| 引用本文: | 樊学平,吕大刚.基于BDNM的桥梁结构可靠度预测[J].哈尔滨工业大学学报,2014,46(2):1.DOI:10.11918/j.issn.0367-6234.2014.02.001 |
| FAN Xueping,L Dagang.Reliability prediction of bridge structures based on BDNM[J].Journal of Harbin Institute of Technology,2014,46(2):1.DOI:10.11918/j.issn.0367-6234.2014.02.001 |
|
| 摘要: |
| 为了结合监测极值应力和应力参数(平均值)的先验模型来对桥梁可靠度进行预测, 认为极值应力随时间变化的动态测量为一个时间序列, 并考虑到贝叶斯动态线性模型(BDLM)的局限性, 引入贝叶斯动态非线性模型(BDNM)对时变极值应力进行预测. 运用BDNM建立了极值应力的状态方程和监测方程, 通过泰勒级数展开技术,将其近似转化为贝叶斯动态线性模型(BDLM), 并通过贝叶斯因子来对应力信息进行监控, 然后结合应力参数的先验信息, 对极值应力的状态参数进行贝叶斯后验概率推断, 建立动态模型对极值应力变化趋势进行预测. 基于监测信息, 考虑到变量估计主观认识的不确定性, 引入折扣因子来确定状态误差方差. 最后利用建立的BDNM和一次二阶矩(FOSM)可靠度方法, 对结构可靠度进行预测, 并通过实例验证了所建模型的合理性和适用性. |
| 关键词: 桥梁健康监测 应力 贝叶斯动态非线性模型 贝叶斯动态线性模型 可靠度预测 |
| DOI：10.11918/j.issn.0367-6234.2014.02.001 |
| 分类号: |
| 基金项目:国家自然科学基金资助项目(50678057). |
|
| Reliability prediction of bridge structures based on BDNM |
|
FAN Xueping, L Dagang
|
|
(School of Civil Engineering, Harbin Institute of Technology, 150090 Harbin, China)
|
| Abstract: |
| To predict the bridge structural reliability based on the monitoring information and the priori model of stress parameters (mean), the dynamic measure of structural stress over time is treated as a time series, and considering the limitation of the BDLM, a Bayesian dynamic nonlinear model (BDNM) is then introduced. State equation and monitoring equation of monitoring stress are established with BDNM. Then the BDNM is approximately transferred into Bayesian dynamic linear model (BDLM) by Taylor series expansion technique, and the monitoring information is monitored by bayes factor. Combining parameters’ prior information with the early stress data containing noise, the stress state parameters are deduced with Bayesian Posterior Probability. A dynamic model is built to forecast the changing trend of structural stress. To allow for the epistemic uncertainty in variance estimation based on monitoring information, a discount factor approach is made for specification of unknown variance. Finally based on the built BDNM and the FOSM method, the structural reliability is predicted, and the feasibility and application of the built model is illustrated by an actual example. |
| Key words: bridge structural health monitoring stress BDNM BDLM reliability prediction |
