Keywords

Intelligent Environmental Decision Support Systems, Drinking water treatment, Data processing

Start Date

27-6-2018 2:00 PM

End Date

27-6-2018 3:20 PM

Abstract

Drinking Water Treatment Plants (DWTP) face complex decision-making in their daily operation. The amount of data generated along the DWTP allows developing data-driven models and knowledge-based models that can be integrated into Intelligent Environmental Decision Support Systems (IEDSS). These systems can be used for predicting the main operational parameters along the water treatment process. At the present study, the procedure for building an IEDSS has been followed with a whole-plant approach. A methodology has been purposed to choose the right modelling technique that attends to the process characteristics and whose results are in tune with the expert knowledge. As a case study, two operations of a DWTP were discussed and data-driven and knowledge-based models were assessed to be integrated into an IEDSS. The chemical dosage rate of pre-oxidation process was analysed and modelled with artificial neural networks and complemented by an expert-based model. On the other hand, the hydraulic management of the advanced treatment by electro-dialysis reversal treatment was modelled with a fuzzy expert system, in which the proposed decisions were the most robust according to the expert knowledge. The outputs of the models adjusted the seasonal and daily changes of the raw water influent under normal circumstances. The IEDSS is being implemented at a full-scale plant that processes real-time data and serve as a useful tool for the plant managers.

Stream and Session

Stream B: (Big) Data Solutions for Planning, Management, and Operation and Environmental Systems

Session B3: Sixth Session on Data Mining as a Tool for Environmental Scientists (S-DMTES-2018)

Download
COinS
 
Jun 27th, 2:00 PM Jun 27th, 3:20 PM

Integrating data-based and knowledge-based models in an Environmental Decision Support System for the management of a Drinking Water Treatment Plant

Drinking Water Treatment Plants (DWTP) face complex decision-making in their daily operation. The amount of data generated along the DWTP allows developing data-driven models and knowledge-based models that can be integrated into Intelligent Environmental Decision Support Systems (IEDSS). These systems can be used for predicting the main operational parameters along the water treatment process. At the present study, the procedure for building an IEDSS has been followed with a whole-plant approach. A methodology has been purposed to choose the right modelling technique that attends to the process characteristics and whose results are in tune with the expert knowledge. As a case study, two operations of a DWTP were discussed and data-driven and knowledge-based models were assessed to be integrated into an IEDSS. The chemical dosage rate of pre-oxidation process was analysed and modelled with artificial neural networks and complemented by an expert-based model. On the other hand, the hydraulic management of the advanced treatment by electro-dialysis reversal treatment was modelled with a fuzzy expert system, in which the proposed decisions were the most robust according to the expert knowledge. The outputs of the models adjusted the seasonal and daily changes of the raw water influent under normal circumstances. The IEDSS is being implemented at a full-scale plant that processes real-time data and serve as a useful tool for the plant managers.