Modeling of membrane fouling in a submerged membrane reactor using support vector regression

AbstractRemoval rate of Fe2+ and Mn2+ using submerged membrane reactor for drinking water in the presence of fulvic acid and iron hydroxide is studied using the data from the experiments obtained from various concentrations of Fe2+, Mn2+, fulvic acid, and iron hydroxide. The relationship between these contaminants and membrane fouling is investigated. In the experiments, flux is kept as constant, and the pressure change with time is observed. To model the relationship, a regression analysis using the support vector regression (SVR) model is presented. Hyperparameter optimization for SVR is important, that is, wrong selection may cause underfitting/overfitting phenomena. In order to find optimal values, grid search method is performed with various parameters such as different kernel functions (radial basis functions, polynomial, linear), cost parameter (C), and scale parameters γ and e. The results obtained by SVR show that proposed method is feasible.

[1]  Zhan Wang,et al.  Use of support vector machine model to predict membrane permeate flux , 2015 .

[2]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[3]  Manoj Khandelwal,et al.  Prediction of mine water quality by physical parameters , 2005 .

[4]  Mohamed Bouamar,et al.  Multisensor system using support vector machines for water quality classification , 2007, 2007 9th International Symposium on Signal Processing and Its Applications.

[5]  R. Gholami,et al.  Heavy metal pollution assessment using support vector machine in the Shur River, Sarcheshmeh copper mine, Iran , 2012, Environmental Earth Sciences.

[6]  Yoshua Bengio,et al.  Random Search for Hyper-Parameter Optimization , 2012, J. Mach. Learn. Res..

[7]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[8]  Max Kuhn,et al.  Applied Predictive Modeling , 2013 .

[9]  Christophe Ley,et al.  Detecting outliers: Do not use standard deviation around the mean, use absolute deviation around the median , 2013 .

[10]  Davut Hanbay,et al.  Prediction of wastewater treatment plant performance based on wavelet packet decomposition and neural networks , 2008, Expert Syst. Appl..

[11]  Vladimir Cherkassky,et al.  Model complexity control for regression using VC generalization bounds , 1999, IEEE Trans. Neural Networks.

[12]  Jingwen Tian,et al.  The Study of Membrane Fouling Modeling Method Based on Support Vector Machine for Sewage Treatment Membrane Bioreactor , 2007, 2007 2nd IEEE Conference on Industrial Electronics and Applications.