A Fleet-Wide Approach for Condition Monitoring of Similar Machines Using Time-Series Clustering
Abstract
The application of machine learning to fault diagnosis allows automated condition monitoring of machines, leading to reduced maintenance costs and increased machine availability. Traditional approaches train a machine learning algorithm to identify specific faults or operational settings. Therefore, these approaches cannot always cope with a dynamic industrial environment. However, an industrial installation often contains multiple machines of the same type, which enables a fleet-based analysis. This type of analysis compares machines to tackle the challenges of a dynamic environment. In this paper a novel method is proposed for analyzing a fleet of machines operating under similar conditions in the same area by using inter-machine comparisons. The proposed methodology consists of two steps. First, the inter-machine difference is calculated using dynamic time warping by using the amount of warping as measure. This method allows comparing the measured signals even when fluctuations are present. Second, a clustering method uses the inter-machine similarity to identify groups of machines that operate in a similar manner. The generation of a fault usually causes a change in the raw signals and diagnostic features. As a result, the inter-machine difference between the faulty machine and the rest of the fleet will increase, leading to the creation of a separate group that contains the faulty machine. The methodology is evaluated and validated on phase current signals measured on a fleet of electrical drivetrains, where a phase unbalance fault is introduced in some of the drivetrains for a specific time duration.
Keywords
Condition monitoring Fleet monitoring Dynamic time warping Clustering Phase unbalanceNotes
Acknowledgments
The authors acknowledge the financial support of VLAIO (Flemish Innovation & Entrepreneurship) through the Baekeland PhD mandate (nr. HBC.2017.0226) and the O&O project REFLEXION (nr. IWT. 150334). Jesse Davis is partially support by the KU Leuven research funds (C14/17/070). The drivetrain fleet data was obtained in the laboratory of the ULB Beams group (http://www.beams.ulb.ac.be) with the support of Dr. Yves Mollet and Prof. Dr. Johan Gyselinck.
References
- 1.Lee J, Kao H-A, Yang S (2014) Service innovation and smart analytics for Industry 4.0 and Big Data environment. In: 6th conference on industrial product-service systems, Windsor, Ontario, Canada, vol 16. Elsevier B.V., pp 3–8Google Scholar
- 2.Siegel D (2013) Prognostics and health assessment of a multi-regime system using a residual clustering health monitoring approach. Ph.D. thesis, University of CincinnatiGoogle Scholar
- 3.Matthews BL, Das S, Bhaduri K, Das K, Martin R, Oza N (2013) Discovering anomalous aviation safety events using scalable data mining algorithms. J Aerosp Inf Syst 10(10):467–475Google Scholar
- 4.Wong MLD, Jack LB, Nandi AK (2006) Modified self-organising map for automated novelty detection applied to vibration signal monitoring. Mech Syst Sig Process 20(3):593–610CrossRefGoogle Scholar
- 5.Chu E, Gorinevsky D, Boyd S (2011) Scalable statistical monitoring of fleet data. In: IFAC proceedings volumes (IFAC-PapersOnline), vol 18. IFAC, pp 13227–13232Google Scholar
- 6.Markou M, Singh S (2003) Novelty detection: a review - part 1: statistical approaches. Sig Process 83(12):2481–2497CrossRefGoogle Scholar
- 7.Brandt T, Grawunder M, Appelrath H-J (July 2016) Anomaly detection on data streams for machine condition monitoring. In: 2016 IEEE 14th international conference on industrial informatics (INDIN), Poitiers. IEEE, pp 1282–1287Google Scholar
- 8.Narwade S, Kulkarni P, Patil CY (2013) Fault detection of induction motor using envelope analysis. Int J Adv Comput Res (ISSN (print)) 2(7):258–262Google Scholar
- 9.Cummings PB, Dunki-Jacobs JR, Kerr RH (1985) Protection of induction motors against unbalanced voltage operation. IEEE Trans Ind Appl IA-21(3):778–792CrossRefGoogle Scholar
- 10.Han T, Yang B-S, Yin Z-J (2007) Feature-based fault diagnosis system of induction motors using vibration signal. J Qual Maintenance Eng 13(2):163–175CrossRefGoogle Scholar
- 11.Cablea G, Granjon P, Bérenguer C, Bellemain P (2015) Online condition monitoring of wind turbines through three phase electrical signature analysis. In: Twelve international conference on condition monitoring and machinery failure prevention technologies CM 2015, Oxford, UKGoogle Scholar
- 12.Zhen D, Wang T, Gu F, Ball AD (2013) Fault diagnosis of motor drives using stator current signal analysis based on dynamic time warping. Mech Syst Sig Process 34(1–2):191–202CrossRefGoogle Scholar
- 13.Wang C, Alvarez SA, Ruiz C, Moonis M (2016) Deviation-based dynamic time warping for clustering human sleep. In: Proceedings of the 9th International Joint Conference on Biomedical Engineering Systems and Technologies, Rome, Italy, vol 4. SCITEPRESS - Science and and Technology Publications, pp 88–95Google Scholar
- 14.Müller M (2007) Dynamic time warping. In: Information retrieval for music and motion, 1 edn. Springer, Heidelberg, pp 69–84Google Scholar
- 15.Bemdt DJ, Clifford J (1994) Using dynamic time warping to find patterns in time series. In: AAAIWS 1994 proceedings of the 3rd international conference on knowledge discovery and data mining, Seattle. AAAI Press, pp 359–370Google Scholar
- 16.Silva DF, Batista GEAPA, Keogh E (2016) On the effect of endpoints on dynamic time warping. In: SIGKDD MiLeTS 2016, San Francisco, California. ACM, p 10Google Scholar
- 17.Liang B (September 2010) A hierarchical clustering based global outlier detection method. In: 2010 IEEE Fifth international conference on bio-inspired computing: theories and applications (BIC-TA), Changsha, China. IEEE, pp 1213–1215Google Scholar
- 18.Meert W (April 2018) wannesm/dtaidistance: v1.0.1 (Version v1.0.1). Zenodo. https://doi.org/10.5281/zenodo.1215946