基于灰色分析和神经网络的爆破振速峰值预测

发布时间：2018-05-02 05:00

本文选题：爆破振动 + BP神经网络　；参考：《中国矿业》2016年S1期

【摘要】：爆破给国家带来了巨大经济效益、给施工提供便利条件的同时避免不了会带来一些负面效应,为此,如何才能既降低爆破振动危害又保证施工进度,己经成为当前隧道工程界亟待解决的一项重要课题。以北京地铁昌平二期05段岩石隧道爆破为例,通过灰色关联分析法确定了对爆破振速峰值有显著影响的指标和输入变量,建立BP神经网络模型,对爆破振动速度峰值进行预测。将结合了灰色关联分析法的BP神经网络模型预报的结果与神经网络模型、传统方法预测的结果相比,其结果为:萨道夫斯基公式的平均误差为18.86%,萨道夫斯基拓展式为16.57%,BP神经网络的误差为14.61%,灰色关联分析神经网络法仅为8.23%。预测结果表明结合灰色关联分析法并运用BP神经网络对爆破振速峰值预测是可行的。
[Abstract]:Blasting has brought great economic benefits to the country. It can not avoid some negative effects while providing convenient conditions for construction. Therefore, how to reduce the damage of blasting vibration and ensure the construction progress, It has become an important subject to be solved urgently in the tunnel engineering field. Taking the blasting of rock tunnel in Changping Phase 05 of Beijing Subway as an example, the indexes and input variables which have significant influence on the peak value of blasting vibration velocity are determined by grey relational analysis, and the BP neural network model is established to predict the peak value of blasting vibration velocity. The prediction results of BP neural network model combined with grey correlation analysis are compared with those of neural network model and traditional method. The results are as follows: the average error of Sadolski formula is 18.86, the error of Sadolski expansion is 16.57 and the error of BP neural network is 14.61, and that of grey correlation analysis neural network is only 8.23. The prediction results show that it is feasible to predict the peak value of blasting vibration velocity by combining grey relational analysis and BP neural network.
【作者单位】：北京科技大学土木与环境工程学院;
【基金】：国家自然科学基金项目资助(编号:51208036) 中央高校基本科研业务费专项资金资助(编号:FRF-TP-15-041A3)
【分类号】：U455.6

【相似文献】