基坑监测动态反演与变形预报组合分析模型研究

With the acceleration of urban development, the scale and depth of foundation pit is increasing in practical engineering. As the internal structure and surrounding environment of foundation pit is complex and changeable, the rock mass deformation and physical characteristic is uncertain. Using the measured data of foundation pit, dynamic prediction and information feedback should be supplied to ensure the safety of construction. Through analyzing the characteristic and applicable conditions of the existing processing methods for monitoring data, the perfect prediction theory system for pit monitoring data will be put forward. This research includes the main contents of the following aspects: (1) With the evaluation of the availability for all pit deformation influencing factors by grey relational analysis model and contingency theory, the major influence factors used in analysis models will be confirmed. (2) After the measured data is preprocessed, the different single prediction models will be chosen to combine effectively according to the sample size and sequential characteristic of measured data. And while each model precision is evaluated, more reliable combination models will be added to establish a model library. (3) In order to improve the precision of deterministic model just as finite element or boundary element method, the physical and mechanical parameters of rock mass will be inversed dynamically with nonlinear combination models based on elastic-plastic constitutive model. (4) With embedded development and remote communication technology, the intelligent monitoring system for foundation pit monitoring will be researched, which has the functions of data collecting, data analysis, data transmission and early warning. This system may raise the timeliness and accuracy of safety monitoring and promote the informatization of foundation pit monitoring.

随着城市建设步伐加快，基坑工程规模与深度不断加大。由于基坑内部结构与周边环境复杂多变，岩体变形及物理特征存在不确定性。为确保施工安全，须根据基坑监测数据提供动态预报与信息反馈。本课题将在分析现有监测数据处理方法特点及适用条件的基础上，建立完善的基坑监测数据分析预报理论体系，主要包括以下研究内容：（1）利用灰关联分析与权变理论，评价基坑变形影响因子的有效度，确定纳入预报模型的主要影响因子；（2）对监测数据预处理后，依据监测数据样本大小及时序特征，选取多种预报模型进行组合分析，建立可靠的组合预报模型库；（3）为提高有限元、边界元等确定性模型精度，以弹塑性本构模型为基础，通过非线性组合模型对岩土物理力学参数进行动态反演；（4）利用嵌入式开发与远程通信技术，研究集数据采集、分析、传输、预警等功能为一体的智能监测系统，提高基坑信息化安全监测时效性和准确性。

随着城市建设步伐的加快，基坑工程规模与深度不断加大。由于基坑内部结构与周边环境复杂多变，岩体变形及物理特征存在不确定性。为确保施工安全，须根据基坑监测数据提供动态预报与信息反馈。本项目在分析现有监测技术的基础上，对基坑变形监测数据预处理、影响因素评定、预报方法优化、岩土参数反演、监控系统研制等关键问题展开了深入研究，主要包括以下内容：（1）引入稳健回归分析、小波阈值去噪等模型，建立了抗差性强的监测数据质量控制体系；（2）利用基于斜率的灰关联分析算法，定量评价了基坑变形影响因素的有效度，用以确定纳入预报模型的主要因子；（3）在优化传统变形分析方法的基础上，基于不同单项模型构建了分步预报、加权预报、多因素预报等组合预报模型库；（4）为提高有限元模拟分析可靠性，以本构模型为基础，建立了基于遗传支持向量机的岩土参数反演模型；（5）利用嵌入式开发与WebGIS技术，研制了集数据采集、处理、传输、分析、预警、监管等功能为一体的智能监测系统，提高了基坑信息化变形监测时效性。

内容获取失败，请点击重试