Discovering weekly seasonality for water demand prediction using evolutionary algorithms
Piotr Lipiński , Patryk Filipiak , Paweł Rychlikowski , Justyna Stańczyk , Joanna Kajewska-Szkudlarek , Janusz Łomotowski , Tomasz Konieczny
AbstractThe modern approach to water supply network management and operation is related to the use of modern solutions from both technical and strategic perspectives. Apart from practices promoted by International Water Association (IWA) (active leakage control, pressure management, speed and quality of repair, pipeline and as- sets management), water demand prediction systems are the future. Preparing short-, medium- and long-term water consumption forecasts is the key factor these days. Short-termsimulations, mainly those covering the period of 24 or 48 hours, are used to optimise the operation of pumping stations and to resolve current exploitation issues, whereas long-term analyses, covering more than one month or year, are said to supportthe decision-making process regarding the design and development of water supply networks. Medium-term predictions, covering weeks, are used to create time schedules for the maintenance of water supply networks and develop failure prevention procedures. Due to the lack of assessment teams and proper assessment tools, the majority of water and sewerage companies store registered time series without thorough study of the data. Information regarding the current operating status of the water supplynetwork included in the series is not properly used and irretrievably lost.
|Publication size in sheets||0.3|
|Book||GECCO '17 Proceedings of the Genetic and Evolutionary Computation Conference Companion ; Berlin, Germany, July 15 - 19, 2017, 2017, Association for Computing Machinery, ISBN 978-1-4503-4939-0|
|Citation count*||1 (2020-09-28)|
* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.