砂土区间地铁盾构施工土体参数反演及其验证

在地铁砂土区间盾构法施工过程中,由于盾构机的掘进和注浆填充对周围的土层有一定的扰动作用,该土层的力学参数产生了一定变化。现场原位试验很难准确测得土体的参数。根据工程重要性和施工的信息化要求,土体的部分力学参数需要利用施工过程中地表沉降监测数据进行位移反演得出。结合沈阳某砂土区间盾构施工监测,利用改进的遗传优化的神经网络算法对其土体参数进行反演,得出较为准确的土体参数,并将其代入数值模型中进行计算验证。结果表明,数值计算结果与监测数据基本吻合。

Inverse Analysis and Verification of Soil Parameters for Metro Shield Construction in Sand layer

The tunneling of the shield machine and the filled grouting could bring disturbance to the surrounding soil in the construction process of metro shield zone, which contributed to some changes of the mechanical parameters of sandy soil. The parameters measured in situ test could hardly be accurate. Parts of the sandy soil mechanical parameters should be got by displacement inverse analysis, which used the monitoring data of surface subsidence during the construction process, according to the importance of the project and the information requirements of the construction. In this paper, we got the more accurate parameters of the shield disturbed sandy soil by using neural network algorithm optimized by genetic algorithm combined with the construction monitoring of sandy shield tunnel in Shenyang, and to verify the parameters by substituting them into the numerical model. The result showed that the calculation result basically coincided with the monitoring data.