Research on Characteristics of Horizontal Atmospheric Diffusion Coefficient Based on BP Neural Network

Abstract:

In order to obtain the actual characteristics of horizontal atmospheric diffusion direction, based on Gauss plume model and the measured data, we use BP neural network to fit the characteristic curve of the diffusion coefficient. We establish a BP neural network, and then we train the network and simulate the diffusion coefficient. According to the simulation results, we compare the characteristic curve with the curve based on the least square method. And the results show that the characteristic curve based on BP neural network has better fitting accuracy. Hence, using the trained neural network to predict the diffusion coefficient has certain theory meaning and actual application value.

• Periodical: Advanced Materials Research (Volumes 791-793)
• Pages: 1605-1608
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.791-793.1605
• Online since: September 2013
• Authors: Pin Shang, Cheng Dong Wu, Ren Ke Han, Wen Jia Ma
• Keywords: Atmospheric Diffusion Coefficient, BP Neural Network (BPNN), Curve Fitting, Gauss Plume Model

References

[1] Akula Venkatram. An examination of the Pasquill-Gifford-Turner dispersion scheme [J]. Atmospheric Environment, 30, 8 (1996): 1283-1290.

DOI: 10.1016/1352-2310(95)00367-3

[2] Turner D B. Workbook of atmospheric dispersion estimates [M]. Boca Reton: CRC Press, (1994).

[3] Noel de Nevers. Air pollution control engineering [M]. New York: McGraw-Hill, (2000).

[4] Xie Shaodong, Zhang Yuanhang, Tang Xiaoyan. Dispersion Models for Vehicle Exhaust Pollutant [J]. Environmental Science, 1999, 1(20): 104-109.

[5] G.A. Davidson. A modified power law representation of the Pasquill-Gifford dispersion coefficients [J]. Journal of the Air & Waste Management Association, 1990, 8, 40(8): 1146-1147.

DOI: 10.1080/10473289.1990.10466761

[6] Irwin, J.S. Estimating plume dispersion: a comparison of several sigma schemes [J]. Journal of Climate and Applied Meteorology, 1983, 22(1): 92-114.

DOI: 10.1175/1520-0450(1983)022<0092:epdaco>2.0.co;2

[7] Yuping Li. A Set of Empirical Formulas for Calculation of Atmospheric Dispersion Coefficients [J]. Transactions of Beijing Institute of Technology, 2009, 10, 29(10): 914-917.

[8] Zhou Hongxiao, Cai Jun, Ren Deguan, et al. An improved algorithm on hidden nodes in multi-layer feed-forward neural networks [J]. Journal of Zhejiang Normal University (Natural Sciences), 2002, 25(3): 268-271.

[9] Sun Fan, Shi Xueqing. Design of BP Networks based on MATLAB [J]. Computer & Digital Engineering, 2007, 35(8): 124-125.

[10] Li Zhongzhi. Simulated formula between water level and water flux based on improved BP neural network [J]. China Rural Water and Hydropower, 2008(10): 30-32.