predicting peak particle velocity by artificial neural networks and multivariate regression analysis - sarcheshmeh copper mine, kerman, iran

ground vibrations caused by blasting are undesirable results in the mining industry and can cause serious damage to the nearby buildings and facilities; therefore, controlling such vibrations has an important role in reducing the environmental damaging effects. controlling vibration caused by blasting can be achieved once peak particle velocity (ppv) is predicted. in this paper, the values of ppv have been predicted and compared using the artificial neural network (ann), multivariate regression analysis (mvra) and empirical relations. the necessary data were gathered from 11 blast operations in sarcheshmeh copper mine, kerman. the neural network input parameters include distance from blast point, maximum charge weight per delay, spacing, stemming and the number of drill-hole rows in each blasting operation. the network is of the multi-layer perception (mlp) type with 24 sets of training data including 2 hidden layers, 1 output layer with the network architecture of {5-11-12-1}, and sigmoid tangent and linear transfer functions. to insure the training accuracy, the network was tested by 6 data sets; the determination coefficient and the average relative error were found to be 0.977 and 8.85%, respectively, showing the mlp network’s high capability and precision in estimating the values of the ppv. to predict these values, mvra and empirical relations were analyzed. the results have revealed that these relations have low capability in estimating the ppv parameter.