| id |
ecaade2023_156 |
| authors |
Luo, Ruizhe, Zhang, Xingzhao, Yang, Luqiao, Yang, Ruyi, Zhang, Fazhuo, Zhang, Ding, Huang, Chenyu and Yao, Jiawei |
| year |
2023 |
| title |
Predicting the Environmental Effects of Urban Morphology and Greenery Using Deep Generative Models |
| source |
Dokonal, W, Hirschberg, U and Wurzer, G (eds.), Digital Design Reconsidered - Proceedings of the 41st Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe 2023) - Volume 2, Graz, 20-22 September 2023, pp. 307–316 |
| doi |
https://doi.org/10.52842/conf.ecaade.2023.2.307
|
| summary |
The increasing urbanization process has led to greater thermal stress on pedestrians, and greenery arrangement can provide an effective mitigation strategy. However, evaluating the environmental impact of different urban morphologies and greenery arrangements using traditional methods requires time-consuming simulations. To address this challenge, we utilized a deep generative model to predict outdoor environmental indicators influenced by greenery and urban morphology. By creating a dataset of Universal Thermal Climate Index, wind speed, temperature, mean radiant temperature, and relative humidity from Envi-met simulations of three building morphologies with randomly distributed greenery arrangements, we found that Building Shadow Exposure (BSE) and Frontal Area Index, as well as BSE and Porosity, had strong interactions. Our study demonstrates that a pix2pix model trained on this dataset can accurately predict the outdoor environment in seconds (R2 > 0.80), making it a promising tool for sustainable urban planning. Thus, our research suggests that deep generative models can accelerate simulation processes and enable more comprehensive studies to support sustainable urban planning in the future. |
| keywords |
Outdoor Environment, Urban Greenery, Urban Morphology, Deep Generative Model |
| series |
eCAADe |
| email |
|
| full text |
 file.pdf (3,058,529 bytes) |
| references |
Content-type: text/plain
|
| last changed |
2023/12/10 10:49 |
|
