山岭隧道钻爆法施工参数控制的优化  

Control optimization of drilling and blasting parameters during mountain tunnel construction period

作　　者：方昱[1,2] 刘开云[1] 

机构地区：[1]安徽省高速公路控股集团有限公司,合肥230088 [2]北京交通大学土木建筑工程学院,北京100044

出　　处：《北京交通大学学报》2015年第1期14-20,共7页JOURNAL OF BEIJING JIAOTONG UNIVERSITY

基　　金：国家自然科学基金资助项目(51378052)

摘　　要：以佛岭长大公路隧道施工为依托,在大量现场测试和光面爆破试验的基础上,引入遗传算法和支持向量回归耦合算法,经对现场测试数据训练,建立了山岭隧道钻爆施工参数、围岩地质条件及爆破后隧道拱顶下沉与围岩松动圈深度之间的非线性支持向量回归智能模型.利用此模型,可以在给定围岩地质条件和拱顶下沉量、松动圈深度等控制指标情况下,利用遗传算法,搜索得到钻爆施工参数的最优解.经佛岭隧道任一断面的实地应用,验证了本文所提出的钻爆施工参数控制优化方法的正确性和可行性.该研究成果为佛岭隧道钻爆施工参数的选择提供了科学依据与技术指导,也可为类似隧道工程的钻爆参数控制优化提供参考.A non-linear support vector regression intelligent model which maps the relation between mountain tunnel drill blasting construction parameters,geological conditions and crown subsidence of surrounding rock and loose circle depth is established ultilizing the genetic algorithm and support vector regression coupling algorithm. The model is applied to the construction of Foling long and large highway tunnel. The model is first trained by the field testing data. Then the model is used to optimize the selection of drill blasting parameters with given control indexes such as geological conditions of surrounding rock,vault subsidence value and loose circle depth. The predicted drill blasting parameters are verified by one random selected cross section in Foling tunnel and the model is proved to be correct and feasible. The model can be used to guide the drill blasting construction in Foling tunnel and can provide references for drilling and blasting parameters control optimization in tunnel with similar conditions.

关 键 词：隧道 钻爆施工 现场测试 人工智能 控制优化 

分 类 号：U455[建筑科学—桥梁与隧道工程]