摘要
为合理控制露天转地下开采爆破振动效应,以大冶铁矿露天转地下开采中深孔爆破工程为实例,综合运用萨道夫斯基公式、考虑高程影响的爆破振动速度预测公式及人工BP神经网络方法,对边坡爆破振动速度进行预测研究,并与现场爆破振动监测结果进行了对比分析。研究结果表明:在存在高程影响的矿山边坡爆破振动速度预测过程中,采用三种预测方法对比发现,BP神经网络模型在爆破振动速度切向、径向、竖向三个方向的预测误差率均在6%以内;同时采用考虑高程影响的改进公式预测时在Z方向上具有较高的精确性,误差率仅为11.89%;而萨道夫斯基公式精确性相对最差。研究结果可用于预测及控制露天转地下开采矿山边坡爆破振动速度。
For reasonable control of the blasting vibration effect in open-pit to underground mining transition, taking Daye iron mine open-pit longhole blasting of underground mining engineering as an example ,the Sadaovsk formula and the effect of elevation and artificial BP neural network method were used to predict the blasting vibration velocity, and finally compared with the site monitoring results. The analysis results show that, among the three kinds of forecasting method, in the prediction of mine slope blasting vibration velocity where elevation effect exists, the BP neural network model in the tangential, radial, vertical direction of three prediction error rate was within 6% ; the modified formula has high accuracy in the Z direction and the prediction error rate was only 11.89% ;and the Sada- ovsk formula shows the wamt accuracy. The resuhs have a guiding role for the prediction and control of the slope blasting vibration velocity in open-pit to underground mining.
出处
《爆破》
CSCD
北大核心
2015年第4期49-54,共6页
Blasting
基金
国家自然科学基金资助项目(41372312
51379194)
湖北省自然科学基金(2014CFB894)
中央高校基本科研业务费专项资金资助项目(CUGL140817)
中国博士后科学基金资助项目(2014M552113)
岩土钻掘与防护教育部工程研究中心开放基金(201401)
关键词
BP神经网络
爆破振动
预测
露天转地下
BP neural network
blasting vibration
prediction
open-pit to underground
