基于滑窗OPTICS算法和DATA-SSI算法的桥梁模态参数智能化识别

Intelligent identification of bridge modal parameters based on sliding-window OPTICS algorithm and DATA-SSI algorithm

针对现有基于数据驱动的随机子空间(data-driven stochastic subspace identification,DATA-SSI)算法存在的不足,无法实现稳定图中真假模态的智能化筛选,提出了一种新的模态参数智能化识别算法。首先通过引入滑窗技术来实现对输入信号的合理划分,以避免虚假模态和模态遗漏现象的出现;其次通过引入OPTICS(ordering points to identify the clustering structure)密度聚类算法实现稳定图中真实模态的智能化筛选,最后将所提算法运用于某实际大型斜拉桥主梁结构的频率和模态振型识别过程中。结果表明,所提改进算法识别的频率值结果与理论值(MIDAS有限元结果)以及实际值(现场动力特性实测结果)间的误差均在5%以内,且识别的模态振型图与理论模态振型图具有很高的相似性。

Here,aiming at shortcomings of existing data-driven stochastic subspace identification(DATA-SSI)algorithm and their being unable to realize intelligent screening of true and false modes in stability diagram,a new intelligent identification algorithm for modal parameters was proposed.Firstly,the sliding window technique was introduced to realize reasonable division of input signals,and avoid occurrence of false modes and mode omissions.Secondly,OPTICS(ordering points to identify clustering structure)density clustering algorithm was introduced to realize intelligent screening of real modes in stability diagram.Finally,the proposed algorithm was applied in modal frequencies and modal shapes identification process of a certain large cable-stayed bridge main girder structure.The results showed that errors among frequency values identified using the proposed improved algorithm and theoretical values(MIDAS finite element results)as well as actual values(on-site dynamic characteristics measurement results)are within 5%;the identified modal shapes have higher similarity to theoretical modal shapes.