ICCM Conferences, The 7th International Conference on Computational Methods (ICCM2016)

Font Size: 
Invited: Damage Location Identification of Simply Supported Steel Truss Bridge Based on Displacement
Shaopu Yang, Jianying Ren, Shaohua LI

Last modified: 2016-05-23

Abstract


Bridge structure damage identification is an important step in bridge structure health monitoring system, but all kinds of damage identification method at present are all complicated and have poor applicability. Therefore, this paper will propose a simple and applicable method of damage identification based on displacement. This has important significance to realize the real-time and exact warning and forecasting the bridge structural health situation. The damage identification indexes are the change percentages of the lower chord panel points maximum deflections and the beam end maximum displacement. The identification model are established respectively using C-Support Vector Classification (C-SVC) and Probabilistic Neural Network (PNN) to identify the damage location, and the two models’ results are analyzed. The numerical example results show that: (1) The damage identification method based on the bridge deflection is feasible. (2) PNN model and SVC model all have good anti-noise capacity and generalization。(3) SVC model is more suitable to be used in site.

Keywords


displacement, damage location identification, SVM, PNN, railway double-track simply supported steel truss bridge

An account with this site is required in order to view papers. Click here to create an account.