摘要
珠三角地区的城际铁路轨道交通系统需要建设大量的盾构隧道,有些隧道需要穿越风化花岗岩与土体混合地层(简称土岩变换地层)。在土岩变换地层中盾构掘进施工时,地层转换造成了盾构机姿态难以调整、超挖以及地面沉降过大等问题;此外,频繁转换的地层条件导致盾构施工刀具磨损严重、换刀频率高等一系列问题,严重降低了施工效率。为此,提出了一种利用混合人工智能模型预测盾构掘进过程性能的计算方法。采用基于遗传算法的自适应模糊推理系统,输入土性参数与前期施工参数,通过自学习与训练来预测施工效率的变化。以穗莞深城际铁路盾构隧道为例,验证了混合模型的适用性和准确性,预测结果表明基于遗传算法的自适应神经模糊推理系统模型的精度较高,可以有效指导施工,提高掘进效率。
There are a large number of shield tunnels in the construction of the intercity railway connecting Guangzhou,Dongguan and Shenzhen,which is constructed in the Pearl River Delta Region and need to pass through the mixed strata of weathered granite and soil(soil-rock mixed strata for short).Shield construction is carried out in the soil-rock composite strata.The upper soft soil layer makes it difficult to adjust the direction of the shield machine,causing over excavation and large ground settlement.Moreover,the frequent conversion of formation conditions leads to a series of problems,such as serious cutter wearing and high frequency of cutter change,which seriously reduces the construction efficiency.In this paper,a hybrid artificial intelligence model was proposed to predict the performance of shield tunneling.In this model,the adaptive fuzzy reasoning system based on genetic algorithm was employed and soil parameters and construction parameters were input.The change of advance rate was predicted through self-learning and training.With the example of a shield tunnel section of the intercity railway connecting Guangzhou,Dongguan and Shenzhen,the applicability and accuracy of the hybrid artificial intelligence algorithm were verified,and the predicted results showed that the accuracy of the GA-ANFIS model was high enough to be applied to the practical engineering project.
作者
罗杰
LUO Jie(China Railway 16th Bureau Group Beijing Metro Engineering Construction Co.Ltd.,Beijing 100018,China)
出处
《铁道建筑技术》
2020年第S01期160-164,共5页
Railway Construction Technology
基金
中铁十六局集团科技研究开发计划课题(ZT16-SZGJHZ-JK-01)
关键词
土压平衡盾构
遗传算法
自适应神经模糊推理系统
掘进速度
earth pressure balance shield
genetic algorithm
adaptive neuro-fuzzy inference system
driving rate
作者简介
罗杰(1977-),男,湖南常德人,高级工程师,主要从事城市地下轨道交通施工技术研究,E-mail:luojie77@outlook.com
