A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Monitoring of Anoxic/Oxic Process for Nitrogen and Chemical Oxygen Demand Removal Using Fuzzy Neural Networks
In this paper, a software sensor based on a fuzzy neural network (FNN) approach was proposed for the real‐time estimation of nutrient concentrations and overcoming the problem of delayed measurements. To improve the FNN performance, fuzzy subtractive clustering was used to identify the model's architecture and optimize the fuzzy rule; meanwhile, a split network structure, applied separately for anaerobic and aerobic conditions, was used with dynamic modeling methods, such as an auto‐regressive model with exogenous inputs. The proposed methodology was applied to a bench‐scale anoxic/oxic process for biological nitrogen removal. It was possible to partially overcome the extrapolation problem of FNNs with the aid of multi‐way principal component analysis, because it has the ability to detect abnormal situations, which could generate extrapolation. Real‐time estimation of chemical oxygen demand, nitrate, and ammonium concentrations based on the model was successfully carried out with the simple online information of the anoxic/oxic system.
Monitoring of Anoxic/Oxic Process for Nitrogen and Chemical Oxygen Demand Removal Using Fuzzy Neural Networks
In this paper, a software sensor based on a fuzzy neural network (FNN) approach was proposed for the real‐time estimation of nutrient concentrations and overcoming the problem of delayed measurements. To improve the FNN performance, fuzzy subtractive clustering was used to identify the model's architecture and optimize the fuzzy rule; meanwhile, a split network structure, applied separately for anaerobic and aerobic conditions, was used with dynamic modeling methods, such as an auto‐regressive model with exogenous inputs. The proposed methodology was applied to a bench‐scale anoxic/oxic process for biological nitrogen removal. It was possible to partially overcome the extrapolation problem of FNNs with the aid of multi‐way principal component analysis, because it has the ability to detect abnormal situations, which could generate extrapolation. Real‐time estimation of chemical oxygen demand, nitrate, and ammonium concentrations based on the model was successfully carried out with the simple online information of the anoxic/oxic system.
Monitoring of Anoxic/Oxic Process for Nitrogen and Chemical Oxygen Demand Removal Using Fuzzy Neural Networks
Huang, Mingzhi (author) / Wan, Jinquan (author) / Ma, Yongwen (author)
Water Environment Research ; 81 ; 654-663
2009-07-01
10 pages
Article (Journal)
Electronic Resource
English
| Taylor & Francis Verlag | 2011
Study of control strategy and simulation in anoxic-oxic nitrogen removal process
| Online Contents | 2005