智能城市2024，Vol.10Issue(1)：53-55,3.DOI:10.19301/j.cnki.zncs.2024.01.016
基于机器学习量化对城市热岛的形态学影响
Quantitative morphological impact of urban heat island based on machine learning
摘要
Abstract
Taking the Pearl River New Town in Guangzhou as an example,this paper uses XGBoost,Shapley Value(SHAP)and Partial Dependency Map(PDP)to evaluate the impact of urban morphology on urban heat island.The results emphasize the differential impact of urban form attributes on heat islands,and optimizing building configurations can reduce heat island risks.Urban layout factors have a threshold impact on heat islands,and it is necessary to control the surface fraction and height of buildings within a limited time to reduce negative impacts.This research model focuses on adapting to spatial planning,resisting extreme urban heat island storms,and providing recommendations for the relationship between urban form and heat island risk through interpretable machine learning.关键词
机器学习/深度学习/城市热岛/城市建筑形态Key words
machine learning/deep learning/urban heat island/urban architectural form分类
资源环境引用本文复制引用
陈紫荆..基于机器学习量化对城市热岛的形态学影响[J].智能城市,2024,10(1):53-55,3.
