PREDICTING WIND PRESSURE ON LOWRISE BUILDINGS WITH PROTECTIVE PARAPETS USING ARTIFICIAL NEURAL NETWORKS
dc.contributor.advisor | Phillips, Brian M | en_US |
dc.contributor.author | Wang, Yanan | en_US |
dc.contributor.department | Civil Engineering | en_US |
dc.contributor.publisher | Digital Repository at the University of Maryland | en_US |
dc.contributor.publisher | University of Maryland (College Park, Md.) | en_US |
dc.date.accessioned | 2020-07-14T05:35:40Z | |
dc.date.available | 2020-07-14T05:35:40Z | |
dc.date.issued | 2020 | en_US |
dc.description.abstract | Wind hazards cause tremendous destruction and threaten people’s safety and economical losses. To improve current provision toward wind load and strengthen buildings against wind forces, it’s vital to properly characterize wind loading on buildings in wind hazard. The turbulence created by the bluff body makes analytical modeling difficult. Therefore, engineers typically turn to wind tunnel tests. This thesis investigates the application of Artificial Neural Networks (ANN) to predict the wind pressure on low-rise buildings with protective parapets. With existing experimental datasets conducted in BLWT, ANN models were trained to model non-linear relationship between inputs, such as tap coordinates and parapet height, and outputs, such as pressure coefficients. The developed model was used to predict pressure coefficients with unseen parapet height to cut down experimental cost. | en_US |
dc.identifier | https://doi.org/10.13016/wl9k-zksw | |
dc.identifier.uri | http://hdl.handle.net/1903/26318 | |
dc.language.iso | en | en_US |
dc.subject.pqcontrolled | Civil engineering | en_US |
dc.subject.pquncontrolled | artificial neural network | en_US |
dc.subject.pquncontrolled | backpropagation | en_US |
dc.subject.pquncontrolled | low rise building | en_US |
dc.subject.pquncontrolled | pressure coefficients | en_US |
dc.title | PREDICTING WIND PRESSURE ON LOWRISE BUILDINGS WITH PROTECTIVE PARAPETS USING ARTIFICIAL NEURAL NETWORKS | en_US |
dc.type | Thesis | en_US |
Files
Original bundle
1 - 1 of 1