Evaluation of Liquefaction Potential Using Conventional and Soft Computing Method

Earthquake damage in the twenty-first century has piqued the interest of numerous scholars and engineers working on enhancing the seismic safety of heavily populated regions. Prayagraj is one of India’s fastest-growing cities, is located on the banks of the Ganga and Yamuna rivers. The river Ganga transports maximum part of alluvial soil, which is an essential factor in determining soil liquefaction potential. Some of the other factors which also affects the liquefaction potential are local site conditions, and water table. The current study focuses on liquefaction potential of soil as determined by semi-empirical approaches suggested by Modified Seed method. The developed soft computing models’ assessment were compared with evaluated Liquefaction Potential which significantly matches with output of models. Therefore ANN & ANFIS models can be used for predicting Liquefaction potential of soils. The Seed’s and Idriss approach are utilized for evaluating soil liquefaction potential since it has a higher estimating capacity than other standard methods. Bore log data from SPT tests done at locations were used to evaluate the liquefaction potential. For training ANN and ANFIS models, 100 datasets from thirty-three bore wells up to a depth of ten meter were gathered, while 26 datasets were retained for verifying the models. The projected findings of ANN and ANFIS models when compared to the Seed’s and Idriss technique suggest that training ANN and ANFIS models were capable of accurately forecasting liquefaction potential.