Structural Strength of Corrugated Steel Web Box Beam Based on Finite Element Analysis

Abstract

Deflection prediction in composite box-girder bridges with corrugated steel webs may be limited in a number of ways. Accurately predicting the complicated behaviour of such structures, including the intricate interaction of diverse materials and geometries, can be difficult and may result in inaccurate predictions. To overcome these issues, in this manuscript Structural Strength of Corrugated Steel Web Box Beam Based on Finite Element Analysis (SS-CSWBB-RBAGCN)is proposed. Originally, the Structural Defects Network (SDNET) 2018 dataset is used to gather the input data. Then, the collected data is fed into pre-processing utilizingSurface Normal Gabor Filter (SNGF).SNGF is used for data cleaning and to find missing values. Then the preprocessed data are given to Relational Bilevel aggregation graph convolutional network(RBAGCN) for predicting deflection behavior of the bridge structure. In general, RBAGCN does not express adapting optimization strategies to determine optimal parameters. Hence, the Tyrannosaurus Optimization Algorithm (TOA) to optimize which accurately predict the deflection behavior of the bridge structure. The proposed SS-CSWBB-RBAGCN method covers 29.36%, 23.27% and 26.42% higher accuracy, 18.46%, 20.38% and 15.45% lower Absolute Percentage Error and 18.26%, 28.41% and 29.41% higher Determination Coefficient compared with existing methods such as, Optimizing Shear Capacity Prediction of Steel Beams with Machine Learning Techniques (SCP-SB-XGB), Prediction analysis of deflection in the construction of composite box-girder bridge with corrugated steel webs based on MEC-BP neural networks(DCCBGB-CSW-BPNN),  and Utilizing Artificial Intelligence Approaches to Determine the Shear Strength of Steel Beams with Flat Webs (AI-SSSB-CFBNN)respectively.