Pedestrian wind flow prediction using spatial-frequency generative adversarial network
作者机构：School of Architecture and Urban PlanningBeijing University of Civil Engineering and ArchitectureBeijing100044China Beijing Key Laboratory of Intelligent Processing for Building Big DataBeijing100044China School of Electrical and Information EngineeringBeijing University of Civil Engineering and ArchitectureBeijing100044China School of Computer Science and TechnologyHarbin Institute of TechnologyHarbin150001China
出 版 物：《Building Simulation》 (建筑模拟（英文）)
年 卷 期：2024年第17卷第2期
页 面：319-334页
核心收录：
学科分类：12[管理学] 1204[管理学-公共管理] 081203[工学-计算机应用技术] 08[工学] 081303[工学-城市规划与设计(含：风景园林规划与设计）] 0835[工学-软件工程] 0813[工学-建筑学] 0833[工学-城乡规划学] 083302[工学-城乡规划与设计] 0812[工学-计算机科学与技术（可授工学、理学学位）]
基 金：This work was financially supported by the Beijing Municipal Natural Science Foundation[No.4232021] the National Natural Science Foundation of China[No.62271036,No.62271035,No.62101022] the Pyramid Talent Training Project of Beijing University of Civil Engineering and Architecture[No.JDYC20220818] theYoung teachers research ability enhancement program of Beijing University of Civil Engineering and Architecture[No.X21083]
主 题：pedestrian wind flow prediction generative adversarial network Gaussian kernel wavelet transform objective function
摘 要：Pedestrian wind flow is a critical factor in designing livable residential environments under growing complex urban *** pedestrian wind flow during the early design stages is essential but currently suffers from inefficiencies in numerical *** learning,particularly generative adversarial networks(GAN),has been increasingly adopted as an alternative method to provide efficient prediction of pedestrian wind ***,existing GAN-based wind flow prediction schemes have limitations due to the lack of considering the spatial and frequency characteristics of wind flow *** study proposes a novel approach termed SFGAN,which embeds spatial and frequency characteristics to enhance pedestrian wind flow *** the spatial domain,Gaussian blur is employed to decompose wind flow into components containing wind speed and distinguished flow edges,which are used as the embedded spatial *** information of wind flow is obtained through discrete wavelet transformation and used as the embedded frequency *** spatial and frequency characteristics of wind flow are jointly utilized to enforce consistency between the predicted wind flow and ground truth during the training phase,thereby leading to enhanced *** results demonstrate that SFGAN clearly improves wind flow prediction,reducing Wind_MAE,Wind_RMSE and the Fréchet Inception Distance(FID)score by 5.35%,6.52%and 12.30%,compared to the previous best method,*** also analyze the effectiveness of incorporating the spatial and frequency characteristics of wind flow in predicting pedestrian wind *** reduces errors in predicting wind flow at large error intervals and performs well in wake regions and regions surrounding *** enhanced predictions provide a better understanding of performance variability,bringing insights at the early design stage to improve pedestrian wind *** proposed spatial-frequen
维普期刊数据库