Evaluation and Prediction of Blast-Induced Ground Vibration Using Gaussian Process Regression at Saindak Copper–Gold Open Pit Mine, Pakistan

In surface mining, blasting is the most common and economical method for rock excavation and fragmentation. Nevertheless, the ground vibration caused by blasting is the major concern in mining, and that can result in significant damages to the nearby environment and structures. Therefore, it is important to precisely forecast the blasting vibration to mitigate and minimize its hazardous effects. This study seeks to develop the advanced machine learning model, namely Gaussian process regression (GPR), for predicting the peak particle velocity (PPV), being the key indicator of blast-induced ground vibration at Saindak Copper–Gold open-pit mine Pakistan, utilizing a dataset of 96 real blast events containing six (6) corresponding blast design parameters, including hole depth (HD), burden (B), spacing (S), stemming length (SL), monitoring distance (D), and maximum charge per delay (MCP). Additionally, the study developed the multiple linear regression (MLR) and empirical model proposed by Indian Standard to compare the predictive capability of GPR. The models’ performances and accuracies were evaluated employing four evaluation metrics such as coefficient of determination (R²), root mean square error (RMSE), mean absolute error (MAE), and mean absolute percent error (MAPE). GPR showed a better predictive capability, exhibiting highest accuracy with (R²), of 0.91 and low (MAE), (RMSE), and (MAPE) values of (0.04), (0.009), and (0.172), respectively, on the testing data. Based on the sensitivity analysis, burden was the most significant parameter influencing the output (PPV). The current research shows that GPR has the potential to effectively estimate the PPV of the ground vibration induced by blasting. In conclusion, the findings can be helpful for geotechnical engineers who want to use advanced machine learning approaches instead of conventional models to predict the PPV of blast-induced ground vibration.