Selection of Optimum Features using PSO- SPG2 for Predicting COD Effluent Level in Wastewater

Wastewaters are waterborne solids and liquids which get discharged into sewers representing the wastes of public life. Wastewater Treatment is the process of removing contaminants from both runoff and domestic wastewater. Chemical Oxygen Demand (COD) is used to measure the degradable organic matter confined in wastewater effluents. An anaerobic process becomes popular method for the treatment of a number of waste streams. Up-flow Anaerobic Filter (UAF) is an anaerobic reactor which is used for the removal and digestion of organic matter present in wastewater. A systematic prediction of agro-food wastewater using Adaptive Neuro-Fuzzy Inference System (ANFIS) was proposed to predict COD effluent of a full-scale anaerobic wastewater treatment plant in accordance with UAF. Wastewater Treatment process initiates with Data collection and pre-processing preceded by normalization technique using Z-Score, which is followed by Feature selection algorithm and the prediction of COD is concluded using ANFIS by the model when applied to the description of UAF treating agro-food industrial wastewaters such as fruit juice wastewater. Prediction of agro-food wastewater can be improved using various feature selection algorithms such as Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and Particle Swarm Optimization (PSO) With Spectral Projected Gradient (SPG2) algorithm. The experimental results proved that Combination of ANFIS with PSO-SPG2 is considered as a best fit to predict the COD effluent level with greater accuracy and with low Mean Square Error values.

[1]  Pradeep Kumar,et al.  High rate anaerobic filter with floating supports for the treatment of effluents from small-scale agro-food industries , 2009 .

[2]  Jyh-Shing Roger Jang,et al.  ANFIS: adaptive-network-based fuzzy inference system , 1993, IEEE Trans. Syst. Man Cybern..

[3]  Yuhui Shi,et al.  Particle swarm optimization: developments, applications and resources , 2001, Proceedings of the 2001 Congress on Evolutionary Computation (IEEE Cat. No.01TH8546).

[4]  José Mario Martínez,et al.  Nonmonotone Spectral Projected Gradient Methods on Convex Sets , 1999, SIAM J. Optim..

[5]  G. Centi,et al.  Copper-pillared clays (Cu-PILC) for agro-food wastewater purification with H2O2 , 2008 .

[6]  Sri Ramakrishna,et al.  FEATURE SELECTION METHODS AND ALGORITHMS , 2011 .

[7]  Marcos Raydan,et al.  The Barzilai and Borwein Gradient Method for the Large Scale Unconstrained Minimization Problem , 1997, SIAM J. Optim..

[8]  Chitralekha Mahanta,et al.  A novel approach for ANFIS modelling based on full factorial design , 2008, Appl. Soft Comput..

[9]  Zbigniew Michalewicz,et al.  Genetic Algorithms + Data Structures = Evolution Programs , 1996, Springer Berlin Heidelberg.

[10]  L. Ladha,et al.  FEATURE SELECTION METHODS AND ALGORITHMS , 2011 .

[11]  Marc M. Van Hulle,et al.  Genetic Algorithm for Feature Subset Selection with Exploitation of Feature Correlations from Continuous Wavelet Transform: a real-case Application , 2004, International Conference on Computational Intelligence.