Abstract: Based on the blasting vibration data of a high rock slope in Quanzhou, and selecting the maximum dose of single-stage, horizontal distance, height difference as impact factors, the process of training and cross-validation and final prediction is made by jointing with support vector machineSVM) and traditional methods. It is confirmed that the support vector machine theory can predict the velocity of blasting vibration well. This research result provides a way to improve the prediction accuracy of blasting vibration. 

Key words: Blasting vibration, Support vector machine, Prediction, Peak particle velocity