Elman神经网络在软土地基沉降分析中的应用

Application of Elman Neural Network to Soft Soil Foundation Settlement Analysis

本文以软土地基原位监测项目的实测资料为基础,应用神经网络在非线性系统建模和仿真等诸多方面的强大功能,引入Elman人工神经网络建立起软土地基沉降分析的数学模型,为工程实践及数值计算提供一种解决方案,探索沉降分析的新思路新方法。通过对不同沉降分析方法所得到计算结果的差别和规律性进行研究,并与实际沉降观测资料进行对比分析和图形拟合,得出该模型能够很好地反映软土地基的沉降规律,可以满足软土地基沉降分析的精度要求,优于其他沉降分析方法,可应用于堤防、大坝、桥梁及道路等类似工程中。

Based on the measured data of in-situ monitoring for the soft soil foundation,using the mathematical model of Elman neural network to settlement analysis.Because powerful function neural network in nonlinear system modeling and simulation and many other features.For the engineering practice and the numerical calculation to provide a solution,and explore new method and new ideas in the settlement data analysis.This paper,research the difference and regularity of variety of settlement analysis results,and comparative analysis with the actual observational data.It is concluded that good resemblances are found for the in-situ monitoring data and the simulation results,meet the accuracy requirements of soft soil foundation settlement a-nalysis.The method’s calculated and measured values of the curve fit better than other analysis methods.The method can be applied to dikes and dams,bridges and roads,and other similar projects.