摘要
常规方法对软土地基沉降的估计存在一定的问题。利用一种新的智能方法来预测软土地基的沉降:通过已经观测到的沉降数据,建立了基于改进观测数据的神经网络预测模型,依此提出了基于部分实测沉降数据来预测短期沉降及最终沉降的方法。算例表明:对高度复杂非线性的土工结构问题,神经网络能够较为精确的解决,而且操作简单、泛化能力强。
Traditional soft foundation settling volume prediction methods have many drawbacks. An intelligent method for soft foundation settling volume prediction was proposed in this paper. Firstly, a neural network prediction method was established based on the modified data observed. Then, short-term and final method was put forward. Case study showsthat, neural network was a efficient tool to settle the complex problems of nonlinear soil structures with simple operation and widely application.
出处
《国外建材科技》
2008年第3期60-62,共3页
Science and Technology of Overseas Building Materials
关键词
软基沉降预测
神经网络
预测模型
settling volume prediction of soft foundation
neural network
prediction model
