Abstract：[Objective] In recent years, under the influence of strong wind, trees and walls collapse, objects fall, and other situations occur from time to time, which seriously affect the safety of community residents. In traditional emergency rescue, the background wind field at the disaster site is unknown, and the accuracy of accident development assessment is affected. In the case of fire, gas leakage, strong wind, and other disasters, decision-makers and rescue teams cannot accurately locate the dangerous areas in the community because of their inability to rapidly obtain accurate background wind field information, which affects the accuracy of the judgment of the disaster scope and development trend. Key dangerous areas in the community under strong wind need to be identified.[Methods] In this study, the wind field of a community in the Shijingshan District of Beijing was taken as an example to conduct scene modeling. To generate the database of the community wind field, the wind field was generated by OpenFOAM, and a shell script was used for a batch of simulations. The speed at the feature points obtained by k-means clustering served as the input, and the wind field served as the output to train the neural network. The selected community feature points could represent the wind field information of the community. The feature point selection and neural network modeling were continuously optimized based on the training and prediction results until the accuracy met the requirements.[Results] Taking the field data of 6 681 points predicted by 10 feature points as an example, the model training test results of 7917 training wind fields and 2026 testing wind fields were as follows: The average relative errors of the predicted values of speeds above 1m/s in the x- and y-axes were 5.8% and 6.2%, respectively. Among them, the average relative error of model prediction between 1m/s and 2m/s is 11.9%, for model prediction between 2m/s and 5m/s was 6.0%, for model prediction between 5m/s and 10m/s was 3.2%, and for model prediction above 10m/s was 3.5%.[Conclusions] Compared with the numerical simulation technology, the neural network model can rapidly generate the background wind field of the community based on the field location data. Compared with the time of the numerical simulation, the time of the neural network model to generate a field is significantly reduced. Unlike the existing neural network model, the proposed model takes actual community points as the feature points for model training and prediction, enabling the installation of sensors and the prediction of real-time wind fields. Therefore, people can organize risk prevention and emergency rescue according to the background wind field, which is of great significance for maintaining community safety.
李聪健, 高航, 刘奕. 基于数值模拟和机器学习的风场快速重构方法[J]. 清华大学学报（自然科学版）, 2023, 63(6): 882-887.
LI Congjian, GAO Hang, LIU Yi. Fast reconstruction of a wind field based on numerical simulation and machine learning. Journal of Tsinghua University(Science and Technology), 2023, 63(6): 882-887.
[1] PATTERSON E B. Poverty, income inequality, and community crime rates[J]. Criminology, 1991, 29(4):755-776. [2] 贾楠,陈永强,郭旦怀,等.社区风险防范的三角形模型构建及应用[J].系统工程理论与实践, 2019, 39(11):2855-2864. JIA N, CHEN Y Q, GUO D H, et al. Construction and application of triangle model for community risk prevention[J]. Systems Engineering:Theory&Practice, 2019, 39(11):2855-2864.(in Chinese) [3] YI F, YU Z W, CHEN H H, et al. Cyber-physical-social collaborative sensing:From single space to cross-space[J]. Frontiers of Computer Science, 2018, 12(4):609-622. [4] KAPUCU N. Disaster and emergency management systems in urban areas[J]. Cities, 2012, 29(S1):S41-S49. [5] 刘奕,张宇栋,张辉,等.面向2035年的灾害事故智慧应急科技发展战略研究[J].中国工程科学, 2021, 23(4):117-125. LIU Y, ZHANG Y D, ZHANG H, et al. Development strategy of smart emergency-response technology for disasters and accidents by 2035[J]. Strategic Study of CAE, 2021, 23(4):117-125.(in Chinese) [6] GHOREYSHI M, JIRASEK A, MILLER T, et al. Implementation and verification of gust modeling in an open-source flow solver[J]. Aerospace Science and Technology, 2019, 92:777-789. [7] LING J L, KURZAWSKI A, TEMPLETON J. Reynolds averaged turbulence modelling using deep neural networks with embedded invariance[J]. Journal of Fluid Mechanics, 2016, 807:155-166. [8] LIU Y, WANG Y Q, DENG L, et al. A novel in situ compression method for CFD data based on generative adversarial network[J]. Journal of Visualization, 2019, 22(1):95-108. [9] LI K, KOU J Q, ZHANG. Deep neural network for unsteady aerodynamic and aeroelastic modeling across multiple Mach numbers[J]. Nonlinear Dynamics, 2019, 96(3):2157-2177. [10] YU J, HESTHAVEN J S. Flowfield reconstruction method using artificial neural network[J]. AIAA Journal, 2019, 57(2):482-498. [11] BHATNAGAR S, AFSHAR Y, PAN S W, et al. Prediction of aerodynamic flow fields using convolutional neural networks[J]. Computational Mechanics, 2019, 64(2):525-545. [12] SEKAR V, JIANG Q H, SHU C, et al. Fast flow field prediction over airfoils using deep learning approach[J]. Physics of Fluids, 2019, 31(5):057103. [13] WU H Z, LIU X J, AN W, et al. A deep learning approach for efficiently and accurately evaluating the flow field of supercritical airfoils[J]. Computers&Fluids, 2020, 198:104393. [14] MEMARZADEH G, KEYNIA F. A new short-term wind speed forecasting method based on fine-tuned LSTM neural network and optimal input sets[J]. Energy Conversion and Management, 2020, 213:112824. [15] CHEN Y, KOPP G A, SURRY D. Prediction of pressure coefficients on roofs of low buildings using artificial neural networks[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2003, 91(3):423-441. [16] BRE F, GIMENEZ J M, FACHINOTTI V D. Prediction of wind pressure coefficients on building surfaces using artificial neural networks[J]. Energy and Buildings, 2018, 158:1429-1441. [17] YAN Y T, DONG X Q, LI J M. Experimental study of methane diffusion in soil for an underground gas pipe leak[J]. Journal of Natural Gas Science and Engineering, 2015, 27:82-89. [18] LI N, LIU Y Q, LI L, et al. Numerical simulation of wind turbine wake based on extended k-epsilon turbulence model coupling with actuator disc considering nacelle and tower[J]. IET Renewable Power Generation, 2020, 14(18):3834-3842. [19] GHOBADI P, NASROLLAHI N. Assessment of pollutant dispersion in deep street canyons under different source positions:Numerical simulation[J]. Urban Climate, 2021, 40:101027.
2023, Vol. 63 
