Font Size: a A A

Construction Control Technique For High-Pier And Long-Span Continuous Rigid-Frame Bridges Based On ANN

Posted on:2006-01-29Degree:MasterType:Thesis
Country:ChinaCandidate:Y XuFull Text:PDF
GTID:2132360182965177Subject:Bridge and tunnel project
Abstract/Summary:PDF Full Text Request
In recent years high-pier and long-span continuous rigid-frame bridges are more and more built due to its merits, which are comfortable travel, easy to be constructed and favorable maintenance. So the bridge design theory, the bridge construction technique and the bridge construction control art are pushed forward. But the progress of the bridge construction control technique is more delay than that of the two formers, because it's restricted by the factor of the immature control theory and the out-of-date monitoring technique. After ANN, i.e. Artificial Neural Network, is used into bridge construction control technique by bridge engineers, a favorable reference and way is provided to bridge construction control. Bridge construction control technique based on ANN is still in the stage of theory discussion and trial period. On the foundation of the above analysis, this dissertation tries to solve some problems emerging during the ANN's application to bridge construction control and to enrich the system of bridge construction control.The vectors of input and output of ANN model are determined randomly and experimentally by bridge engineers at the present time. So the factors causing the elevation errors during geometric control of bridge superstructure are analyzed as emphasis according to the specific details of bridge construction control in this dissertation. After introducing two concepts: Gray Correlations and Principal Component Analysis, this dissertation confirms the number of neural unit of input layer using non-structural analysis technique at last. Not only all kinds of factors leading to elevation errors are on full considerations, but also the aim that ensure all vectors independent, reduce the dimension of all vectors and the scale of network is obtained.This dissertation's achievement is applied to the geometric control of Xu-Shui-Gou Bridge and a satisfied result is got.This dissertation presents temperature adjustment methods for high pier and long cantilever superstructure after analyzing the temperature complex influence to control technique for the construction of high-pier long-span continuous rigid-frame bridges.
Keywords/Search Tags:continuous rigid-frame bridge, construction control, artificial neural network, gray correlations analysis, principal component analysis
PDF Full Text Request
Related items