Modelling sump pump energy usage patterns using artificial intelligence
Abstract
Water typically seeps into the basements of buildings that are close to lakes or situated in low-lying areas, and hence these places require continuous pumping to avoid flooding. Such situations very often lead to high energy consumption which is not well-understood. This research models and analyzes the energy usage patterns of sump pumps installed at Basement 3 of Asia Pacific College (APC) through artificial intelligence techniques. High-resolution (5-minute) and daily aggregated datasets were recorded by SCT013 current sensors and then were converted to energy consumption values. Four prediction models—linear regression, random forest, support vector machine (SVM), and long short-term memory (LSTM)— were trained and tested using root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R²) metrics. For the 5-minute dataset, random forest achieved RMSE=20.78, MAE=2.05, and R²=0.226, whereas LSTM obtained RMSE=18.22, MAE=1.56, and negative R². For the daily dataset, random forest achieved RMSE=299.43, MAE=222.98, R²=0.091, while LSTM reached RMSE=289.80, MAE=208.04, and R²=0.014. The models indicate strong diurnal patterns and show the effect of temporal resolution on predictive performance. All reported LSTM results correspond to direct one-step-ahead forecasting and therefore characterize short-horizon predictive accuracy rather than multi-step deployment performance. The developed framework serves as a steppingstone towards future intelligent pump control and energy-efficient operation.
Keywords
Energy forecasting; Long short-term memory; Machine learning; Random forest; Smart buildings; Sump pumps
Full Text:
PDFDOI: https://doi.org/10.11591/eei.v15i6.12025
Refbacks
- There are currently no refbacks.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Bulletin of Electrical Engineering and Informatics (BEEI)
ISSN: 2089-3191
,
e-ISSN: 2302-9285
This journal is published by the
Institute of Advanced Engineering and Science (IAES)
.