A Novel Carbon Emission Calculation Method for Power System Based on Personalized Transformer with Two-Stage Training

Bixuan Gao; Riwei Zhang; Xiangyu Kong; Gaohua Liu; Kaijie Fang; Meimei Duan

doi:10.1016/j.eng.2026.01.005

Engineering ›› :202601005 DOI: 10.1016/j.eng.2026.01.005

Research

research-article

A Novel Carbon Emission Calculation Method for Power System Based on Personalized Transformer with Two-Stage Training

Author information +

History +

PDF

Abstract

Accurately and comprehensively calculating carbon emissions in the power system is a fundamental pre-requisite for achieving low-carbon energy transitions. Existing carbon flow theory-based methods pri-marily concentrate on emissions from the grid and load sides, while the methods for generation-side emissions often rely on costly continuous monitoring systems or imprecise default emission factors. However, in practice, most power generation units cannot achieve real-time and precise carbon emission measurement, leading to generation side data deviations that affect overall computational accuracy. To address the limitations, this paper proposes a novel carbon emission measurement method based on heterogeneous data and personalized Transformer, applicable to various power generation units. This method has several key innovations: ① extends traditional total electricity production based models are extended to a multi-feature framework to capture similarities and differences in time-series data among various generator units; ② an improved Transformer is designed that integrates short-term rela-tionship extraction, long-term differential identification and long-term and short-term feature fusion modules to enhance the multi-feature based emission mapping process, and ③ a two stage training pro-tocol is adopted, with self-supervised pretraining of feature extractors followed by fine tuning, to accel-erate convergence and improve accuracy. Experiments on real-world generation-unit data show that the proposed method reduces average RMSE by 22.3% relative to a standard Transformer and by 15.9% rela-tive to Informer. Further validation utilizing the Institute of Electrical and Electronics Engineers (IEEE) 30-bus test case confirms the effectiveness and applicability of the model for carbon emission measurement across all segments of the power system.

Keywords

Power system carbon emissions / Carbon emission calculation / Transformer / Multi-frequency feature mapping

Cite this article

Download citation ▾

Bixuan Gao, Riwei Zhang, Xiangyu Kong, Gaohua Liu, Kaijie Fang, Meimei Duan. A Novel Carbon Emission Calculation Method for Power System Based on Personalized Transformer with Two-Stage Training. Engineering 202601005 DOI:10.1016/j.eng.2026.01.005

登录浏览全文

4963

注册一个新账户忘记密码

References

Publishing order | Descend order by publishing year | Descend order by cited within

[1]	Liu Y, Li Y, Zhou C, Song J, Deng H, Du E, et al. Overview of carbon measurement and analysis methods in power systems. Proc CSEE 2024; 44(6):2220-36 [Chinese].

[2]	Kang C, Du E, Li Y, Zhang N, Chen Q, Guo H, et al. Key scientific problems and research framework for carbon perspective research of new power systems. Power Syst Technol 2022; 46(03):821-33.

[3]	Wu C, Zhou Y, Gan W, Wu J. Robust scheduling of a pulp and paper mill considering flexibility provision from steam power generation. Appl Energy 2025;377:124595.

[4]	Gao B, Kong X, Liu G, Xiang T, Gao Y, Luo S, et al. Monitoring high-carbon industry enterprise emission in carbon market: a multi-trusted approach using externally available big data. J Clean Prod 2024;466:142729.

[5]	Zhang N, Li Y, Huang J, Li Y, Du E, Li M, et al. Carbon measurement method and carbon meter system for whole chain of power system. Autom Electr Power Syst 2023; 47(9):2-12.

[6]	Zhou T, Kang C, Xu Q, Chen Q. Preliminary investigation on a method for carbon emission flow calculation of power system. Autom Electr Power Syst 2012; 36(11):44-9.

[7]	Sun X, Bao M, Ding Y, Hui H, Song Y, Zheng C, et al. Modeling and evaluation of probabilistic carbon emission flow for power systems considering load and renewable energy uncertainties. Energy 2024;296:130768.

[8]	Lai G, Ye Q, Chen W, Hu Z, Hong L, Wang Y, et al. Electricity-carbon modeling of flat glass industry based on correlation variable. Energy Rep 2022; 8 (13):1265-74.

[9]	Zhou C, Lin X, Wang R, Song B. Real-time carbon emissions monitoring of high-energy-consumption enterprises in Guangxi based on electricity big data. Energies 2023; 16(13):5124.

[10]	Zhang S, Chen H, Wang B, Yu R, Ma Z, Miao Y. Carbon emission monitoring based on analysis from “electricity-carbon” relationship of cement enterprises. China Environ Sci 2023; 43(7):3787-95.

[11]	Zhou P, Ang BW, Wang H. Energy and CO₂ emission performance in electricity generation: a non-radial directional distance function approach. Eur J Oper Res 2012; 221(3):625-35.

[12]	Liu J, Zhao H, Wang S, Liu G, Zhao J, Dong ZY. Real-time emission and cost estimation based on unit-level dynamic carbon emission factor. Energy Convers Econ 2023; 4(1):47-60.

[13]	Liu Z, Sun T, Yu Y, Ke P, Deng Z, Lu C, et al. Near-real-time carbon emission accounting technology toward carbon neutrality. Engineering 2022;14:44-51.

[14]	Liu G, Liu J, Zhao J, Wen F, Xue Y.A real-time estimation framework of carbon emissions in steel plants based on load identification. In:Proceedings of 2020 International Conference on Smart Grids and Energy Systems (SGES); 2020 Nov 26-28; Perth, Australia. IEEE; 2020. p. 988-93.

[15]	Liu J, Liu G, Zhao H, Zhao J, Qiu J, Dong ZY. A real-time carbon emission estimation framework for industrial parks with non-intrusive load monitoring. Sustain Energy Technol Assess 2023;60:103482.

[16]	Liu G, Liu J, Zhao J, Qiu J, Mao Y, Wu Z, et al. Real-time corporate carbon footprint estimation methodology based on appliance identification. IEEE Trans Industr Inform 2023; 19(2):1401-12.

[17]	Yang C, Liang G, Liu J, Liu G, Yang H, Zhao J, et al. A non-intrusive carbon emission accounting method for industrial corporations from the perspective of modern power systems. Appl Energy 2023;350:121712.

[18]	Chen F, Gao Y, Wang J, Wu M, Zhang W, Teng F. Real-time carbon emission estimation for industrial users with low RMSE based on NILM and evolutionary algorithm. IEEE Trans Instrum Meas 2024;73:1-11.

[19]	Zeng J, He G, Li Y, Du E, Zhang N, Zhu J. A short-term carbon emission accounting method using electricity data based on convolutional neural networks and light gradient boosting machine combined model in power industry. J Shanghai Jiaotong Univ 2025;6:746-57 [Chinese].

[20]	Xia Y, Sun G, Wang Y, Yang Q, Wang Q, Ba S. A novel carbon emission estimation method based on electricity-carbon nexus and non-intrusive load monitoring. Appl Energy 2024;360:122773.

[21]	Wang Y, Wu H, Dong J, Liu Y, Qiu Y, Zhang H, et al. TimeXer: empowering transformers for time series forecasting with exogenous variables. arXiv:2402.19072; 2024.

[22]	Wu H, Hu T, Liu Y, Zhou H, Wang J, Long M. TimesNet: temporal 2D-variation modeling for general time series analysis. arXiv:2210.02186; 2022.

[23]	Zhang H, Li B, Su SF, Yang W, Xie L. A novel hybrid transformer-based framework for solar irradiance forecasting under incomplete data scenarios. IEEE Trans Industr Inform 2024; 20(6):8605-15.

[24]	Breunig MM, Kriegel HP, Ng RT, Sander J. LOF:identifying density-based local outliers. In: ACM International Conference on Management of Data and Symposium on Principles of Database Systems; 2000 May 15-18; Dallas, TX, USA. ACM; 2000. p. 93-104.

[25]	García-Laencina PJ, Sancho-Gómez JL, Figueiras-Vidal AR, Verleysen M. K nearest neighbours with mutual information for simultaneous classification and missing data imputation. Neurocomputing 2009; 72(7-9):1483-93.

[26]	Kang C, Zhou T, Chen Q, Wang J, Sun Y, Xia Q, et al. Carbon emission flow from generation to demand: a network-based model. IEEE Trans Smart Grid 2015; 6 (5):2386-94.

[27]	Wang Y, Wu L, Wang S. A fully-decentralized consensus-based ADMM approach for DC-OPF with demand response. IEEE Trans Smart Grid 2017; 8 (6):2637-47.

[28]	Ma K, Nie X, Yang J, Zha L, Li G, Li H. A power load forecasting method in port based on VMD-ICSS-hybrid neural network. Appl Energy 2025;377:124246.

[29]	Chen H, Lu T, Huang J, He X, Sun X. An improved VMD-EEMD-LSTM time series hybrid prediction model for sea surface height derived from satellite altimetry data. J Mar Sci Eng 2023; 11(12):2386.

[30]	Liang Y, Zhang D, Zhang J, Hu G. A state-of-the-art analysis on decomposition method for short-term wind speed forecasting using LSTM and a novel hybrid deep learning model. Energy 2024;313:133826.

[31]	Liu J, Huang X, Li Q, Chen Z, Liu G, Tai Y. Hourly stepwise forecasting for solar irradiance using integrated hybrid models CNN-LSTM-MLP combined with error correction and VMD. Energy Convers Manage 2023;280:116804.

[32]	Bai S, Kolter JZ,Koltun V. An empirical evaluation of generic convolutional and recurrent networks for sequence modeling. arXiv:1803.01271; 2018.

[33]	Zhang T, Huang Y, Hui L, Liang Y. A hybrid electric vehicle load classification and forecasting approach based on GBDT algorithm and temporal convolutional network. Appl Energy 2023;351:121768.

[34]	Qi S, Cheng S, Tan X, Feng S, Zhou Q. Predicting China’s carbon price based on a multi-scale integrated model. Appl Energy 2022;324:119784.

[35]	Wei D, Deng Y. Redistributed invariant redundant fractional wavelet transform and its application in watermarking algorithm. Expert Syst Appl 2025;262:125707.

[36]	Gao B, Huang X, Shi J, Tai Y, Zhang J. Hourly forecasting of solar irradiance based on CEEMDAN and multi-strategy CNN-LSTM neural networks. Renew Energy 2020;162:1665-83.

[37]	Shao Z, Gao S, Zhou K, Yang S. A new multiregional carbon emissions forecasting model based on a multivariable information fusion mechanism and hybrid spatiotemporal graph convolution network. J Environ Manage 2024;352:119976.

[38]	Wu B, Zeng H, Wang Z, Wang L. Interpretable short-term carbon dioxide emissions forecasting based on flexible two-stage decomposition and temporal fusion transformers. Appl Soft Comput 2024;159:111639.

[39]	Zhang Y, Chen B, Pan G, Zhao Y. A novel hybrid model based on VMD-WT and PCA-BP-RBF neural network for short-term wind speed forecasting. Energy Convers Manage 2019;195:180-97.

[40]	Wang Y, Wu L. On practical challenges of decomposition-based hybrid forecasting algorithms for wind speed and solar irradiation. Energy 2016;112:208-20.

[41]	Sun J, Kong X, Yang Z, Xiang T, Wang Y, Gao Y, et al. Development of an urban carbon emission accounting and dynamic warning framework using multi-source cross-domain big data: a case study of municipalities. China. J Clean Prod 2025;517:145873.

[42]	Li Y, Zhang N, Du E, Liu Y, Cai X, He D. Mechanism study and benefit analysis on power system low carbon demand response based on carbon emission flow. Proc CSEE 2022; 42(8):2830-42 [Chinese].