Gated Residual Fusion Framework for Mitigating Error Accumulation in Multistep NOx Prediction for Utility CFB Boilers
5 affiliations
- 1School of Intelligent Manufacturing, Jianghan University, Wuhan 430056, P. R. China.
- 2Hubei Engineering Research Center for Intelligent Detection and Control of Specialized Equipment, Jianghan University, Wuhan 430056, P. R. China.
- 3School of Engineering, University of Edinburgh, Edinburgh EH9 3FG, U.K.
- 4School of Energy and Power, Jiangsu University of Science and Technology, Zhenjiang 212100, Jiangsu, P. R. China.
- 5School of Electrical Engineering, Nanjing Vocational University of Industry Technology, Nanjing 210023, P. R. China.
Abstract
Accurate multistep prediction of nitrogen oxide (NOx) emissions is essential for combustion perception and optimization in utility boilers. Under deep peak-shaving conditions, frequent operating condition transitions induce mixed operating data, strong variable coupling, and progressive error accumulation, restricting the long-horizon stability of conventional models. To address these issues, a gated residual fusion framework was constructed by using data from a 300 MW circulating fluidized bed (CFB) boiler. A preprocessing strategy integrating rolling signal denoising, correlation analysis, and Random Forest ranking selected 30 variables from 162 candidates. iTransformer and TimesNet were coupled to extract complementary dynamic representations, and sequence decomposition fusion, parallel fusion, and gated residual fusion were compared. In the proposed architecture, the dominant NOx evolution is captured by the iTransformer backbone, while the TimeNet residual branch provides correction for local deviations that are insufficiently represented by the backbone. Compared with standalone VMD-iTransformer and VMD-TimesNet, RMSE was reduced by 73.99% and 78.37% at step 1 and by 39.86% and 49.51% at step 6, respectively. Compared with eight baseline models, the proposed model achieved the best average performance, with RMSE, MAE, MAPE, and R2 values of 3.2228 mg/m3, 1.8930 mg/m3, 5.2338%, and 0.9185. Results demonstrate peak recovery, abrupt change tracking, and long-horizon stability.
Related
- Matching Wind-Solar Portfolios with Electricity Consumption Across Climates: A Physical Framework to Minimize Residual Demand
- Photocatalytic strategies for mitigating microplastic pollution and greenhouse gas emissions
- Mitigating Heavy Metal Uptake in Wastewater-Irrigated Radish (Raphanus sativus L.) through Biochar and L-Methionine Application
- Reservoir eutrophication prediction with explainable framework integrating temporal convolutional networks and bidirectional sequence modeling
- A cross-regional machine learning transfer prediction framework for heat risk based on environmental characteristics of mature new towns
BibTeX
@article{liao2026gated,
title = {{Gated Residual Fusion Framework for Mitigating Error Accumulation in Multistep NOx Prediction for Utility CFB Boilers}},
author = {Guangting Liao and Cong Yu and Yukun Zhu and Wei Fan and Haiquan Yu and Qiang Wang},
journal = {ACS Omega},
year = {2026},
month = oct,
doi = {10.1021/acsomega.6c09912},
url = {https://pubs.acs.org/acsodf/article/11/41/64249/5446711/Gated-Residual-Fusion-Framework-for-Mitigating},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - JOUR
TI - Gated Residual Fusion Framework for Mitigating Error Accumulation in Multistep NOx Prediction for Utility CFB Boilers
AU - Guangting Liao
AU - Cong Yu
AU - Yukun Zhu
AU - Wei Fan
AU - Haiquan Yu
AU - Qiang Wang
PY - 2026
DA - 2026/10/08
JO - ACS Omega
DO - 10.1021/acsomega.6c09912
UR - https://pubs.acs.org/acsodf/article/11/41/64249/5446711/Gated-Residual-Fusion-Framework-for-Mitigating
AB - Accurate multistep prediction of nitrogen oxide (NOx) emissions is essential for combustion perception and optimization in utility boilers. Under deep peak-shaving conditions, frequent operating condition transitions induce mixed operating data, strong variable coupling, and progressive error accumulation, restricting the long-horizon stability of conventional models. To address these issues, a gated residual fusion framework was constructed by using data from a 300 MW circulating fluidized bed (CFB) boiler. A preprocessing strategy integrating rolling signal denoising, correlation analysis, and Random Forest ranking selected 30 variables from 162 candidates. iTransformer and TimesNet were coupled to extract complementary dynamic representations, and sequence decomposition fusion, parallel fusion, and gated residual fusion were compared. In the proposed architecture, the dominant NOx evolution is captured by the iTransformer backbone, while the TimeNet residual branch provides correction for local deviations that are insufficiently represented by the backbone. Compared with standalone VMD-iTransformer and VMD-TimesNet, RMSE was reduced by 73.99% and 78.37% at step 1 and by 39.86% and 49.51% at step 6, respectively. Compared with eight baseline models, the proposed model achieved the best average performance, with RMSE, MAE, MAPE, and R2 values of 3.2228 mg/m3, 1.8930 mg/m3, 5.2338%, and 0.9185. Results demonstrate peak recovery, abrupt change tracking, and long-horizon stability.
SN - 2470-1343
ER -