基于BP神经网络模型的爆破飞石最大飞散距离预测研究

摘要

首先将BP神经网络模型引入爆破飞石距离的预测研究,以最小抵抗线、炸药单耗、单孔最大药量作为影响爆破飞石最大距离的主要因素,建立了爆破飞石预测的BP神经网络模型,然后以某露天矿山深孔台阶松动爆破为例,利用爆破施工过程中收集的原始资料和爆破飞石监测数据,对建立的BP神经网络模型进行了训练,最后应用经训练的BP神经网络模型对爆破飞石距离进行了预测.与实测值比较后发现,BP网络模型的预报结果非常接近实测值,能够满足工程实践的要求,是一种有效的预测爆破飞石最大距离的方法.%BP neural network was applied to predict the distance of blast flyrock. The minimum burden, the specific charge and the maximum explosive quantity of single hole were considered as the main factors to establish the back-propagation neural network model. Furthermore, the BP neural network was trained, by taking the deep-hole loosening bench blasting in a open-pit mine as an experimental object by using the raw information and monitoring data of blast flyrock collected in the detonation process. Finally, the trained model was applied to predict the maximum distance of blast flyrock. Results show that the forecast data by the BP neural network model was close to the actual case. The model gives an effective way to predict the maximum distance of flyrock.