(1 北京工业大学空间结构研究中心，北京 100124；2 北京工业大学城市与工程安全减灾省部共建教育部重点实验室，北京 100124)

［摘要］为了提高诊断效率与诊断结果的可靠性，根据空间网格结构的构成特点提出了面向子结构的损伤定位方法，即根据网格结构的组成规律，将其细分成子结构，采用概率神经网络识别损伤可能发生的子结构。以某单层柱面网壳试验模型为例进行模拟损伤的定位研究，论述了训练样本确定的优先准则，为提高损伤诊断的准确率采取了并集策略。计算结果表明，面向子结构的初步损伤定位法对大型网格结构进行损伤定位是可行的，为现有的面向节点的损伤定位提供了有效的补充，且该方法具有一定的工程实用价值。

［关键词］空间网格结构; 单层柱面网壳; 面向子结构; 损伤定位; 概率神经网络

中图分类号：TU393.3      文献标识码：A      文章编号：1002-848X(2014)06-0079-06

Substructure-oriented damage localization method of the spatial lattice structure

Liu Caiwei1, Zhang Yigang1, 2, Wu Jinzhi1

 (1 Spatial Structures Research Center, Beijing University of Technology, Beijing 100124, China; 2 Key Laboratory of Urban Security and Disaster Engineering of China Ministry of Education, Beijing University of Technology, Beijing 100124, China)

Abstract: In order to improve the damage diagnosis efficiency and reliability of damage diagnosis results, substructure-oriented damage localization method was proposed based on the characteristics of spatial lattice structures. The lattice structure was subdivided into substructures based on the formation rule of lattice structure and probabilistic neural network was used to locate the substructures that were possible to be damaged. Take a single-layer cylindrical reticulated shell test model as an example, simulated damage localization study was carried out. The priority criteria of training samples were discussed and union strategy was taken in order to improve the diagnostic accuracy. The results show that the substructure-oriented preliminary damage localization to detect the defect position in the large lattice structure is feasible and the method has a certain practical value, providing an effective supplement for substructureoriented damage localization method.

Keywords: spatial lattice structure; single-layer latticed cylindrical shell; substructure-oriented; damage localization; probabilistic neural network

*国家自然科学基金资助项目（51278009）。

作者简介：刘才玮，博士研究生，Email:03150053@163.com。

参考文献

［1］BIJAN SAMALI, ULRIKE DACKERMANN, JIANCHUN LI. Location and severity identification of notch-type damage in a two-storey steel framed structure utilizing frequency response functions and artificial neural network［J］. Advances in Structural Engineering, 2012, 15 (5):743-757.

［2］张毅刚,吴金志,梁鑫.面向节点的网格结构损伤定位方法［J］.北京工业大学学报, 2004,30(2): 171-175.

［3］JIANG SHAOFEI, ZHANG CHUNMING, YAO JUAN. Eigen-level data fusion model by integrating rough set and probabilistic neural network for structural damage detection［J］. Advances in Structural Engineering, 2011, 14 (2):333-350.

［4］张育智,李乔,单德山.复杂结构损伤识别的广义子结构法［J］.西南交通大学学报, 2009,44(2): 160-165.

［5］LAM H F, KO J M, WONG C W. Localization of damaged structural connections based on experimental modal and sensitivity analysis［J］. Journal of Sound and Vibration, 1998, 210 (1):91-115.

［6］LI HUIJUN,LIU CHUNGUANG，JIA LINGLING. Reliability and sensitivity analysis of double-layer spherical latticed shell［J］. International Journal of Space Structures, 2011, 26(1): 19-29.

［7］吴金志. 基于动力检测的网格结构损伤识别研究［D］. 北京：北京工业大学, 2005.

［8］LI JIANCHUN, DACKERMANN U,XU YOULIN,et al. Damage identification in civil engineering structures utilizing PCA-compressed residual frequency response functions and neural network ensembles［J］. Structural Control and Health Monitoring, 2011, 18(2): 207-226.