Forecasting ground vibration due to rock blasting: a hybrid intelligent approach using support vector regression and fuzzy C-means clustering

Ground vibration due to rock blasting is one of the main concerns in the mining operation and may cause severe damages to nearby structures and population. Therefore, accurate prediction of ground vibration plays an important role in the minimization of environmental effects of blasting in mines. Nowadays, application of artificial intelligence techniques has grown dramatically for predicting ground vibration. The researchers found that these techniques are suitable tools for accurate prediction. In this study, a hybrid model is proposed for estimating ground vibration using support vector regression (SVR) and fuzzy C-means clustering (FCM). The model is developed based on blasting data compiled from Sarcheshmeh copper mine, Iran. To demonstrate the efficiency of FCM-SVR model, this model is compared with SVR model (without data clustering) and United States Bureau of Mines empirical equation (with and without data clustering). The results show that SVR intelligent technique is more efficient and accurate than the empirical equation, and data clustering plays an effective role in accurate prediction of blast-induced ground vibration.