The Application of Support Vector Regression Models which Based on MATLAB on the Simulation of Wastewater Treatment Plant

Article Preview

Abstract:

Mathematical modeling and simulation technology is a very important wastewater treatment plant data processing analysis tools. To analysis the wastewater treatment plant operation process, this study was realized on MATLAB and LIB-SVM tools. Determine the five feed water quality indexes (water inflow, pH, temperature, COD, MLSS) as input variables and the effluent COD concentration as output variable. Within the SVM modeling, through the GA algorithm, PSO algorithm and the grid search method to separately carried on the model parameters optimization. Through the verification results show that SVM model predicted the COD concentration average relative error is 0.0165, which has higher accuracy, which can meet the actual requirements of water quality prediction.

You might also be interested in these eBooks

Info:

Periodical:

Pages:

197-201

Citation:

Online since:

February 2013

Export:

Price:

Permissions CCC:

Permissions PLS:

Сopyright:

© 2013 Trans Tech Publications Ltd. All Rights Reserved

Share:

Citation:

[1] Henze M., Gujer W., Mino T.: Activated Sludge models ASMI ASM2, ASM2d and ASM3(London: IWAPublishing, 1987).

DOI: 10.2166/9781780402369

Google Scholar

[2] Gernaey K. V., Loosdreeht M. C. M., Henze M., Environmental Modelling & Software, Vol(19) 2004, pp.763-783.

Google Scholar

[3] F. Lim, Min W., Zhe Y. China Resources Comprehensive Utilization, Vol(25) 2007, pp.21-23.

Google Scholar

[4] Vapnik V.N.: The Nature of Statistical Learning Theory (NewYork: Springer-Verlag, 1995).

Google Scholar

[5] Ivanciuc O., in: Applications of Support vector Machines in Chemistry. (Weinheim, 2007) pp.291-400.

Google Scholar

[6] Sathiya S., Keerthi C. J. Neural Computation, Vol. (15)2003, pp.1667-1689.

Google Scholar

[7] JingXu L., HuaiPing C., YueJin-T., Journal of system simulation. Vol. (19) 2007, pp.1540-1547.

Google Scholar

[8] LIBSVM: a library for support Vector Machines [OL]. Chih-Chung http: /www. csie. ntu. edu. tw.

Google Scholar