基于回归分析在爆破振动速度预测中的应用与研究

Application and Study of Regression Analysis on Blasting Vibration Velocity Prediction

为研究露天矿爆破开采地震波的变化规律,以江西德兴露天铜矿爆破开采工程为背景,运用线型萨氏公式、高程改进式与多元二次非线性回归拟合等方法对比研究振动速度矢量和,并结合速度方向性、主频全面研究振速的变化特性。结果表明:在该矿山工况下,发生高程效应但很微小,近200 m距离内垂向速度作用大于径向速度为主要因素;预测速度矢量和时,多元二次非线性回归拟合精确性为19%高于萨式公式和高程改进式;在各向速度拟合中,高程改进式拟合垂向速度精确性较好为25%,其它两向速度中多元二次非线性回归拟合比其他两种方法提高10%。建立完整的振速规律回归模型,有利于工程的技术优化与振动作用的控制。

In order to study the seismic waves regularity in the process of mining blasting,the sardowski formula,elevation improvement formula and multiple quadratic nonlinear regression fitting method were applied to analysis the vibration velocity vector sum. Besides,the characteristic of vibration velocity was also studied by velocity directivity and main frequency. The results showed that the elevation effect appeared,but it＇s small. When the distance between blasting charge and testing point was within 200 m,the vertical velocity was main factor and bigger than radial velocity. For velocity vector sum,the accuracy of multiple quadratic nonlinear regression fitting was 19%,which was higher than the results calculated by the linear sardowski formula and elevation improvement formula. As to vertical velocity fitting,the elevation improvement formula was the best,and the accuracy was nearly 25%. And the accuracy of the results by multiple quadratic nonlinear regression fitting of other two direction velocity was 10% over the other two methods. A conclusion could be drawn that establishing a complete regression model of blasting vibration velocity was contributed to the engineering technical optimization and blasting vibration control.