Machine learning-based predictive maintenance of a wire drawing machine using vibration and motor current signals
EKSPLOATACJA I NIEZAWODNOSC-MAINTENANCE AND RELIABILITY, cilt.29, sa.1, 2027 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 29 Sayı: 1
- Basım Tarihi: 2027
- Doi Numarası: 10.17531/ein/225078
- Dergi Adı: EKSPLOATACJA I NIEZAWODNOSC-MAINTENANCE AND RELIABILITY
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Applied Science & Technology Source, Compendex, Directory of Open Access Journals
- Açık Arşiv Koleksiyonu: AVESİS Açık Erişim Koleksiyonu
- Recep Tayyip Erdoğan Üniversitesi Adresli: Evet
Özet
Predictive maintenance improves reliability and reduces downtime in modern manufacturing systems. However, many studies rely on laboratory datasets or single-component monitoring, limiting their applicability to complex industrial environments. This study proposes a predictive maintenance framework for a multi-pass wire drawing machine using vibration and motor current signals from a real industrial production line. A Composite Health Index (CHI) is developed to transform multi-motor sensor data into an interpretable machine-level degradation indicator. The extracted features are used to train ensemble machine learning models including Random Forest, Extra Trees, and XGBoost, whose outputs are combined through a weighted hybrid ensemble model. A decision-layer mechanism with smoothing and temporal filtering is applied to reduce false alarms while preserving detection capability. Experimental results show that the model achieves a recall of 0.90 and an F1-score of 0.75, demonstrating its effectiveness for industrial applications.