Proposing of a new soft computing-based model to predict peak particle velocity induced by blasting

Estimation of ground vibration induced by blasting operations is an important task to control the safety issues at the surface mines and civil projects. By reviewing the previous studies, some empirical and soft computing models have been proposed to estimate blast-induced ground vibrations. The main goal of this research is to propose a new predictive model in the field of ground vibration estimation. For this aim, the group method of data handling (GMDH) model which is a type of neural network, is proposed with respect to input parameters including the stemming length, powder factor, burden to spacing ratio, distance from the blast-face, blast-hole depth and maximum charge per delay. Also, the peak particle velocity, as the most common descriptor for evaluating the ground vibration, was selected as the output. The required datasets were collected from a quarry in Penang, Malaysia, using 102 blasting operations. Several criteria such as root mean square error (RMSE) and coefficient of determination (R²) were utilized to determine the reliability of the GMDH. Based on the obtained results, the GMDH forecasting technique with R² of 0.911 and RMSE of 0.889 can be presented as a powerful technique in predicting the blast-induced ground vibration.