Evaluation of dump slope stability of a coal mine using artificial neural network

Abstract

Dump slope stability is critical in the design and operation of any open pit mines having significant impacts on safety and on the economics of an open pit project. The issues related to the stability of dumps is catching attention worldwide from quite some time, which is quite important for the safe working in and around monstrous structures as well as restricted availability of land. In the present paper, artificial neural network (ANN) has been used to calculate the factor of safety of dump slope of a coal mine. A three-layer, feed-forward back-propagation neural network having optimum hidden neurons is used to model. The input parameters are dump slope geometry, geotechnical properties and hydrological conditions have been used to evaluate the stability of dump slope. Six input parameters and one output parameters have been trained using various algorithms. New dump slope data sets have been used for the validation and comparison of the factor of safety of dump slope. Factor of safety has also been determined using numerical modeling technique. Results were compared between factor of safety, calculated from the numerical modeling tool and predicted value of factor of safety by ANN. It was found a very closer agreement between the predicted and calculated values of factor of safety for the dump slope with ANN over numerical modelling.