Predicting potential of blast-induced soil liquefaction using neural networks and neuro-fuzzy system

In recent years, controlled blasting has turned into an ecient method for evaluation of soil liquefaction on a real scale and of ground improvement techniques. Predicting blast-induced soil liquefaction using collected information can be an eective step in the study of blast-induced liquefaction. In this study, to estimate residual pore pressure ratio, rst, a multi-layer perceptron neural network is used in which error (RMS) for the network was calculated as 0.105. Next, a neuro-fuzzy network, ANFIS, was used for modeling. Dierent ANFIS models are created using Grid Partitioning (GP), subtractive clustering (SCM), and Fuzzy C-Means clustering (FCM). Minimum error is obtained using FCM at about 0.081. Finally, Radial Basis Function (RBF) network is used. Error of this method was about 0.06. Accordingly, RBF network has better performance. Variables, including ne-content, relative density, eective overburden pressure, and SPT value, are considered as input components, and residual pore pressure ratio, Ru, was used as the only output component for designing prediction models. In the next stage, the network output is compared with the results of a regression analysis. Finally, sensitivity analysis for RBF network is tested, and its results reveal that 0v 0 and SPT are the most eective factors for determining Ru.