Electricity Consumption Forecasting: An Approach Using Cooperative Ensemble Learning with SHapley Additive exPlanations

2026-05-24

AI Summary

This study addresses the growing challenges in electricity management and environmental sustainability driven by rapidly increasing energy consumption in higher education institutions, necessitating accurate medium- to long-term forecasting approaches. Leveraging seven years of electricity usage and meteorological data from two campuses of the Federal Institute of Paraná in Brazil, we propose a novel method—Weaker Separator Booster—that integrates SHAP-based interpretability analysis with a cooperative ensemble strategy combining LSTM, Random Forest, Support Vector Regression, and XGBoost models, whose hyperparameters are optimized via genetic algorithm and particle swarm optimization. Experimental results demonstrate superior predictive performance, achieving sMAPE scores of 13.90% and 18.72% on the IFPR-Palmas and Coronel Vivida campuses, respectively. The analysis further reveals a consistent pattern wherein lagged electricity consumption variables dominate predictive power, while meteorological factors exert only marginal influence, underscoring both the method’s accuracy and its interpretability advantages.

Abstract

Electricity expense management presents significant challenges, as this resource is susceptible to various influencing factors. In universities, the demand for this resource is rapidly growing with institutional expansion and has a significant environmental impact. In this study, the machine learning models long short-term memory (LSTM), random forest (RF), support vector regression (SVR), and extreme gradient boosting (XGBoost) were trained with historical consumption data from the Federal Institute of Paraná (IFPR) over the last seven years and climatic variables to forecast electricity consumption 12 months ahead. Datasets from two campuses were adopted. To improve model performance, feature selection was performed using Shapley additive explanations (SHAP), and hyperparameter optimization was carried out using genetic algorithm (GA) and particle swarm optimization (PSO). The results indicate that the proposed cooperative ensemble learning approach named Weaker Separator Booster (WSB) exhibited the best performance for datasets. Specifically, it achieved an sMAPE of 13.90% and MAE of 1990.87 kWh for the IFPR-Palmas Campus and an sMAPE of 18.72% and MAE of 465.02 kWh for the Coronel Vivida Campus. The SHAP analysis revealed distinct feature importance patterns across the two IFPR campuses. A commonality that emerged was the strong influence of lagged time-series values and a minimal influence of climatic variables.

Problem

• Electricity Consumption Forecasting
• University Energy Management
• Environmental Impact
• Time-Series Prediction
• Energy Demand

Innovation

• Cooperative Ensemble Learning
• SHAP
• Weaker Separator Booster
• Hyperparameter Optimization
• Electricity Consumption Forecasting