Please use this identifier to cite or link to this item: http://localhost:8081/xmlui/handle/123456789/8804
Title: COMPUTER AIDED DESIGN OF .INTERIOR PANEL OF GIRDER BRIDGE USING ARTIFICIAL NEURAL NETWORKS
Authors: Giri, V.
Keywords: CIVIL ENGINEERING;COMPUTER AIDED DESIGN;GIRDER BRIDGE;ARTIFICIAL NEURAL NETWORKS
Issue Date: 2006
Abstract: In civil engineering, there are a number of situations where sufficient data is available in the form of design aids, charts and tables but the use of these data requires manual reading of data and hence development of fully computer based design procedure becomes difficult. On the other hand, by using available data, the Artificial Neural Network (ANN) can be trained and once training is over, the developed network can be used directly in the design procedures. The objective of the present work is to develop a computer aided design model for the design of interior panel subjected to patch loading. A model developed by using ANNs, which are trained for data obtained from Pigeaud's curves for getting moment coefficient in short span (ml) and moment coefficient in long span (m2) of the interior panel of girder bridge subjected to patch loading. For different span ratios (K), ANN out put shows good match with target values. Inherently the artificial neural network is a computationally efficient tool and hence the developed computer model will be very useful in optimum design of interior panel of girder bridge. The automation of the design of interior panel has been done using C++ language with the use of the developed ANN models. for prediction of moment coefficients. The results are promising and the developed computer model is quite. useful to speedup the design process of bridge deck slab.
URI: http://hdl.handle.net/123456789/8804
Other Identifiers: M.Tech
Research Supervisor/ Guide: Upadhyay, Akhil
metadata.dc.type: M.Tech Dessertation
Appears in Collections:MASTERS' DISSERTATIONS (Civil Engg)

Files in This Item:
File Description SizeFormat 
CEDG12569.pdf3.15 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.