Endüstriyel Makinelerin Arıza Durumlarına Göre Segmentasyonu: K-means ve Fuzzy C-means Algoritmaları ile RFM Analizi

Abstract

Bu çalışma makinelerin segmentasyonunu, bakım ve arıza kayıtlarına dayalı olarak RFM analizi ile değerlendirdikten sonra K-means ve Fuzzy C-means kümeleme algoritmaları kullanarak değerlendirmeyi amaçlamaktadır. Her bir makinenin arıza geçmişi makinelerin bakım ve arıza verileri analiz edilerek incelenmiştir. Makinelerin segmentasyonunu değerlendirmek amacıyla Arıza Frekansı, Toplam Arıza Süresi ve Son Arıza zamanı gibi parametreler kullanılmıştır. Bu parametreler, müdahale edilmesi gereken makinelerin belirlenmesini ve makinelerin operasyonel sağlık durumlarını anlaşılmasını sağlamıştır. Makine verileri üzerinde RFM analizi uygulandıktan sonra K-means ve Fuzzy C-means algoritmaları kullanılarak kümeleme yapılmıştır. Bu çalışma, makinelerin bakım süreçlerini optimize etmek, arıza eğilimlerini daha doğru tahmin etmek, operasyonel verimliliği artırmak ve maliyetleri düşürmek için veri odaklı bir yaklaşım sunmaktadır. Çalışma sonuçları David-Bouldin Index, Dunn Index, Calinski-Harabasz Index gibi metrikler kullanılarak kıyaslanmış ve en iyi kümelemeyi yapan algoritma seçilmiştir. Sonuçlar, makinelerin segmentlere ayrılmasını ve her segment için özel bakım ve iyileştirme stratejilerinin geliştirilmesini sağlamaktadır.

Keywords

• RFM Analizi
• K-means
• Fuzzy C-means
• Kestirimci Bakım

Thanks

Bu çalışmada kullanılan veri seti, Birinci Otomotiv firması tarafından sağlanmıştır. Kendilerine, değerli katkıları ve destekleri için teşekkür ederiz. Verilerin sağlanması, bu araştırmanın gerçekleştirilmesinde büyük bir yardımcı olmuştur ve çalışmanın başarısına önemli bir katkı sağlamıştır.

Segmentation of Industrial Machines Based on Fault Conditions: RFM Analysis with K-means and Fuzzy C-means Algorithms

Abstract

This study aims to evaluate machine segmentation based on maintenance and failure records using RFM analysis, followed by clustering with K-means and Fuzzy C-means algorithms. The failure history of each machine was analyzed based on maintenance and failure data. Parameters such as Failure Frequency, Total Failure Time, and Last Failure Time were used to assess the segmentation of the machines. These parameters helped identify machines that require intervention and better understand their operational health status. After applying RFM analysis to the machine data, clustering was performed using K-means and Fuzzy C-means algorithms. This study presents a data-driven approach to optimize machine maintenance processes, more accurately predict failure trends, increase operational efficiency, and reduce costs. The results were compared using metrics such as the David-Bouldin Index, Dunn Index, and Calinski-Harabasz Index, and the best clustering algorithm was selected. The findings facilitate the segmentation of machines and the development of specific maintenance and improvement strategies for each segment.

Keywords

• RFM Analysis
• K-means
• Fuzzy C-means
• Predictive Maintenance

References

1. [1] Shokrani A, Dhokia V, Newman ST. Environmentally conscious machining of difficult-to-machine materials with regard to cutting fluids. International Journal of Machine Tools and Manufacture. 2012; 57: 83–101.
2. [2] Sivakumar K, Mathan Kumar P, Amarkarthik A, Jegadheeswaran S, Shanmugaprakash R. Empirical modeling of material removal rate and surface roughness of OHNS steel using Cu-TiB2Tool in EDM. Materials Today: Proceedings. 2021; 45: 2725–2729.
3. [3] Matoorian P, Sulaiman S, Ahmad MMHM. An experimental study for optimization of electrical discharge turning (EDT) process. Journal Of Materials Processing Technology. 2008; 204: 350–356.
4. [4] Pant P, Bharti PS. Electrical Discharge Machining (EDM) of nickel-based nimonic alloys: A review. Materials Today: Proceedings. 2020; 25: 765–772.
5. [5] Bishnoi P, Sahu M. Optimization of the Keyseat Design with Consideration of Effect of Stress Concentration on Different Materials. International Journal of Engineering Research & Technology. 2014; 3: 477–481.
6. [6] Kishor HP, Raghu T. Desıgn Analysıs of A Keyless Couplıng. International Journal of Recent Research in Civil and Mechanical Engineering. 2014; 1: 37–43.
7. [7] Verma V, Sajeevan R. Multi Process Parameter Optimization of Diesinking EDM on Titanium Alloy (Ti6Al4V) Using Taguchi Approach. Materials Today: Proceedings. 2015; 2: 2581–2587.
8. [8] Klocke F, Mohammadnejad M, Holsten M, Ehle L, Zeis M, Klink A. A Comparative Study of Polarity-related Effects in Single Discharge EDM of Titanium and Iron Alloys. Procedia CIRP. 2018; 68: 52–57.

9. [9] Bhaumik M, Maity K. Effect of different tool materials during EDM performance of titanium grade 6 alloy. Engineering Science and Technology, An International Journal. 2018; 21: 507–516.
10. [10] Singh NK, Singh Y, Sharma A, prasad R. Experimental investigation of flushing approaches on EDM machinability during machining of titanium alloy. Materials Today: Proceedings. 2021; 38: 139–145.
11. [11] Prakash J, Equbal MdI, Equbal A. Investigative study on dimensional accuracy of machined cavity formed during EDM of AISI 1035. Materials Today: Proceedings. 2021; 47: 3217–3220.
12. [12] Kumar P, Dewangan S, Pandey C. Analysis of surface integrity and dimensional accuracy in EDM of P91 steels. Materials Today: Proceedings. 2020; 33: 5378–5383.
13. [13] Sanchez JA, Lopez de Lacalle LN, Lamikiz A, Bravo U. Dimensional accuracy optimisation of multi-stage planetary EDM. International Journal of Machine Tools and Manufacture. 2002; 42: 1643–1648.
14. [14] Çakıroğlu R, Günay M. Comprehensive analysis of material removal rate, tool wear and surface roughness in electrical discharge turning of L2 tool steel. Journal Of Materials Research and Technology. 2020; 9: 7305–7317.
15. [15] Singh H, Shukla DK. Optimizing electric discharge machining parameters for tungsten-carbide utilizing thermo-mathematical modelling. International Journal of Thermal Sciences. 2012; 59: 161–175.
16. [16] Unses E, Çoğun C. Improvement of Electric Discharge Machining (EDM) Performance of Ti-6Al-4V Alloy with Added Graphite Powder to Dielectric. Strojniski Vestnik - Journal of Mechanical Engineering. 2015; 61.
17. [17] Azhiri RB, Bideskan AS, Javidpour F, Tekiyeh RM. Study on material removal rate, surface quality, and residual stress of AISI D2 tool steel in electrical discharge machining in presence of ultrasonic vibration effect. The International Journal of Advanced Manufacturing Technology. 2019; 101: 2849–2860.
18. [18] Koenig W. The Flow Fields in the Working Gap with Electro Discharge Machining. Annals Of The CIRP. 1977; 25: 71–75.
19. [19] Schumacher BM. About the Role of Debris in the Gap During Electrical Discharge Machining. CIRP Annals. 1990; 39: 197–199.
20. [20]Obara H, Abe M, Ohsumi T. Control of Wire Breakage during Wire EDM -1st Report: Monitoring of Gap Signals According to Discharged Location. International Journal of Electrical Machining. 1999; 4: 53–58.
21. [21] Cakiroglu R, Günay M. Elektro Erozyonla Tornalama Yöntemiyle İşlenen Soğuk İş Takım Çeliğinin Yorulma Ömrünün Tahmini. Politeknik Dergisi. 2021; 24: 495–502.