The present study has been conducted in surface limestone mine to select the controllable blasting design parameters affecting the peak particle velocity (PPV) for assessment of ground vibration. The study aims at using two prominent statistical tools, namely, principal component analysis (PCA) and stepwise selection and elimination (SSE) techniques for identifying the key controllable parameters affecting the PPV. For determining the correlation coefficient between blasting design parameters and PPV from 44 field scale trial blasts, multi-variate linear regression (MLR) technique was done. It was found that the PCA and SSE eliminated a huge number of controllable parameters to identify the most important parameters, affecting the PPV. The PCA eliminated 6 number of parameters while SSE eliminated 5 number of parameters, and the coefficients of determination (R²) obtained were 0.616 and 0.584 for PCA and SSE respectively. The predictor equations were evolved, and these equations were used to validate the PPV results for another set of 21 field scale trial blasts. The predictor equations have been found to be fairly accurate in predicting the PPV values. Further, the PCA technique provides very near prediction of PPV with high degree of correlation in comparison to SSE technique. The paper highlights the role of state-of-art statistical tools in selecting the blasting design parameters affecting the PPV in field-scale blasting.

本研究已在地表石灰石矿山中进行，以选择影响峰值粒子速度（PPV）的可控爆破设计参数来评估地面振动。这项研究旨在使用两种重要的统计工具，即主成分分析（PCA）和逐步选择和消除（SSE）技术来识别影响PPV的关键可控参数。为了确定44个现场规模试验爆破的爆破设计参数与PPV之间的相关系数，进行了多元线性回归（MLR）技术。发现PCA和SSE消除了大量可控制的参数以识别最重要的参数，从而影响了PPV。PCA消除了6个参数，而SSE消除了5个参数，对于PCA和SSE，获得的R ²）分别为0.616和0.584。演化了预测器方程，这些方程用于验证另一组21个现场规模试验爆炸的PPV结果。已经发现预测器方程式在预测PPV值方面相当准确。此外，与SSE技术相比，PCA技术提供了高度相关的PPV预测。本文重点介绍了最先进的统计工具在选择影响现场规模爆破中PPV的爆破设计参数中的作用。