基于灰色神经网络模型的基坑开挖引发周边地面沉降预测分析

在城市地下工程建设中,深基坑开挖引起的周围地表土沉降问题越来越受到人们的重视。地表沉降将引起邻近建、构筑物破坏,从而造成经济损失。因此,预测基坑周围土体未来一段时间的沉降,对及时采取治理措施具有重要意义。文章针对GM(1,1)模型地面沉降预测精度较低的问题,利用神经网络对灰色预测模型进行组合,生成灰色神经网络模型,并进行预测分析,结果表明,利用灰色神经网络模型预测的沉降值,比单独的灰色GM(1,1)模型预测的沉降值具有更高的精度。

Analysis of Periphery Ground Subsidence Initiated by Foundation Pit is Excavated Based on Grey Neural Network Model

In city underground engineering construction, deep foundation pit excavate causes top-soil subside around that pay attention to by the people more and more. It causes near building and structure, and so on to destroy to subside in the earth＂ s surface, thus it results in the economic losses. Accordingly, it has important meanings to predict ground subside of foundation pit predict in some time of future to take the administration measure in time. This text aims at the lower precision of ground subsidence predicts of GM（1,1） model, and utilize neural network to make grey prediction model up, to produce getting grey neural network model, and go on prediction analysis. The result indicates, subsiding value that the grey neural network model predicted has higher precision than the subsiding value that the single grey GM（1,1） model predicted.