摘要
为解决新桥矿大块率高、炸药单耗高及爆破效率低等问题。在对爆破工艺改进的基础上设计有限的爆破试验（13组试验）获取样本，并建立BP神经网络预测模型（隐含层节点数取9），以最小抵抗线形、孔间距0、周边孔距Z作为输入因子，以炸药单耗、大块率作为输出因子预测、优选爆破参数。优化推荐W=0．8m、a=1m、Z=0．8m．对应的炸药单耗为0．2001kg／t，仅为原工艺的50％；大块率为5．2091％，仅为原工艺的20％：生产效率提高了约65％。该方法采用有限的试验与智能预测相结合，实现低成本获取真实样本，并提高了预测精度。
In order to solve the problem of high block rate, high unit explosive consumption and low blasting effi- ciency in Xin-Qiao mine, 13 samples were obtained from limited blasting tests on the basis of improved blasting te- chology. The blasting parameters BP neural networks pediction model with 9 hidden layer nodes＇ was established. Taking the minimum burden W, hole spacing a, contour hole distance Z as the input factors and the unit explosive consumption, block rate as the output factor. The recommended parameters were W = 0.8 m,a = 1 m,Z -- 0.8 m,and the explosive specific charge was O. 2001 kg/t, which is 50% of the original process ; the boulder yield was 5. 2091% ,only took up 20% of the original process; the production efficiency was increased by 65%. Combined with finite test and intelligent prediction, the method can achieve better samples with low cost. In addition, the prediction accuracy was improved.
作者
赵彬
张德明
康虔
王新民
ZHAO Bin1,2 ,ZHANG De-ming1 ,KANG Qian3, WANG Xin-min1(School of Resources and Safety Engineering,Central South University, Changsha 410083, China; 2. Minrnetals Exploration and Development Company Limited,China Minmetals Corporation, Beijing 100010, China;3. School of Environment Protection and Safety Engineering, University of South China, Hengyang 421001, China)
出处
《爆破》
CSCD
北大核心
2018年第1期86-89,115,共5页
Blasting
基金
国家自然科学基金（11472311）
湖南省安全开采重点试验室开放基金（201203）
关键词
爆破参数
爆破试验
BP神经网络
优化预测
blasting parameter
blasting test
BP neural network
prediction optimization
