Water supply system real time operational management Demo Case #1 Athens Greece The platform was integrated with several new applications, aiming to provide decision support on optimal sluice gates (flow regulation structures) settings, early warning for high turbidity events, forecasts on turbidity level, and estimations of water supply volumes. The optimal operation and scheduling of the sluice gates was supported by an application following a “grey-box” approach that estimates the sluice gates openings for a target flow. Furthermore, two deep neural network models were built to forecast the turbidity level at the two most downstream sensors of the system under study, using as predictors turbidity measurements from the most upstream sensors. Finally, an application was developed that provides one-day ahead forecasts of the total daily water supply volumes, with focus on the annual and weekly seasonality of water outflows, during periods of exceptional demand. Fiware4Water delivers a fully functional Fiware-enabled system architecture, which can easily be extended to cover other applications, taking advantage of the connectors developed. Similarly, third-party applications, models, tools and services can take advantage of the installed Fiwareenabled infrastructure to communicate seamlessly in a bidirectional way, following Fiware standardization protocols and not ad-hoc approaches, which lead to fragmented and isolated solutions. Vasiliki Polychniatou EYDAP Demo case 1 leader Panagiotis Kossieris NTUA Christos Makropoulos KWR Back to content Back to the list scientific focus
RkJQdWJsaXNoZXIy NTYxMTg3