A Case-Based Reasoning Solution for Urban Drinking-Water Quality Control

The Urban Drinking Water Supply (UDWS) systems play a vital role nowadays to support global city management, especially in the context of the fast growth of urbanization. In Europe, there is a natural conflict between urban population growth and the aging drinking water supply infrastructure. In particular, according to European standards, water should be drinkable from the taps. The requirements from these standards not only bring easement in our modern daily life but also are examining the robustness of the risk warning and control response in the UDWS system. Water quality as the key factor in the urban water supply is by default strictly regulated in the municipality management. Currently, the most common water quality control is relying on periodic tests for various quality indicators. In addition, further processing in the treatment plant, including measures and doses, is relatively fixed and weak to respond to the risk in UDWS. In order to address these problems, we propose a Case-Based Reasoning (CBR) approach for water quality risk control in UDWS systems. First, we build a case-based analysis computational scheme for the risk control in the UDWS system. Then we provide our methods in feature selection and similarity evaluation between different cases. Finally, the results of an industrial case study are presented. This case is based on our data collection from the UDWS system in Oslo city in Norway. The discussion and analysis show the feasibility of this approach and practically provide decision support for water treatment actions.