The Prediction of SO₂ Pollutant Concentration Using a RBF Neural Network

Abstract:

Air pollution may cause pernicious effects on human health, and is a widespread problem in the world. Air quality management systems have became an important research issue with strong implications for inhabitants’ health. Monitoring and forecasting of air quality indicators plays an important role in the management systems. Artificial intelligent techniques are successfully used in modelling of highly complex and nonlinear phenomena. In this paper, a model, which is radial basis function (RBF) neural network, is established to estimate the impact of meteorological indicators on SO₂. The proposed model achieves 9.91% in mean absolute percentage error (MAPE) compared to real observation data sequence. For air quality, it could be a promising candidate for forecasting the air quality indicators data sequence.

• Periodical: Applied Mechanics and Materials (Volumes 55-57)
• Pages: 1392-1396
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.55-57.1392
• Online since: May 2011
• Authors: Cheng Yuan Wang, Wen Yu Zhang, Ju Jie Wang, Wen Fang Zhao
• Keywords: Air Quality, Forecasting, Measurement, Modeling, RBF Neural Network

References

[1] Akkoyunlu A, Erturk F. Evaluation of air pollution trends in istanbul. Int J Environ Pollut 2003 (18): 98-388.

[2] Elbir T, Muezzinoglu A, Bayram A. Evaluation of some air pollution indicators in Turkey. Environ Int 2000 (26): 5-10.

[3] Tayanc M. An assessment of spatial and temporal variation of sulfur dioxide levels over Istanbul, Turkey. Environ Pollut 2000 (107): 9-61.

DOI: 10.1016/s0269-7491(99)00131-1

[4] McCollister GM, Wilson K R. Linear stochastic models for forecasting daily maxima and hourly concentrations of air pollutants. Atmospheric Environment 1974 (9): 23-416.

DOI: 10.1016/0004-6981(75)90127-4

[5] Yilmaz Yildirim , Mahmut Bayramoglu. Adaptive neuro-fuzzy based modelling for prediction of air pollution daily levels in city of Zonguldak. Chemosphere 2006 (63): 1575-1582.

DOI: 10.1016/j.chemosphere.2005.08.070

[6] Wei-Zhen Lu, etc., Potential assessment of a neural network model with PCA/RBF approach for forecasting pollutant trends in Mong Kok urban air, Hong Kong. Environmental Research 2004(96): 79-87.

DOI: 10.1016/j.envres.2003.11.003

[7] D.S. Broomhead, D. Lowe. Multivariable functional interpolation and adaptotive networks. complex Systems 1988 (11): 321-355.

[8] C. M. Bishop, Improving the generalization properties of radial basis function neural networks. Neural Comput 1991(3): 579-588.

DOI: 10.1162/neco.1991.3.4.579

[9] D. K. WeddingII, K. J. Cios, Time series forecasting by combining RBF networks certainty factors and the Box-Jenkins model. Neuro computing 1996 (10): 149-168.

DOI: 10.1016/0925-2312(95)00021-6

[10] L. Yu, W. Huang, K. K. Lai, S. Y. Wang, A reliability-based RBF network ensemble model for foreign exchange rates prediction, in: I. King, etal. (Eds. ). ICONIP 2006, Part III, Lecture Notes in Computer Science, Vol. 4234, 2006, pp.380-389.

DOI: 10.1007/11893295_43

[11] Lean Yu , Kin Keung Lai , Shouyang Wang. Multistage RBF neural network ensemble learning for exchange rates forecasting. Neuro computing 2008 (71): 3295-3302.

DOI: 10.1016/j.neucom.2008.04.029

[12] Chien-Cheng Lee, Yu-Chun Chiang, Cheng-Yuan Shih, Chun-Li Tsai. Noisy time series prediction using M-estimator based robust radial basis function neural networks with growing and pruning techniques. Expert Systems with Applications 2009 (36): 4717–4724.

DOI: 10.1016/j.eswa.2008.06.017