. Many of these simplified methods are based on finding the liquefaction boundary. As the liquefaction classification problem is highly nonlinear in nature, it is difficult to develop a comprehensive model taking into account all the independent variables, such as the seismic and soil properties, using conventional modeling techniques. These various simplified procedure have been developed, using case studies that liquefied or not during earthquake, to estimate liquefaction potential of soils. In order to address liquefaction engineering, this paper proposed to use an artificial neural network. ANN has the capability to train itself with available data sets and extrapolate outcome for unknown scenario based on the training. It is particularly helpful for large data sets when human brain is inefficient. Various ANN models have already been in used for liquefaction assessment. However, this paper is more objective in applying ANN in liquefaction prediction in 3D dataset. In this study, a neural network approach is used to evaluate seismic liquefaction potential based on actual 3D field records have done. First, the data with 3D parameters used for training and testing. Second, the inputs for the model are selected on their physical meaning with respect to liquefaction. Finally, the contribution strengths of the parameters calculated to see which parameter more affects the liquefaction potential of the area. Also, with this saved model, it can be used as a forecasting tool for analysis 3D liquefaction potentials in a short way.