基于改进DATA-SSI和聚类分析的桥梁结构模态参数识别

发布时间：2018-10-19 18:05

【摘要】：桥梁结构的模态参数主要用于桥梁结构损伤识别及使用性能评估,在桥梁健康监测系统中具有十分重要的意义。目前较为常用的模态参数识别算法是随机子空间识别算法,主要包括协方差驱动的随机子空间识别算法和数据驱动的随机子空间识别算法。在利用随机子空间算法识别桥梁结构的模态参数时,首先需人为依据稳定图来进行系统阶次的确定,通过对稳定图定阶的不断研究,发现其依然存在模态失真和计算量大这两方面的缺陷,为此提出了新的定阶算法以实现系统阶次的自适应确定。其次需要人为参与稳定图中真实模态的辨识,这不仅降低参数识别效率,还导致筛选的结果带有一定主观性。为了避免人为参与定阶和真实模态的筛选,提高模态参数识别的效率,在现有随机子空间算法的基础上引入聚类分析算法,提出了一种改进的模态参数自动化识别算法,以实现系统阶次的自动化确定和稳定图中真实模态的自动化筛选。再通过一座试验桥验证了真实模态存在的规律。最后将改进算法运用于识别实际大型斜拉桥的结构参数。结果表明:改进算法能够实现实际桥梁结构的模态参数自动化识别,且识别结果具有可靠性。
[Abstract]:The modal parameters of bridge structure are mainly used for damage identification and performance evaluation of bridge structure, which is of great significance in bridge health monitoring system. At present, the commonly used modal parameter identification algorithms are random subspace identification algorithm, which mainly includes covariance-driven random subspace identification algorithm and data-driven random subspace recognition algorithm. In identifying the modal parameters of bridge structure by using stochastic subspace algorithm, it is necessary to determine the order of the system based on the stability graph. It is found that there are still defects in mode distortion and computational complexity. A new order determination algorithm is proposed to realize the adaptive order determination of the system. Secondly, it is necessary to take part in the identification of real modes in the stability map, which not only reduces the efficiency of parameter identification, but also leads to the subjectivity of the selection results. In order to avoid artificial participation in the selection of order and real modes and to improve the efficiency of modal parameter identification, an improved automatic modal parameter identification algorithm is proposed by introducing clustering analysis algorithm based on existing stochastic subspace algorithms. In order to realize the automatic determination of system order and the automatic screening of real modes in the stability diagram. The existence of real modes is verified by an experimental bridge. Finally, the improved algorithm is applied to identify the structural parameters of large cable-stayed bridges. The results show that the improved algorithm can realize the automatic identification of the modal parameters of the actual bridge structure, and the identification results are reliable.
【作者单位】：浙江工业职业技术学院;
【基金】：浙江省住房与城乡建设厅科研项目(2014Z126) 浙江省教育厅科研项目(Y201432555) 绍兴市科技计划项目(2014B70003)
【分类号】：U446.3

【相似文献】