基于GA⁃BP模型的复合地层掘进参数研究  

Research on Tunneling Parameters of Composite Strata Based on GA⁃BP Models

作　　者：李卓遥 LI Zhuoyao(China Railway 19th Bureau Group Rail Traffic Engineering Co.,Ltd.,Beijing 101320,China)

机构地区：[1]中铁十九局集团轨道交通工程有限公司,北京101320

出　　处：《施工技术（中英文）》2024年第13期19-25,共7页Construction Technology

基　　金：中铁十九局集团有限公司科技研究开发项目:地铁盾构隧道基岩起伏凸起段控制爆破预处理技术研究与应用。

摘　　要：为探索混合花岗岩复合地层地质参数与盾构掘进参数的关系,依托东莞地铁1号线隧道工程,基于松山湖站—大朗站区间地质勘探数据与现场掘进数据,采用BP神经网络构建预测模型,并用遗传算法对模型进行优化,实现了从地层参数到掘进参数的预测。结果表明:通过数据预处理、遗传算法优化后,平均误差小于6%,最大误差小于25%,基本达到工程要求。通过遗传算法优化的BP神经网络模型预测掘进参数,可以很好地指导实际工程,对未来盾构机无人化、智能化施工提供参考。To explore the relationship between geological parameters of mixed granite composite strata and shield tunneling parameters,based on the tunnel project of Dongguan Metro Line 1,the geological exploration data and on⁃site tunneling data of Songshanhu Station⁃Dalang Station,a BP neural network is used to construct the prediction model,and the genetic algorithm is used to optimize the model and realize the prediction from stratum parameters to tunneling parameters.The results show that after data preprocessing and genetic algorithm optimization,the average error is less than 6%,and the maximum error is less than 25%,which basically meet the engineering requirements.The BP neural network model,optimized by a genetic algorithm,is used to predict the tunneling parameters,which can guide the actual project well and provide references for the unmanned and intelligent construction of the shield machine in the future.

关 键 词：地铁 盾构 复合地层 神经网络 遗传优化 地层参数 掘进参数 预测 

分 类 号：U459[建筑科学—桥梁与隧道工程]