基于桥梁应变影响线的车辆荷载自动识别理论与算法研究

The vehicle load identification based on the response of dynamic strain of bridge, which is an significant theoretical foundation of Bridge Weigh-In-Motion technology, has become an increasingly hot topic relevant to bridge health monitoring and testing, condition assessment and load statistics. This project focuses on vehicle load identification and algorithm investigation based on bridge strain influence line; A hybrid function, composed of sectioned polynomial and harmonics with adjusted amplitude, are employed to fit initial influence line for dynamic calibration test. The bridge strain influence line with quasi-static characteristic can be extracted. The axles information automatic identification method based on wavelet transform is proposed to focus on fuzzy initial signal. The obtained axles information can be used to match with the real axles information. Along the line of quasi-static characteristic influence line extraction and axles information automatic identification and matching, the theory and algorithm used to improve the precision of vehicle load identification are further explored from the viewpoint of choice of influence line, vehicle weigh for multi-vehicle, etc. The vehicle-bridge coupling vibration model is built based on surface to surface contact to investigate the effects on vehicle load identification from type of bridge, span, the location of testing, road surface roughness, speed of vehicle, etc. The theory and algorithm of quasi-static influence line extraction and vehicle load automatic identification are examined by numerical simulation and field test. The obtained results can be used for bridge health and safety assessment, bridge design and management. The theoretical and practical value is of importance.

根据桥梁动应变响应进行车辆荷载识别是与桥梁监测、状态评估、荷载统计相关联的热点问题，也是发展桥梁动态称重技术的重要理论基础。本项目开展基于桥梁应变影响线的车辆荷载自动识别理论与算法研究；采用由分段多项式和可调幅谐波组成的混合函数对动态标定测试的初始影响线进行拟合，分离并提取具有准静态特性的桥梁应变影响线；针对模糊原始信号提出基于小波变换的车轴信息自动识别方法，并与真实车轴信息进行匹配分析；在准静态影响线提取研究和车轴信息自动识别与匹配的基础上，从影响线选定、多车辆称重等方面进一步探索提升车辆荷载识别精度的理论与算法；建立基于面-面接触的车-桥耦合振动模型，研究桥梁类型、跨度、测点位置、路面不平度、行车速度等因素对车辆荷载识别的影响；采用数值模拟和现场试验验证准静态影响线提取和车辆荷载自动识别理论与算法。研究成果可为桥梁结构评估、桥梁设计与管理提供依据，具有良好的理论与工程应用价值。

本项目开展基于桥梁应变影响线的车辆荷载自动识别理论与算法研究。采用由分段多项式和可调幅谐波组成的混合函数对动态标定测试的初始影响线进行拟合，分离并提取具有准静态特性的桥梁应变影响线；针对模糊原始信号提出基于小波变换的车轴信息自动识别方法；从影响线选定、多车辆称重等方面进一步探索提升车辆荷载识别精度的理论与算法。本项目主要的创新性成果包括：.（1）桥梁动态称重理论算法深入研究，及B-WIM的适用范围与现阶段的测试精度水平研究。.（2）桥梁固有特征-影响线的识别与提取研究，以车致桥梁动力响应为对象，建立合理的拟合函数模型，采用最小二乘拟合方法计算提取桥梁结构的实际影响线信息。并开展理论和试验验证。.（3）基于信号的桥梁结构的损伤识别研究，结合桥梁实际影响线信息进行损伤识别，相较传统的基于模态的损伤识别而言，该方法所用信号接近实际，更容易准确获得，且多信号结合的优势明显。.项目从B-WIM理论方法的深入研究出发，开展桥梁实际影响线的提取的系列研究，提取的桥梁实际影响线可作为评估桥梁承载能力、桥梁损伤情况研究的重要参考。课题研究成果可实现对桥梁结构和桥上移动车辆荷载的同时实时监测，为桥梁结构评估、管养维护等提供科学依据。.在课题研究期间，课题组在国内外高水平期刊已发表学术论文7篇，其中SCI论文4篇，EI论文3篇；获授权国家专利3项；培养硕士研究生3名。