新桥矿爆破工艺与参数优化

发布时间：2019-02-13 12:22

【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W、孔间距a、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=0.8 m、a=1 m、Z=0.8 m,对应的炸药单耗为0.2001 kg/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。
[Abstract]:In order to solve the problems of high bulk ratio, high explosive unit consumption and low blasting efficiency in Xinqiao Mine, a limited blasting test (13 sets of tests) was designed on the basis of the improvement of blasting technology. The prediction model of BP neural network is established. The minimum resistance line W, hole spacing a, peripheral hole spacing Z are taken as input factors, explosive unit consumption and bulk rate are used as output factors, and blasting parameters are selected. It is recommended that the unit consumption of 0.2001 kg/t, is only 50 of the original process, and the bulk rate is 5.2091, which is only 20% of the original process. The production efficiency has been improved by about 65%. The method combines limited experiments with intelligent prediction to obtain real samples at low cost and improve prediction accuracy.
【作者单位】：中南大学资源与安全工程学院;中国五矿集团公司五矿勘查开发有限公司;南华大学环境保护与安全工程学院;
【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)
【分类号】：TD235.4

【相似文献】