深層学習による建物壁面の風圧係数分布予測に関する基礎的検討　未学習の建物形状における予測精度の検討

抄録

In recent years, the use of natural ventilation is increasing due to the promotion of energy saving. However, in the design stage, performance simulation is often not performed because of the labor and cost involved in simulation. Therefore, in this report, for the purpose of constructing a simple method for predicting the wind pressure coefficient, the wind pressure coefficient distribution on the building wall is predicted by deep learning and the prediction accuracy is examined.