复合地层中盾构施工掘进参数相关性及预测研究
Research on the correlation and prediction of shield tunneling parameters in compound strata
何华飞 1林雪冰 1胡朋 1贾柏源 2陈逸民 2宋克志2
作者信息
- 1. 北京市政建设集团有限责任公司,北京 100045
- 2. 鲁东大学土木工程学院,山东烟台 264025
- 折叠
摘要
在盾构掘进过程中,众多掘进参数交织在一起,规律错综复杂.科学地分析这些参数之间的规律与相关性,对于盾构的实时控制至关重要.依托济南地铁R2左线工程,基于机器学习方法对复合地层中盾构施工的掘进参数进行研究.首先,分析地层参数与盾构掘进参数之间的关系,通过相关性分析发现,6个掘进参数中,盾构的总推进力和刀盘扭矩与其他参数的相关性尤为显著.然后,利用支持向量机算法,对不同地层中盾构总推进力和刀盘扭矩进行预测.预测结果与实际监测结果基本一致,证明这种方法的有效性.最后,为进一步提高预测精度,采用PSO算法对盾构掘进参数预测模型进行优化.优化后的模型总推进力R2 提升约3%,刀盘扭矩R2 提升约10%,优化效果十分显著.该项研究可为复合地层中盾构施工参数选择提供重要参考.
Abstract
There are numerous excavation parameters present in the process of shield tunneling,and the principles between excavation parameters are intertwined and complex.Scientific analysis of the principles and correlations between excavation parameters is significant for real-time control of shield tunneling.Based on the R2 left line project of Jinan metro,the excavation parameters of shield construction in compound strata were studied by machine learning methods.First,the relationship between geological parameters and shield tunneling parameters was analyzed,and correlation analysis was used to obtain the correlation between six tunneling parameters,in which the total thrust and cutter head torque of the shield were strongly correlated with other parameters.Then,the support vector machine algorithm was used to predict the total thrust and cutter head torque of the shield in different strata,and the predicted results were generally consistent with the actual monitoring results.Finally,the PSO algorithm was used to optimize the shield tunneling parameter prediction model.The results of the optimized model showed that the total R2 thrust increased approximately 3%,and cutter head torque increased approximately 10%,indicating that the optimization effect is significant.The current research results have certain reference significance for the selection of shield construction parameters in compound strata.
关键词
盾构施工/复合地层/机器学习/掘进参数/相关性分析/支持向量机Key words
shield construction/compound strata/machine learning/tunneling parameter/correlation analysis/support vector machine引用本文复制引用
出版年
2024
NETL
NSTL
维普
万方数据