基于时变监测数据的降雨滑坡多目标随机反分析

发布时间：2018-03-18 03:28

本文选题：边坡　切入点：流-固耦合　出处：《岩土力学》2017年11期 　论文类型：期刊论文

【摘要】：基于非饱和土流-固耦合理论和贝叶斯理论,建立了边坡的非饱和土流-固耦合随机反演模型,提出了基于马尔科夫链的多目标随机反分析方法,利用位移和孔隙水压力时变监测数据进行多目标和单目标随机反演,并对反演结果进行比较分析。结果表明,多目标随机反分析参数后验分布标准差较单目标随机反分析明显减小。单目标随机反分析只对本目标进行优化,对其他目标的预测误差较大。多目标随机反分析能同时对所有目标进行优化,反演结果对所有目标误差均较小,95%置信区间较单目标明显收窄,采用不同类型监测数据的多目标随机反分析所得结果更为可靠,预测更为准确。
[Abstract]:Unsaturated soil flow solid coupling theory and Bias theory based on the established slope of unsaturated soil fluid-solid coupled stochastic inversion model, a multi-objective stochastic back analysis method based on Markov chain, the displacement and pore water pressure variable multi-objective and single objective stochastic inversion and the inversion results of monitoring data. Comparative analysis. The results show that the multi-objective stochastic back analysis parameters of the posterior distribution of standard deviation than the single objective stochastic back analysis is significantly reduced. The single objective stochastic back analysis only to the multi-objective optimization, prediction error on other targets. A multi-objective stochastic back analysis can simultaneously on all targets were optimized, the inversion results all the errors are less than the single target, 95% confidence interval was significantly narrowed, using different types of monitoring data of multi objective stochastic back analysis can provide more reliable results, the prediction is more accurate.

【作者单位】：上海交通大学海洋工程国家实验室;上海交通大学船舶海洋与建筑工程学院;国网上海电力设计有限公司;高新船舶与深海开发装备协同创新中心;同济大学地下工程系;机械工业勘察设计研究院;
【基金】：国家重点基础研究发展计划973项目(No.2014CB049100) 国家自然科学基金项目(No.51422905,No.41372275,No.51679135) 中组部青年拔尖人才计划~~
【分类号】：P642.22;TU433

【相似文献】