工程爆破

2016, v.22(05) 57-63

[打印本页] [关闭]
本期目录(Current Issue) | 过刊浏览(Past Issue) | 高级检索(Advanced Search)

基于爆破振动的岩质边坡损伤神经网络预测
Prediction of blast-induced damage depth for rock slope based on monitored vibration and neural network model

邹玉君;严鹏;卢文波;陈明;王高辉;
ZOU Yu-jun;YAN Peng;LU Wen-bo;CHEN Ming;WANG Gao-hui;State Key Laboratory of Water Resources and Hydropower Engineering Science,Wuhan University;Key Laboratory of Rock Mechanics in Hydraulic Structural Engineering Ministry of Education,Wuhan University;

摘要(Abstract):

岩石高边坡的爆破开挖会对保留岩体造成损伤,岩体损伤过大可能导致边坡失稳,需严格控制并准确确定开挖损伤深度,因此,提出一种快速精确的损伤深度预测方法。以白鹤滩水电站左岸834.0~770.0 m高程坝肩槽边坡爆破开挖为背景,利用六个开挖梯段的多高程、多爆心距爆破振动监测及损伤深度声波检测的数据,建立基于振动峰值的爆破损伤深度BP神经网络预测模型,对高边坡爆破损伤深度进行实时预测。该方法利用不同部位及不同爆心处的质点峰值振动峰值作为主回归变量,同时还考虑最大单响药量和岩体强度的影响。结果表明,当开挖区域坡体岩性相似且无长大软弱结构面发育时,运用神经网络模型及多高程实测爆破振动预测本梯段爆破损伤深度的方法简便可行,预测精度可满足实际工程需求。作为传统爆破损伤声波检测的补充,可大大减轻现场声波测试工作量。
The blasting excavation of high rock slope in large-scale hydropower projects leads to damages on the reserved rock mass.Such damages may cause slope failure,so the blast-induced damage depth should be strictly controlled and precisely determined and it is urgently needed to find an efficient and accurate method to determine damage depth.During blasting excavation of the left bank slope between altitude of 834.0 m and 770.0 m of the Bai-he-tan Hydropower Station,the vibration caused by the first to the sixth bench blasting are monitored at different points and the blasting damage depths are also obtained by sonic wave testing.Then the BP artificial neural network model is established for real-time prediction of damage depth based on monitored vibration.This method takes the vibration at different distances and altitudes to the blast center as main regression variable,and the maximum explosives per delay and rock mass strength are also considered.The result indicates that if the lithology of each bench were similar and there were no large structural planes existing,the method that applying BP artificial neural network model presented with monitored vibration is convenient and feasible.The prediction accuracy of damage depth can meet the requirement of practical project,and the method for supplementury will significantly reduce massive traditional sonic wave testing workload.

关键词(KeyWords): 岩石高边坡;爆破振动;BP神经网络;爆破损伤;实时预测
High rock slope;Blasting vibration;BP neural network;Blasting damage;Real-time prediction

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金杰出青年基金项目(51125037);国家自然科学基金面上项目(51179138)

作者(Author): 邹玉君;严鹏;卢文波;陈明;王高辉;
ZOU Yu-jun;YAN Peng;LU Wen-bo;CHEN Ming;WANG Gao-hui;State Key Laboratory of Water Resources and Hydropower Engineering Science,Wuhan University;Key Laboratory of Rock Mechanics in Hydraulic Structural Engineering Ministry of Education,Wuhan University;

Email:

DOI:

参考文献(References):

扩展功能
本文信息
服务与反馈
本文关键词相关文章
本文作者相关文章
中国知网
分享