Development of imperialist competitive algorithm in predicting the particle size distribution after mine blasting

Proper rock fragmentation is one of the most important aims in surface mines as well as tunneling projects. The main purpose of the current study is to forecast rock fragmentation through imperialist competitive algorithm (ICA). Shur river dam region, in Iran, was considered and 80 sets of data, including D 80 , as a standard for evaluating the fragmentation, maximum charge per delay, spacing, burden, powder factor, stemming and rock mass rating were prepared. For comparison aims, artificial neural network was also developed and the predicted values by ICA model was then compared to ANN results. In the other words, two forms of ICA models, i.e., ICA-linear and ICA-power models as well as ANN were employed for predicting the D 80 . To compare the performance capacity of the ICA and ANN models, several statistical evaluation criteria, such as variance account for (VAF), R -square ( R 2 ), root mean square error (RMSE) were computed. Finally, it was demonstrated that the ICA-power model with the R 2 of 0.947, VAF of 93.96% and RMSE of 1.23 was more suitable and acceptable model for predicting the D 80 than the ICA-linear model with the R 2 of 0.943, VAF of 93.49% and RMSE of 1.28 and the ANN model with the R 2 of 0.897, VAF of 88.78% and RMSE of 1.68 and had the capacity to generalize.