A Machine Learning-Based Ridge-Meta Ensemble Model for Electrical Load Forecasting in Integrated Energy Systems
Abstract
Reliable electrical load forecasting is the backbone for precise operation and intelligent scheduling of Integrated Energy Systems (IES). This study proposes a ridge-meta ensemble approach using ridge regression to merge three gradient boosting techniques: XGBoost, LightGBM, and CatBoost, and predict electrical energy consumption. By fusing the three gradient boosting models with a ridge regression meta-learner, the framework balances high accuracy and computational efficiency. Validation uses 2021-2023 hourly electrical load data from Arizona State University's Tempe Campus, a real-world dataset with complex meteorological and temporal correlations. Test results show the ridge-meta ensemble model achieves an R-squared value of 0.8973, with a mean absolute error (MAE) of 1079.31 kilowatts and a root mean square error (RMSE) of 1490.39 kilowatts. Compared to traditional linear regression, this represents a 46.15% reduction in MAE and a 42.93% reduction in RMSE. Through SHAP (SHapley Additive exPlanations) analysis, we found that wet-bulb temperature and seasonal fluctuations are the primary factors influencing load predictions. With low computational overhead, the proposed model offers a feasible solution for real-time IES scheduling and management.
Read the paper
Related studies
- Chaos and quantum thermalization
- The Dalton quantum chemistry program system
- Cord Blood Leptin and Adiponectin as Predictors of Adiposity in Children at 3 Years of Age: A Prospective Cohort Study
- Quantitative metabolomics by H-NMR and LC-MS/MS confirms altered metabolic pathways in diabetes
- Stable adiabatic quantum batteries
- A novel self-lipid antigen targets human T cells against CD1c(+) leukemias
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.