An improved tool of water data analytics for flowmeters data

This paper presents an improved tool for data validation and reconstruction of flowmeters. These sensors are installed in the Catalonia regional water network from Barcelona (Spain). Here a new time series model with exogenous variable is proposed with excellent results for data validation. It is postulated that the integration of the electronics alarms, along with other tests about the daily data accumulated and a later analysis of the data reconstruction allow to improve the results of the existing tools. This is accomplished by decreasing the false alarms and missing alarms of more than 6000 hourly data retrieved from more than 200 flowmeters each day. This new tool provides reliable information daily reliable information of the state of the water network. This information could potentially contribute to optimally control and manage this large and complex water network.
Postprint (published version)