基于多源影响线信息融合的梁桥损伤诊断方法研究

A damage identification method of beam bridges based on multi-source influence line information fusion

摘要:

在梁桥损伤评估与性能评价时,针对使用影响线类方法存在单一测点数据导致识别结果的鲁棒性和准确性较差的问题,提出一种基于多源影响线信息融合的梁桥损伤识别新方法,并利用某实际钢混组合梁桥进行仿真验证。首先通过提取挠度、应变、转角、支座反力等影响线,分别构建影响线曲率差指标,然后基于D-S理论对多源影响线信息进行融合,并引入欧式距离赋权来提高距离测点空间更近处的置信程度,提出一种可用于梁桥结构损伤识别的融合指标,最后结合模型算例,计算得到包含四种不同数据的融合指标,讨论其测点位置、〖JP3〗损伤程度和噪声水平对损伤诊断结果的影响。研究表明:多源影响线信息融合损伤识别方法比传统单一损伤识别在定位、定性上更精确;融合转角、支座反力、应变信息的损伤识别模式效果明显优于其他融合模式,且在10%噪声干扰下仍具有较好的噪声鲁棒性。

Abstract:

For damage identification and performance evaluation of beam bridges, the data at a single measuring point will result in poor robustness and accuracy of identification results. To address this problem, a new damage identification method based on the information fusion of a multi-source influence line was proposed and verified by an actual steel-concrete composite beam bridge. The curvature differences of the influence lines of deflection, strain, rotation angle, and support reaction force were constructed. Then, based on the D-S evidence theory, the multi-source influence line information was fused, and a fusion index that could be used for damage identification of beam bridge structures was proposed by introducing Euclidean distance weighting. Finally, the fusion indexes containing four kinds of data were obtained by taking the model example, and the effects of measuring point location, damage degree, and noise level on the damage diagnosis results were discussed. The results show that the damage identification method based on the fusion of multi-source influence line information is more accurate than the traditional method. The damage identification model with the combination of angle, support reaction force, and strain information is obviously superior to other fusion modes, and still has good noise robustness under 10% noise interference.