A Review Paper on Different Effects of Artificial Neutral Network on Prediction of Various Properties of Soil Used in the Construction

This project generally deals with the study of the properties of black cotton soil that has extraordinary and uncertain behaviours which are due to its complex and indefinite physical processes that are associated during the formation of various materials. In this project, we are dealing with the complexities that are exhibited considering the geotechnical behaviour. The uncertainties that the geotechnical engineers are confronted with are mainly inconsistency, loading effects, timing effects, construction effects, and errors in soil boring. An engineer has to be skilled and competent to interpret the results of the laboratory tests for applying it in the field. Hence, to cope up with solving some of the geotechnical problems, the traditional engineering modelling techniques that cannot interpret the results accurately has been replaced with the Artificial Neural Networks (ANN). The ANN model utilizes the ANN tool in the MATLAB software. The data presented to the neural network through the input layer has given response through output layer after passing through the hidden layers. The intermediary layers or the hidden layers are locations where complicated problems are linked with input and output. These hidden layers provide the network with high degree of non-linearity. Through the application of ANN the properties of soil like the compression index, shear strength, permeability, soil compaction, lateral earth pressure, etc. are estimated. These are mainly influencing factors for describing the interaction of soil with the other earth materials during the construction phase which leads to the success, economy and safety of work. To overcome some of the complex issues that are associated with black cotton soil, properties like Liquid Limit, Plasticity Index, Maximum Dry Density and optimum Moisture content have been also studied.