EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Comparative Study of Decision-Level Fusion Strategies for Multi-Sensor CNN-Based Bearing Fault Diagnosis

Iman Makrouf iD, Mourad Zegrari iD, Khalid Dahi iD, Demba Diallo iD, Meryem Abtane, Ilias Ouachtouk

Entropy · 12.09.2026 · 28 (9) · 1020

DOI: 10.3390/e28091020 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Entropy, online erschienen am 12.09.2026. Verfasst von Iman Makrouf und 5 weiteren Autor:innen. Beteiligte Einrichtung: Digital Engineering for Leading Technology and Automation, National Higher School of Arts and Crafts Casablanca, Hassan II University of Casablanca, …. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 29 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Bearing fault diagnosis increasingly relies on multiple sensors, since compound or multi-location faults can produce signatures that are only partially captured by a single sensor. The decision-level fusion (DLF) of independently trained models offers a practical way to combine such complementary information, yet systematic comparisons across DLF techniques remain scarce, particularly for measurements from different sensor locations. This paper benchmarks six DLF strategies, i.e, Max, Average, Majority Voting, Weighted Sum, Dempster–Shafer, and Stacking, on a dual-branch one-dimensional convolutional neural network (1D-CNN) with each branch trained end-to-end on vibration signals from a distinct bearing location. On a two-sensor test bench covering seven health conditions, all methods exceed 99.7% accuracy on clean signals, while Dempster–Shafer fusion proves markedly more robust under noise, retaining up to 84% accuracy at a 5 dB signal-to-noise ratio (SNR). A conflict-coefficient analysis further provides an interpretable account of when fusion succeeds, linking performance to the confidence complementarity between branches.

Schlagwörter

Stichwörter der Autor:innen

Dempster–Shafer theoryconvolutional neural networkdecision-level fusionmulti-sensor fusionnoise stress

Autor:innen und Institutionen

  • Digital Engineering for Leading Technology and Automation, National Higher School of Arts and Crafts Casablanca, Hassan II University of Casablanca, Casablanca 20670, Morocco.
  • Complex Systems and Interactions, Ecole Centrale Casablanca, Bouskoura 27182, Morocco.
  • GeePs, CNRS, CentraleSupélec, Université Paris-Saclay, 91190 Gif-sur-Yvette, France.

ORCID-Profile: Iman Makrouf, Mourad Zegrari, Khalid Dahi, Demba Diallo

Zitieren

APA

Makrouf, I., Zegrari, M., Dahi, K., Diallo, D., Abtane, M., & Ouachtouk, I. (2026). Comparative Study of Decision-Level Fusion Strategies for Multi-Sensor CNN-Based Bearing Fault Diagnosis. Entropy, 28(9), 1020. https://doi.org/10.3390/e28091020

Vancouver

Makrouf I, Zegrari M, Dahi K, Diallo D, Abtane M, Ouachtouk I. Comparative Study of Decision-Level Fusion Strategies for Multi-Sensor CNN-Based Bearing Fault Diagnosis. Entropy. 2026;28(9):1020. doi:10.3390/e28091020

BibTeX
@article{Makrouf2026,
  author = {Makrouf, Iman and Zegrari, Mourad and Dahi, Khalid and Diallo, Demba and Abtane, Meryem and Ouachtouk, Ilias},
  title = {Comparative Study of Decision-Level Fusion Strategies for Multi-Sensor CNN-Based Bearing Fault Diagnosis},
  journal = {Entropy},
  year = {2026},
  volume = {28},
  number = {9},
  pages = {1020},
  doi = {10.3390/e28091020}
}
Kostenlos starten →