现有混凝土桥梁结构剩余寿命预测方法研究  

Research on Remaining Life Evaluation Methods for Existing Concrete Bridges

作　　者：刘斌云[1] 霍达[1] 

机构地区：[1]北京工业大学,北京100022

出　　处：《工程建设与设计》2008年第2期78-81,共4页Construction & Design for Engineering

基　　金：北京市自然科学基金资助项目(8031001)

摘　　要：在回顾桥梁剩余寿命预测方法的基础上,探讨了利用人工神经网络对现有混凝土桥梁结构剩余寿命进行全桥模糊预测的方法。该方法从全桥的角度综合考虑各种因素影响,利用因果分析图及数学模糊规则,事先归纳出影响混凝土桥梁整体寿命的几种主要因素,并利用自编程序对几种主要因素的相关数据进行神经网络训练。训练结果表明,只要网络结构选择合理,配合正常的检测制度与专门的数据采集系统,其对全桥剩余寿命进行模糊预测的结果是具有实用价值的。On the basis of a review on the remaining life evaluation methods for existing concrete bridges, a full bridge fuzzy prediction process based on artificial neural networks （ANN） theory is put forward and discussed in this paper. Considering all kinds of influence on the remaining life of a full bridge, several main factors are beforehand abstracted according to mathematic fuzzy rule, then the correlative datum of these factors are utilized to be trained by an self-compiled ANN program. The datum training results revealed the ANN program＇s validity in remaining life prediction if only a proper net structure be selected together with a normal examine scheme ＆ a special datum-collecting system.

关 键 词：桥梁 剩余寿命 全桥模糊预测 人工神经网络(ANN) 

分 类 号：U448.33[建筑科学—桥梁与隧道工程] U446[交通运输工程—道路与铁道工程]