论文部分内容阅读
为了尽量减小由隧道开挖引起的地面沉降而带来的风险,需要在隧道施工过程中可靠地预测地表的变形量。该文采用改进的方法来选择平移和伸缩因子的初始值,利用小波神经网络分析预测隧道施工中的地表沉降量,并在预测中考虑了地表平均压力、盾构机平均穿透深度、填充泥浆度等外界因素对地表沉降的影响。结果表明,利用改进的方法来选择初始的平移和伸缩因子,提高了函数的逼真性能,并减小了估计误差。
In order to minimize the risk of land subsidence caused by tunnel excavation, it is necessary to reliably predict the deformation of the ground surface during tunnel construction. This paper adopts an improved method to select the initial values of translation and scaling factors, predicts surface subsidence in tunnel construction by using wavelet neural network, and considers average surface pressure, average penetration depth of shield machine, Degree and other external factors on the impact of surface subsidence. The results show that using an improved method to select the initial translation and scaling factors, the real performance of the function is improved and the estimation error is reduced.