An Ensemble Model of Attention-Enhanced N-BEATS and XGBoost for District Heating Load Forecasting
Accurate heat load forecasting is essential for the efficiency of District Heating Systems (DHS). Still, it is challenged by the need to model long-term temporal dependencies and nonlinear relationships with weather and other factors. This study proposes a hybrid deep learning framework combining an...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-07-01
|
| Series: | Energies |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1996-1073/18/15/3984 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849239847808532480 |
|---|---|
| author | Shaohua Yu Xiaole Yang Hengrui Ye Daogui Tang Hamidreza Arasteh Josep M. Guerrero |
| author_facet | Shaohua Yu Xiaole Yang Hengrui Ye Daogui Tang Hamidreza Arasteh Josep M. Guerrero |
| author_sort | Shaohua Yu |
| collection | DOAJ |
| description | Accurate heat load forecasting is essential for the efficiency of District Heating Systems (DHS). Still, it is challenged by the need to model long-term temporal dependencies and nonlinear relationships with weather and other factors. This study proposes a hybrid deep learning framework combining an attention-enhanced Neural Basis Expansion Analysis for Time Series (N-BEATS) model and eXtreme Gradient Boosting (XGBoost). The N-BEATS component, with a multi-head self-attention mechanism, captures temporal dynamics, while XGBoost models non-linear impacts of external variables. Predictions are integrated using an optimized weighted averaging strategy. Evaluated on a dataset from 103 heating units, the model outperformed 13 baselines, achieving an MSE of 0.4131, MAE of 0.3732, RMSE of 0.6427, and R<sup>2</sup> of 0.9664. This corresponds to a reduction of 32.6% in MSE, 32.0% in MAE, and 17.9% in RMSE, and an improvement of 5.1% in R<sup>2</sup> over the best baseline. Ablation studies and statistical tests confirmed the effectiveness of the attention mechanism and ensemble strategy. This model provides an efficient solution for DHS load forecasting, facilitating optimized energy dispatch and enhancing system performance. |
| format | Article |
| id | doaj-art-e9dcb28151d545e9b273721f1d0cfd8c |
| institution | Kabale University |
| issn | 1996-1073 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Energies |
| spelling | doaj-art-e9dcb28151d545e9b273721f1d0cfd8c2025-08-20T04:00:49ZengMDPI AGEnergies1996-10732025-07-011815398410.3390/en18153984An Ensemble Model of Attention-Enhanced N-BEATS and XGBoost for District Heating Load ForecastingShaohua Yu0Xiaole Yang1Hengrui Ye2Daogui Tang3Hamidreza Arasteh4Josep M. Guerrero5School of Intelligent Manufacturing, Nanjing University of Science and Technology, Nanjing 210094, ChinaSchool of Cyber Science and Engineering, Nanjing University of Science and Technology, Nanjing 210094, ChinaSchool of Cyber Science and Engineering, Nanjing University of Science and Technology, Nanjing 210094, ChinaSchool of Transportation and Logistics Engineering, Wuhan University of Technology, Wuhan 430063, ChinaCenter for Research on Microgrids (CROM), Huanjiang Laboratory, Shaoxing 311800, ChinaCenter for Research on Microgrids (CROM), Huanjiang Laboratory, Shaoxing 311800, ChinaAccurate heat load forecasting is essential for the efficiency of District Heating Systems (DHS). Still, it is challenged by the need to model long-term temporal dependencies and nonlinear relationships with weather and other factors. This study proposes a hybrid deep learning framework combining an attention-enhanced Neural Basis Expansion Analysis for Time Series (N-BEATS) model and eXtreme Gradient Boosting (XGBoost). The N-BEATS component, with a multi-head self-attention mechanism, captures temporal dynamics, while XGBoost models non-linear impacts of external variables. Predictions are integrated using an optimized weighted averaging strategy. Evaluated on a dataset from 103 heating units, the model outperformed 13 baselines, achieving an MSE of 0.4131, MAE of 0.3732, RMSE of 0.6427, and R<sup>2</sup> of 0.9664. This corresponds to a reduction of 32.6% in MSE, 32.0% in MAE, and 17.9% in RMSE, and an improvement of 5.1% in R<sup>2</sup> over the best baseline. Ablation studies and statistical tests confirmed the effectiveness of the attention mechanism and ensemble strategy. This model provides an efficient solution for DHS load forecasting, facilitating optimized energy dispatch and enhancing system performance.https://www.mdpi.com/1996-1073/18/15/3984district heating forecastingN-BEATSmulti-head self-attentionXGBoostmodel ensemble |
| spellingShingle | Shaohua Yu Xiaole Yang Hengrui Ye Daogui Tang Hamidreza Arasteh Josep M. Guerrero An Ensemble Model of Attention-Enhanced N-BEATS and XGBoost for District Heating Load Forecasting Energies district heating forecasting N-BEATS multi-head self-attention XGBoost model ensemble |
| title | An Ensemble Model of Attention-Enhanced N-BEATS and XGBoost for District Heating Load Forecasting |
| title_full | An Ensemble Model of Attention-Enhanced N-BEATS and XGBoost for District Heating Load Forecasting |
| title_fullStr | An Ensemble Model of Attention-Enhanced N-BEATS and XGBoost for District Heating Load Forecasting |
| title_full_unstemmed | An Ensemble Model of Attention-Enhanced N-BEATS and XGBoost for District Heating Load Forecasting |
| title_short | An Ensemble Model of Attention-Enhanced N-BEATS and XGBoost for District Heating Load Forecasting |
| title_sort | ensemble model of attention enhanced n beats and xgboost for district heating load forecasting |
| topic | district heating forecasting N-BEATS multi-head self-attention XGBoost model ensemble |
| url | https://www.mdpi.com/1996-1073/18/15/3984 |
| work_keys_str_mv | AT shaohuayu anensemblemodelofattentionenhancednbeatsandxgboostfordistrictheatingloadforecasting AT xiaoleyang anensemblemodelofattentionenhancednbeatsandxgboostfordistrictheatingloadforecasting AT hengruiye anensemblemodelofattentionenhancednbeatsandxgboostfordistrictheatingloadforecasting AT daoguitang anensemblemodelofattentionenhancednbeatsandxgboostfordistrictheatingloadforecasting AT hamidrezaarasteh anensemblemodelofattentionenhancednbeatsandxgboostfordistrictheatingloadforecasting AT josepmguerrero anensemblemodelofattentionenhancednbeatsandxgboostfordistrictheatingloadforecasting AT shaohuayu ensemblemodelofattentionenhancednbeatsandxgboostfordistrictheatingloadforecasting AT xiaoleyang ensemblemodelofattentionenhancednbeatsandxgboostfordistrictheatingloadforecasting AT hengruiye ensemblemodelofattentionenhancednbeatsandxgboostfordistrictheatingloadforecasting AT daoguitang ensemblemodelofattentionenhancednbeatsandxgboostfordistrictheatingloadforecasting AT hamidrezaarasteh ensemblemodelofattentionenhancednbeatsandxgboostfordistrictheatingloadforecasting AT josepmguerrero ensemblemodelofattentionenhancednbeatsandxgboostfordistrictheatingloadforecasting |