城市隧道地表沉降测量优化预测技术研究

Study on Optimizing Prediction Technology in Surface Settlement Measurement of City Tunnel

为了提高城市隧道地表沉降的预测精度,在传统的神经网络模型中,引入一种有预测功能的灰色模型,优化传统的地表沉降预测方法。该方法首先利用岩土体地表沉降的随机过程建立模型,通过建立起地表沉降与随机变量之间的非线性映射关系,使得神经网络就在地表沉降的预测中,具备非线性数据逼近能力,保证测量的准确性。试验表明,该方法提高了城市隧道地表沉降的预测精度。

Abstract ： In order to improve prediction accuracy of the surface settlement for city tunnel, the authors put forward a kind of gray space model, which can optimize traditional prediction method of surface settlement. Firstly, the method is to build the model by random process of surface settlement of rock and soil mass, the neural network has nonlinear data approximation ability in surface settlement prediction by building the nonlinear mapping relationship between surface settlement and random variables. The test shows that combination model improves prediction accuracy of surface settlement for city tunnel.