深層学習を用いた単純形状建物における風圧係数予測手法に関する研究

書誌事項

タイトル別名
  • シンソウ ガクシュウ オ モチイタ タンジュン ケイジョウ タテモノ ニオケル フウアツ ケイスウ ヨソク シュホウ ニカンスル ケンキュウ
  • A Method for Predicting Wind Pressure Coefficient on Simple Shaped Building Walls Using Deep Learning

この論文をさがす

抄録

Synopsis : In recent years, due to the progress of global warming, the efficient operation of buildings is becoming more and more important. Natural ventilation has been attracting attention in the air environment field as a method to reduce energy consumption in the operation phase. In particular, it is important to design a wind-forced Ventilation that takes into account the wind pressure on the building wall that is predicted in advance. Wind tunnel experiments are an effective method for predicting wind pressure on building walls, but they require a great deal of time and effort, including preparation. In this study, deep learning was used to predict the distribution of wind pressure coefficients on a simple building wall, and the results were compared with the results of previous wind tunnel experiments to verify the usefulness of deep learning techniques in the design of wind ventilation. As a result, we were able to predict the wind pressure coefficient with good accuracy in the case studied in this research. However, its applicability in current practice is very limited. In order to construct a more general prediction method, it is necessary to repeat the study of various factors in the future.

収録刊行物

詳細情報 詳細情報について

問題の指摘

ページトップへ