基于桥梁挠度监测的斜拉桥拉索损伤识别预测函数法研究

项目名称： 基于桥梁挠度监测的斜拉桥拉索损伤识别预测函数法研究

项目编号： No.51278315

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 建筑科学

项目作者： 苏木标

作者单位： 石家庄铁道大学

项目金额： 80万元

中文摘要： 对于斜拉桥而言，拉索是关键受力构件。在迄今发现的斜拉桥病害中，拉索出现损伤的病害占绝大部分。如何实时掌握斜拉索病害损伤，科学评估拉索健康状态，对保证斜拉桥运营安全具有重要的理论意义和实用价值。本项目拟研究一种基于桥梁挠度监测的斜拉桥拉索损伤识别新方法，主要研究内容包括：⑴模拟计算斜拉桥拉索损伤前后在多种荷载（含温度、风）工况下梁体挠度及索力，分析梁体挠度和索力随环境温度、风、车辆荷载、拉索损伤情况而变化的规律以及挠度对拉索损伤的敏感性；⑵采用支持向量机构建以实测梁体挠度、环境温度、风、车辆荷载为自变量，索力为因变量的预测函数；⑶提出基于梁体挠度监测数据识别斜拉索是否损伤、损伤位置和损伤程度的判决标准和方法；⑷斜拉桥模型试验，验证拉索损伤识别方法的有效性。本项目研究可为斜拉桥管理和养护部门及时、准确识别斜拉索损伤，决定是否需要换索提供一种经济、实用、可操作性强的新方法，具有广阔的应用前景。

中文关键词： 斜拉桥；挠度监测；预测函数法；支持向量机；损伤识别指标

英文摘要： Stay cables are the key tension members for a calbe-stayed bridge. Up to now, the damage of stay cables are primary in the diseases of cable-stayed bridges. How to master the cable damages in time and scientifically assess the health state of cables has an important theoretical significance and practical utility to guarantee the operation safety of a stayed-cable bridge. In this project, a new damage identification method for cables of calbe-stayed bridges based on deflection monitoring will be researched. The main research contents include: (1) to simulate and calculate the tension of cables and the deflection of the girder when cables of the cable-stayed bridges are in good or damaged condition under a variety of loads (including temperature and wind), to analyse the change rule of the deflection of the girder and the cable force with the variation of ambient temperature, wind, vehicle loads and the damage of cables, and to study the sensibility of the deflection on the damage of cables; ⑵ with support vector machine, to establish the predictive function whose independent variables are deflections of the girder, the ambient temperature and the loads, and dependent variables are the tension of cables; (3) to propose the judgment standard and methods for identifying whether stay-cables are damaged, and detecting

英文关键词： Cable-stayed bridge；Deflection monitoring；Prediction function method；Support Vector Machine(SVM)；Damage identification index