Keywords

Fuzzy, neural network, environmental decision support system

Location

Session C1: VI Data Mining for Environmental Sciences Session

Start Date

13-7-2016 2:50 PM

End Date

13-7-2016 3:10 PM

Description

The development of an environmental decision support system (EDSS) by means of two different aims to support the operators’ decisions in the drinking water treatment plant (DWTP), equipped with the biggest electrodialysis reversal (EDR) in the world has been tested. A fuzzy artificial neural network model (fuzzy) and an artificial neural network (ANN) have been compared for optimizing the decision: how to manage water blending ratios from EDR and conventional treatment of the drinking water plant, evaluating Llobregat River characteristics (inlet DWTP), current operating conditions of DWTP, weather conditions and distribution requirements on-line. This tool has been tested among 4,5 months in the facility, showing better tendencies regarding the fuzzy model in comparison to the ANN model. A further application of the fuzzy and ANN process models, tested and validated, will be integrated into a process-controlled architecture of the EDSS.

 
Jul 13th, 2:50 PM Jul 13th, 3:10 PM

Fuzzy vs neural network models for environmental decision support system implementation aiming to standardise the multiparametric decision in a Drinking Water Plant with Electrodialysis Reversal

Session C1: VI Data Mining for Environmental Sciences Session

The development of an environmental decision support system (EDSS) by means of two different aims to support the operators’ decisions in the drinking water treatment plant (DWTP), equipped with the biggest electrodialysis reversal (EDR) in the world has been tested. A fuzzy artificial neural network model (fuzzy) and an artificial neural network (ANN) have been compared for optimizing the decision: how to manage water blending ratios from EDR and conventional treatment of the drinking water plant, evaluating Llobregat River characteristics (inlet DWTP), current operating conditions of DWTP, weather conditions and distribution requirements on-line. This tool has been tested among 4,5 months in the facility, showing better tendencies regarding the fuzzy model in comparison to the ANN model. A further application of the fuzzy and ANN process models, tested and validated, will be integrated into a process-controlled architecture of the EDSS.