基于参数区间反演修正混合模型的混凝土坝位移监控指标确定方法  

Method for determining displacement monitoring index of concrete dam based on the hybrid model considering parameter interval inversion modification

作　　者：魏博文[1] 袁冬阳 李火坤[1] 徐镇凯[1] WEI Bowen;YUAN Dongyang;LI Huokun;XU Zhenkai(School of Civil Engineering and Architecture,Nanchang University,Nanchang,Jiangxi 330031,China)

机构地区：[1]南昌大学建筑工程学院

出　　处：《岩石力学与工程学报》2018年第A02期4151-4160,共10页Chinese Journal of Rock Mechanics and Engineering

基　　金：国家自然科学基金资助项目(51669013,51779115);江西省研究生创新专项资金项目(YC2017–S013)~~

摘　　要：针对混凝土坝位移监控指标拟定中,结构监测时序及性能退化等因素对其敏感性的影响问题,为提高监控指标对混凝土服役安全性态预警的可靠性,结合区间分析理论、粗糙集理论与神经网络原理提出一种基于参数区间反演修正混合模型的混凝土坝位移监控指标确定方法。利用大坝原型观测资料建立混合模型为基础,结合粗糙集理论与神经网络原理建立具备区间参数反演功能的混凝土坝位移监控混合模型,在对混凝土坝坝体和坝基材料参数区间反演的基础上,利用区间反演所得的最不利参数组合拟定混凝土坝位移监控指标。工程实例分析表明,该方法充分融合了粗糙集理论对不确定性数据挖掘与神经网络非线性自分析的优势,具有良好的鲁棒性,反演所得的混凝土坝坝体和坝基变形模量区间参数值合理有效,所拟监控指标为混凝土坝长期服役和运行管理提供了理论依据与决策支持。此外,所建立的分析模型经一定的改进和拓展,可推广应用于其它结构变形等监控指标的拟定。Considering that displacement monitoring index of concrete dam is affected by the sensitivity of structural monitoring time series and performance degradation,in order to improve the reliability of the displacement monitoring index on the early-warning function for safety behavior of concrete dam service,a method for determining displacement monitoring index of concrete dam based on the hybrid model considering parameter interval inversion modification is carried out on the combination of interval analysis theory,rough theory and neural network theory in the paper. The hybrid model with the function of interval parameter inversion is established on the basis of dam prototype monitoring data,which combines rough theory and neural network theory. The material parameter intervals of concrete dam and bedrock are inversed by the model. Concrete dam displacement monitoring index is determined by the most unfavorable combination of parameters inversed by interval inversion. Examples show that the method combines the uncertain data excavation ability of rough theory and self-learning ability on nonlinear problem of neural network,it has good robustness. Material parameter intervals of concrete dam inversed by the method are reasonable and effective,the monitoring index carried out by the method provides a theoretical basis and decision support for long service and operation management of concrete dam. In addition,the method can be used for determining displacement monitoring indexes of other structures after some expansion and improvement.

关 键 词：水利工程 混凝土坝 位移监控 区间分析 粗糙神经网络 指标拟定 

分 类 号：TV31[水利工程—水工结构工程]