基于动态贝叶斯网络的钢桥检测和维修优化研究

Optimization Study of Steel Bridge Inspection and Maintenance Based on Dynamic Bayesian Network

钢桥作为桥梁的一种主要形式,在车辆与风荷载等反复作用下,极易产生疲劳破坏。为了实现可持续发展,既要保证桥梁结构在整个使用寿命期内安全运营,又要使得养护成本最低,应进行钢桥的检测与维护优化研究。本文提出了一种钢桥检测和维修优化框架,使用GeNIe软件建立了动态贝叶斯网络模型,并用Paris公式来验证了此模型计算的正确性。研究结果表明,当裂纹扩展速度较为缓慢时,使用动态贝叶斯网络模型计算的值较为准确;当裂纹扩展速度较为迅速时,使用动态贝叶斯网络模型计算的值与Paris公式计算的值会存在偏差,但偏差并不大,在可接受的范围内。

As a main form of bridges,steel bridges are prone to fatigue failure under repeated actions of vehi-cle and wind loads.In order to achieve sustainable development,it is necessary to ensure the safe operation of the bridge structure throughout its service life,and to minimize the maintenance cost.Thus,optimization study of steel bridge inspection and maintenance should be performed.In this paper,an optimization framework for inspec-tion and maintenance of steel bridges was proposed.The dynamic Bayesian network model is established using software Genie,and the model is validated using the Paris formula.The results show that when the crack propaga-tion speed is slow,the value calculated by the dynamic Bayesian network model is more accurate;When the crack growth rate is relatively fast,there is a deviation between the values calculated by the dynamic Bayesian network model and the Paris formula,however,the deviation is not large and still within the acceptable range.