【摘要】 以上海地区某复杂深基坑工程为分析对象,将现场监测数据与神经网络方法相结合,对施工监测当天以后第三次的监测数据做出预测。结合施工动态风险预测方法,按实测监测数据和预测变形数据分别计算出相应的风险等级。本文特点是将预测的变形数据转化为风险水平,可以使施工人员根据不同的风险水平制定相应的预防措施,从而为深基坑工程的施工安全预警提供依据。通过本文的预测分析,可以较为直观地对深基坑工程的施工风险进行定量描述,对施工风险的发展趋势做出合理的预判。更多还原

【Abstract】 In this paper,based on the collected on-site monitoring data of a complex deep excavation in Shanghai,a neural network model is proposed to predict the third time monitoring data of the day of construction.Combined with analysis of dynamic risk,based on the measured and predicted monitoring data,the corresponding risk levels for both data can be determined,respectively.The new feature of this study is transforming the predicted data into predicted risk,which will help the practical engineers to make appropriate preventive measures in advance according to different levels of risks and provide the basis for security alerts of deep excavation construction.Through examples presented in this paper,the present model provides a more direct description of the risk levels of deep foundation projects,and produces a reasonable prediction of the developing trend of the risk.更多还原