Study on Carbon Emission Prediction for Steel Frame Shell Structures Based on BP Neural Networks
About the author:
China Construction Sixth Engineering Bureau Civil Engineering Co., Ltd.,Tianjin 300450, China
Abstract:
The design of large span spatial building structures and the magnitude of loads have asignificant impact on the carbon emissions of the structural system. To study the carbon emission law ofsteel frame mesh shells, numerical simulations were conducted to analyze the structural systems ofdifferent roof steel frame mesh shells. The material consumption and carbon emissions of mesh shells andgrids with different spans and materials were calculated, and the carbon emissions of mesh shells andgrids were predicted using BP neural networks. The research results indicate that by calculating thematerial consumption, stress ratio, and stability of the mesh shell and grid structure system through staticanalysis, it is determined that the structural system is in a stable state under full load. According to thecalculation of the carbon emissions of the grid shell and grid structure system,when the structural systemmaterial is aluminum alloy, its carbon emissions are smaller than other materials. Based on the BP neuralnetwork model, the carbon emissions of the mesh shell and grid structure were predicted. The errorbetween the predicted value and the actual value was ≤ 5%, and the accuracy met the predictionrequirements. In the production and installation process of roof materials, advanced constructiontechniques can be used to improve construction efficiency, thereby reducing energy consumption andmaterial waste during the construction process.