论文部分内容阅读
文章研究并搭建了隧道施工过程中包括传感器、人工监测及地质超前等信息的多元信息的采集系统,将所搭建的系统应用到抚松隧道工程实际中,对采集到的多元信息进行了综合分析。另一方面,将差异进化(DE)算法分别于BP神经网络算法及有限元法(FEM)相结合,实现了DE-ANN算法及DE-FEM反演程序。运用其中的DE-ANN算法对采集的传感器信息进行了时间序列预测,用其中的DE-FEM反演程序对抚松隧道大量断面进行了力学参数反演工作,指导了该工程的施工。
This paper studies and builds a multi-information collection system including sensors, manual monitoring and geological advancement in the process of tunnel construction. The system is applied to the Fusong tunnel project, and a comprehensive analysis of the collected multi-information is carried out . On the other hand, the DE-ANN algorithm and the DE-FEM inversion program are realized by combining the differential evolution (DE) algorithm with the BP neural network algorithm and the finite element method (FEM) respectively. The DE-ANN algorithm was used to predict the collected sensor information, and the inversion of mechanical parameters of a large number of sections of Fusong tunnel was carried out by using the DE-FEM inversion program to instruct the construction of the project.