EUVIMEDEUVIMED entdecken
Crossref

An efficient long-term daily PM2.5 concentration prediction method based on decomposition-ensemble and recursive error correction: A case study of the Beijing area

Yuanxun Cheng, Qingsong Hu, Dong Wu, Wenjie Gao, Liudi Wang, Hu Zhang

Environmental Research · 2026

DOI: 10.1016/j.envres.2026.124484 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Environmental Research, 2026. Verfasst von Yuanxun Cheng und 5 weiteren Autor:innen. Beteiligte Einrichtung: School of Information and Control Engineering, China University of Mining & Technology, Xuzhou, Jiangsu 221116, China. Hauptthemen laut MEDLINE: Particulate Matter, Environmental Monitoring, Air Pollutants, Air Pollution. Die Arbeit nennt 52 Literaturangaben. Gefördert von: National Natural Science Foundation of China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurate prediction of fine particulate matter (PM2.5) concentration is not only a key tool for guiding public health protection and intelligent travel decision-making, but also provides a scientific basis for tracing pollution sources and evaluating the environmental governance policies. However, existing approaches lack a systematic analysis and utilization of the characteristics of PM2.5 prediction error sequences. Therefore, this paper proposes a PM2.5 concentration prediction method based on decomposition-ensemble and Recursive Error Correction (REC). We show that the prediction error sequence approximately follows a normal distribution and exhibits predictability. Based on this property, a REC mechanism is proposed to suppress error accumulation within the prediction error sequence. Nevertheless, the presence of outliers in the error sequence can degrade the performance of REC. To address this issue, we incorporate the Olive-Hawkins estimator into the extreme learning machine to improve the robustness of output weight estimation, thereby enhancing the stability and accuracy of error sequence processing. Finally, the proposed method reconstructs the final prediction results by combining the preliminary forecasts obtained through decomposition-ensemble modeling with the correction values generated by REC. This paper forecasts the daily average PM2.5 concentrations in Beijing for the next 1, 5, and 10 days. The results demonstrate that the proposed model achieves superior predictive performance in long-term forecasting, consistently outperforming the comparative models. Notably, the incorporation of the REC mechanism yields an accuracy improvement of over 40%. The proposed method enhances the accuracy of long-term PM2.5 forecasting, providing valuable insights and tools for environmental pollution prevention and control.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Particulate MatterEnvironmental MonitoringAir PollutantsAir PollutionBeijingPrediction Algorithms

Stichwörter der Autor:innen

Air pollutionPM(2.5) concentration predictionRecursive error correctionVariational mode decomposition

Substanzen

Particulate MatterAir Pollutants

Autor:innen und Institutionen

  • School of Information and Control Engineering, China University of Mining & Technology, Xuzhou, Jiangsu 221116, China.
  • School of Information and Control Engineering, China University of Mining & Technology, Xuzhou, Jiangsu 221116, China. Electronic address: hqsong722@163.com.
  • School of Airspace Science and Engineering, Shandong University, Weihai, Shandong 264209, China.
  • Jinan USR IOT Technology Limited, Jinan, Shandong 250000, China.

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Cheng, Y., Hu, Q., Wu, D., Gao, W., Wang, L., & Zhang, H. (2026). An efficient long-term daily PM2.5 concentration prediction method based on decomposition-ensemble and recursive error correction: A case study of the Beijing area. Environmental Research. https://doi.org/10.1016/j.envres.2026.124484

Vancouver

Cheng Y, Hu Q, Wu D, Gao W, Wang L, Zhang H. An efficient long-term daily PM2.5 concentration prediction method based on decomposition-ensemble and recursive error correction: A case study of the Beijing area. Environmental Research. 2026. doi:10.1016/j.envres.2026.124484

BibTeX
@article{Cheng2026,
  author = {Cheng, Yuanxun and Hu, Qingsong and Wu, Dong and Gao, Wenjie and Wang, Liudi and Zhang, Hu},
  title = {An efficient long-term daily PM2.5 concentration prediction method based on decomposition-ensemble and recursive error correction: A case study of the Beijing area},
  journal = {Environmental Research},
  year = {2026},
  doi = {10.1016/j.envres.2026.124484}
}
Kostenlos starten →