Modeling and analysis of pumps in a wastewater treatment plant: A data-mining approach

摘要

A data-mining approach is proposed to model a pumping system in a wastewater treatment plant. Two parameters, energy consumption and wastewater flow rate after the pumping system, are used to evaluate the performance of 27 scenarios while the pump was operating. Five data-mining algorithms are applied to identify the relationships between the outputs (energy consumption and wastewater flow rate) and the inputs (elevation level of the wet well and the speed of the pumps). The accuracy of the flow rate and energy consumption models exceeded 90%. The derived models are deployed to optimize the pump system. The computational results obtained with the proposed models are discussed.