Modeling of activated sludge process using various nonlinear techniques: a comparison study
This paper presents a comparison study between radial basis function neural network (RBFNN), feed forward multilayer perceptron neural network (MLPNN) and adaptive neuro-fuzzy (ANFIS) technique to model the activated sludge process (ASP). All of these techniques are based on the nonlinear autoregres...
Saved in:
Main Authors: | , , |
---|---|
Format: | Conference or Workshop Item |
Language: | English |
Published: |
2015
|
Subjects: | |
Online Access: | http://eprints.utm.my/id/eprint/61997/1/NorhalizaAbdWahab2015_ModelingofActivatedSludgeProcessusingVariousNonlinear.pdf http://eprints.utm.my/id/eprint/61997/ http://www.utm.my/iicist/ |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | This paper presents a comparison study between radial basis function neural network (RBFNN), feed forward multilayer perceptron neural network (MLPNN) and adaptive neuro-fuzzy (ANFIS) technique to model the activated sludge process (ASP). All of these techniques are based on the nonlinear autoregressive with eXogenous input (NARX) structure. The ASP inputs and outputs data are generated from activated sludge model 1 (ASM1). This work will cover the dissolved oxygen (DO), substrate and biomass modeling. The performances of the model are evaluated based on R2, mean square error (MSE) and root mean square error RMSE. The simulation result shows that ANFIS with NARX structure given a better performance compared with the other modeling techniques. |
---|