Modification and prediction of blast-induced ground vibrations based on both empirical and computational techniques

Ground vibration (GV) is a blasting consequence and is an important parameter to control in mining and civil projects. Previous GV predictor models have mainly been developed considering two factors; charge per delay and distance from the blast-face. However, mostly the presence of the water as an influential factor has been neglected. In this paper, an attempt has been made to modify United State Bureau of Mines model (USBM) by incorporating the effect of water. For this purpose, 35 blasting operations were investigated in Chadormalu iron mine, Iran and required blasting parameters were recorded in each blasting operation. Eventually, a coefficient was calculated and added in USBM model for effect of water. To demonstrate the capability of the suggested equation, several empirical models were also used to predict measured values of PPV. Results showed that the modified USBM model can perform better compared to previous models. By establishing new parameter in the USBM model, a new predictive model based on gene expression programming (GEP) was utilized and developed to predict GV. To show capability of GEP model in estimating GV, linear multiple regression (LMR) and non-linear multiple regression (NLMR) techniques were also performed and developed using the same datasets. The results demonstrated that the newly proposed model is able to predict blast-induced GV more accurately than other developed techniques.