DocumentCode
3365906
Title
Development of deformation prediction system of large bridge based on MATLAB and radial basis function neural network
Author
Tian, Linya ; Wang, Gang ; Yu, Xiaotao
Author_Institution
Coll. of Geosci. & Eng., Hohai Univ., Nanjing, China
fYear
2010
fDate
26-28 June 2010
Firstpage
4613
Lastpage
4616
Abstract
Affected by many factors, deformation of large bridge will take place in the course of usage, so the development of deformation prediction system will contribute to the maintenance and management of large bridge. As the multiple-input and multiple-output characteristic of the RBF neural network, which also has the nature of universal approximation and optimal approximation, MATLAB combined with RBF network was choosed to develope the deformation prediction system of large bridge. The improved algorithm of RBF network parameters was studied, some key issues in the system development were solved, the choice of input factors and the preprocessing method of sampled data were discussed, the main function of the system was elaborated. This system has some popularization and application values to the deformation prediction and safety supervising of other similar projects.
Keywords
bridges (structures); condition monitoring; deformation; mathematics computing; radial basis function networks; structural engineering computing; MATLAB; deformation prediction system development; large bridge; optimal approximation; radial basis function neural network; sampled data preprocessing method; universal approximation; Bridges; Data preprocessing; Educational institutions; Electronic mail; Engineering management; Geology; MATLAB; Neural networks; Radial basis function networks; Safety; bridge; deformation predication system; development; matlab; radial basis function neural network;
fLanguage
English
Publisher
ieee
Conference_Titel
Mechanic Automation and Control Engineering (MACE), 2010 International Conference on
Conference_Location
Wuhan
Print_ISBN
978-1-4244-7737-1
Type
conf
DOI
10.1109/MACE.2010.5536600
Filename
5536600
Link To Document