A short-term hourly water demand forecasting algorithm is needed in order to ensure a stable and safe supply of water. Unlike daily or monthly water demand forecasting, there are a large amount of fluctuation of hourly water demand. Hourly water demand is affected by short time period and abnormal data caused by the sensor, communication, and water treatment plant problems. An effective refinement method that detects and corrects the abnormal data among the historical data is needed to achieve accurate and practical hourly water demand forecasting. In this paper, we suggest an abnormal data refinement out of a confidence interval (ADR-CI) method and an error percentage correction (EPC) method. These methods try to distribute and revise the incoming hourly water demand and past water demand data. The proposed methods are verified by the experiments in a real water supply plant during a year.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.