斜拉桥结构损伤识别的概率可靠度法

基于静力响应测试数据,结合正则化技术和假设检验,提出了测试数据不完整且存在观测噪声条件下斜拉桥结构损伤识别的概率可靠度法。采用正则化技术降低斜拉桥结构损伤识别问题的不适定性和求解困难,综合利用参数估计和数据摄动技术获得结构参数的统计特征,在此基础上根据假设检验识别损伤位置和损伤程度。以一斜拉桥结构的损伤识别仿真分析验证了方法的正确性和有效性。研究表明,所提出的方法能够在测试数据存在观测噪声的情况下准确识别损伤。

Statistical Damage Assessment Method of Cable-stayed Bridges Based on Static Response Test Data

The method of statistical structural damage assessment for cable-stayed bridges is developed based on static response test data, which is uncompleted and noise-polluted. The regularization technique is adopted to alleviate the ill-posedness and solution difficulty of the damage assessment problem. The parameter estimation technique and data perturbation method are introduced to obtain the statistical characteristics of structural parameters. The hypothesis test approach is employed to locate damages and assess their severity. A numerical simulation case of a cable-stayed bridge is studied to demonstrate the validity and feasibility of the suggested technique in actual engineering application. It is demonstrated that the method presented in this paper can assess damage correctly in the case of test data polluted by measurement noises.