Change search
Link to record
Permanent link

Direct link
BETA
Publications (8 of 8) Show all publications
Saari, J., Strömbergsson, D., Lundberg, J. & Thomson, A. (2019). Detection and identification of windmill bearing faults using a one-class support vector machine (SVM). Measurement, 137, 287-301
Open this publication in new window or tab >>Detection and identification of windmill bearing faults using a one-class support vector machine (SVM)
2019 (English)In: Measurement, ISSN 0263-2241, E-ISSN 1873-412X, Vol. 137, p. 287-301Article in journal (Refereed) Published
Abstract [en]

The maintenance cost of wind turbines needs to be minimized in order to keep their competitiveness and, therefore, effective maintenance strategies are important. The remote location of wind farms has led to an opportunistic maintenance strategy where maintenance actions are postponed until they can be handled simultaneously, once the optimal opportunity has arrived. For this reason, early fault detection and identification are important, but should not lead to a situation where false alarms occur on a regular basis. The goal of the study presented in this paper was to detect and identify wind turbine bearing faults by using fault-specific features extracted from vibration signals. Automatic identification was achieved by training models by using these features as an input for a one-class support vector machine. Detection models with different sensitivity were trained in parallel by changing the model tuning parameters. Efforts were also made to find a procedure for selecting the model tuning parameters by first defining the criticality of the system and using it when estimating how accurate the detection model should be. Method was able to detect the fault earlier than using traditional methods without any false alarms. Optimal combination of features and model tuning parameters was not achieved, which could identify the fault location without using any additional techniques.

Place, publisher, year, edition, pages
Elsevier, 2019
Keywords
Novelty detection, Wind turbine, Bearing fault diagnostics
National Category
Other Civil Engineering Tribology (Interacting Surfaces including Friction, Lubrication and Wear)
Research subject
Operation and Maintenance; Machine Elements
Identifiers
urn:nbn:se:ltu:diva-72839 (URN)10.1016/j.measurement.2019.01.020 (DOI)000464553200027 ()2-s2.0-85060852953 (Scopus ID)
Note

Validerad;2019;Nivå 2;2019-02-11 (svasva)

Available from: 2019-02-11 Created: 2019-02-11 Last updated: 2019-04-29Bibliographically approved
Strömbergsson, D., Marklund, P., Berglund, K., Saari, J. & Thomson, A. (2019). Mother wavelet selection in the discrete wavelet transform for condition monitoring of wind turbine drivetrain bearings. Wind Energy, 22(11), 1581-1592
Open this publication in new window or tab >>Mother wavelet selection in the discrete wavelet transform for condition monitoring of wind turbine drivetrain bearings
Show others...
2019 (English)In: Wind Energy, ISSN 1095-4244, E-ISSN 1099-1824, Vol. 22, no 11, p. 1581-1592Article in journal (Refereed) Published
Abstract [en]

Although the discrete wavelet transform has been used for diagnosing bearing faults for two decades, most work in this field has been done with test rig data. Since field data starts to be made more available, there is a need to shift into application studies.

The choice of mother wavelet, ie, the predefined shape used to analyse the signal, has previously been investigated with simulated and test rig data without consensus of optimal choice in literature. Common between these investigations is the use of the wavelet coefficients' Shannon entropy to find which mother wavelet can yield the most useful features for condition monitoring.

This study attempts to find the optimal mother wavelet selection using the discrete wavelet transform. Datasets from wind turbine gearbox accelerometers, consisting of enveloped vibration measurements monitoring both healthy and faulty bearings, have been analysed. The bearing fault frequencies' excitation level has been analysed with 130 different mother wavelets, yielding a definitive measure on their performance. Also, the applicability of Shannon entropy as a ranking method of mother wavelets has been investigated.

The results show the discrete wavelet transforms ability to identify faults regardless of mother wavelet used, with the excitation level varying no more than 4%. By analysing the Shannon entropy, broad predictions to the excitation level could be drawn within the mother wavelet families but no direct correlation to the main results. Also, the high computational effort of high order Symlet wavelets, without increased performance, makes them unsuitable.

Place, publisher, year, edition, pages
John Wiley & Sons, 2019
Keywords
bearing failure, condition monitoring, discrete wavelet transform, mother wavelet selection, wind turbine field measurements
National Category
Other Civil Engineering Tribology (Interacting Surfaces including Friction, Lubrication and Wear)
Research subject
Machine Elements; Operation and Maintenance
Identifiers
urn:nbn:se:ltu:diva-75777 (URN)10.1002/we.2390 (DOI)000480192400001 ()
Note

Validerad;2019;Nivå 2;2019-12-06 (johcin)

Available from: 2019-08-30 Created: 2019-08-30 Last updated: 2019-12-06Bibliographically approved
Saari, J. & Odelius, J. (2018). Detecting operation regimes using unsupervised clustering with infected group labelling to improve machine diagnostics and prognostics. Operations Research Perspectives, 5, 232-244
Open this publication in new window or tab >>Detecting operation regimes using unsupervised clustering with infected group labelling to improve machine diagnostics and prognostics
2018 (English)In: Operations Research Perspectives, ISSN 2214-7160, Vol. 5, p. 232-244Article in journal (Refereed) Published
Abstract [en]

Estimating the stress level of components while operation modes are varying is a key issue for many prognostic models in condition monitoring. The identification of operation profiles during production is therefore important. Clustering condition monitoring data with regard to operation regimes will provide more detailed information about the variation of stress levels during production. The distribution of the operation regimes can then support prognostics by revealing the cause-and-effect relationship between the operation regimes and the wear level of components.

In this study unsupervised clustering technique was used for detecting operation regimes for an underground LHD (load-haul-dump machine) by using features extracted from vibration signals measured on the front axle and the speed of the Cardan axle. The clusters were also infected with a small portion of the data to obtain the corresponding labels for each cluster. Promising results were obtained where each sought-for operation regime was detected in a sensible manner using vibration RMS values together with speed.

Place, publisher, year, edition, pages
Elsevier, 2018
National Category
Other Civil Engineering
Research subject
Operation and Maintenance
Identifiers
urn:nbn:se:ltu:diva-70391 (URN)10.1016/j.orp.2018.08.002 (DOI)000452765000023 ()2-s2.0-85051647575 (Scopus ID)
Note

Validerad;2018;Nivå 2;2018-08-27 (andbra)

Available from: 2018-08-15 Created: 2018-08-15 Last updated: 2019-04-23Bibliographically approved
Saari, J. (2018). Machinery diagnostic techniques for maintenance optimization. (Doctoral dissertation). Luleå: Luleå University of Technology
Open this publication in new window or tab >>Machinery diagnostic techniques for maintenance optimization
2018 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

One of the future challenges of machinery diagnostics and prognosticsis to prepare for the Internet of Things (IoT), where it is possible to change and improve existing approaches drastically. An intensifying application of the IoT will increase the use of embedded sensors and, therefore, create a demand for diagnostic tools where manual workis minimized and is mainly handled by smart algorithms. The auto-mated anomaly detection of large assets and their components with a system of smart algorithms needs proper optimization. Foremost, it is critical to avoid machinery failures, since they can interrupt production, cause unbearable production losses for the business and, even more, can put the lives of personnel in danger if a catastrophic failure occurs. On the other hand, if all the components are repeatedly creating false alarms, the verification of these incidents may be overwhelming. This research studied how a one-class SVM can be optimized by tuning the algorithm to function properly by taking the criticality of the system into consideration. Another topic dealt with was how a one-class SVM can be used for identifying the location of faults by carefully selecting proper input features. Furthermore, a method was tested where a variational Bayesian for Gaussian mixture algorithm was used for pre-processing and separating the condition monitoring data into operation mode classes. Later these classes can be used for improving the time for acquiring the condition monitoring data or to give more information as to how prognostic algorithms should be selected. In addition, a method was tested which involved the use of a Random Forest for feature selection and for the creation of indifference to load or other similar external factors by comparing separate classes with each other. Overall, the idea is that all of these tech-niques can be combined and merged in order to improve machinery diagnostic tools and prepare for the coming era of digitalization

Place, publisher, year, edition, pages
Luleå: Luleå University of Technology, 2018
Series
Doctoral thesis / Luleå University of Technology 1 jan 1997 → …, ISSN 1402-1544
National Category
Other Civil Engineering
Research subject
Operation and Maintenance
Identifiers
urn:nbn:se:ltu:diva-71376 (URN)978-91-7790-248-5 (ISBN)978-91-7790-249-2 (ISBN)
Public defence
2018-11-30, F1031, Luleå tekniska universitets, Luleå, 10:30
Opponent
Supervisors
Available from: 2018-10-30 Created: 2018-10-30 Last updated: 2018-11-21Bibliographically approved
Saari, J., Lundberg, J., Odelius, J. & Rantatalo, M. (2018). Selection of features for fault diagnosis on rotating machines using random forest and wavelet analysis. Insight (Northampton), 60(8), 434-442
Open this publication in new window or tab >>Selection of features for fault diagnosis on rotating machines using random forest and wavelet analysis
2018 (English)In: Insight (Northampton), ISSN 1354-2575, E-ISSN 1754-4904, Vol. 60, no 8, p. 434-442Article in journal (Refereed) Published
Abstract [en]

Identification of component faults using automated condition monitoring methods has a huge potential to improve the prediction of machine failures. The ongoing development of the Internet of Things (IoT) will support and benefit feature selection and improve preventative maintenance decision making. However, there may be problems with the selection of features that best describe a specific fault and remain valid even when the operation mode is changing (for example different levels of load). In this study, features were extracted from vibration signals using wavelet analysis; a feature subset was selected using the random forest ensemble technique. Three different datasets were created where the load of the system was changing while the rotational speed remained the same. The tests were repeated five times by first recording the nominal condition and then introducing four faults: angular misalignment; offset misalignment; partially broken gear tooth failure; and macro-pitting of the gear. To improve previous feature selection techniques, a method is proposed where, before training a classifier, the most promising features are compared at different degrees of torsional load. The results indicate that the proposed method of using random forests to select top variables can help to choose good features that may not have been considered in manual feature selection or in individual load zones.

Place, publisher, year, edition, pages
British Institute of Non-Destructive Testing, 2018
National Category
Other Civil Engineering
Research subject
Operation and Maintenance; Centre - SKF-LTU University Technology Cooperation
Identifiers
urn:nbn:se:ltu:diva-70433 (URN)10.1784/insi.2018.60.8.434. (DOI)000441327800006 ()2-s2.0-85051538361 (Scopus ID)
Note

Validerad;2018;Nivå 2;2018-08-16 (andbra)

Available from: 2018-08-16 Created: 2018-08-16 Last updated: 2019-03-26Bibliographically approved
Saari, J., Odelius, J., Lundberg, J. & Rantatalo, M. (2015). Using wavelet transform analysis and the support vector machine to detect angular misalignment of a rubber coupling. In: Sulo Lahdelma and Kari Palokangas (Ed.), Maintenance, Condition Monitoring and DiagnosticsMaintenance Performance Measurement and Management: . Paper presented at MCMD and MPMM 2015 conference, Oulu, Finland, 30 Sep - 5 Oct 2015 (pp. 117-126).
Open this publication in new window or tab >>Using wavelet transform analysis and the support vector machine to detect angular misalignment of a rubber coupling
2015 (English)In: Maintenance, Condition Monitoring and DiagnosticsMaintenance Performance Measurement and Management / [ed] Sulo Lahdelma and Kari Palokangas, 2015, p. 117-126Conference paper, Published paper (Other academic)
Abstract [en]

Shaft misalignment is a common problem for many types of rotating systems. It can cause machine breakdowns due to the premature failure of bearings or other components. Common diagnostic approaches rely on detecting increasing vibration response spectra at multiples of the shaft speed. However, in many time-variant systems, such as wind turbines, the speed and load vary considerably, which can make spectrum analysis insufficient. In this paper, a method for detecting shaft misalignment by using wavelet analysis is proposed. The method was experimentally evaluated in a laboratory test rig for four different operating conditions by varying the rotational speed and load. An angular misalignment was introduced between a hydraulic pump (load) and a medium-sized industrial gearbox connected with a rubber coupling. Vibration data were collected by using two accelerometers mounted in an axial and a radial direction directly on the gearbox casing. The features extracted from wavelet representation were classified by using a support vector machine algorithm. The detection of misalignment and the sensitivity of the proposed method are presented using validation data and confusion matrices. The final results of the confusion matrices clearly indicate that this method can detect misalignment even when the speed and load vary. The proposed method can be used for systems which are connected with shafts and there are many similar systems (comprising an electric motor, a gearbox and a centrifugal pump) working under the same circumstances.

Keywords
Shaft misalignment, wavelet tranform, SVM
National Category
Reliability and Maintenance Other Civil Engineering
Research subject
Operation and Maintenance
Identifiers
urn:nbn:se:ltu:diva-66432 (URN)
Conference
MCMD and MPMM 2015 conference, Oulu, Finland, 30 Sep - 5 Oct 2015
Available from: 2017-11-07 Created: 2017-11-07 Last updated: 2018-05-07Bibliographically approved
Mishra, M., Saari, J., Galar, D. & Leturiondo, U. (2014). Hybrid Models for Rotating Machinery Diagnosis and Prognosis: Estimation of Remaining Useful Life (ed.). Paper presented at . Luleå: Luleå tekniska universitet
Open this publication in new window or tab >>Hybrid Models for Rotating Machinery Diagnosis and Prognosis: Estimation of Remaining Useful Life
2014 (English)Report (Other academic)
Abstract [en]

The purpose of this literature review is to summarise the various technologies that can be used for machinery diagnosis and prognosis. The review focuses on Condition Based Maintenance (CBM) in machinery systems, with a short description of the theory behind each technology; it also includes references to state-of-the-art research into each theory. When we compare technologies, especially with respect to cost, complexity, and robustness, we find varied abilities across technologies. The machinery health assessment for CBM deployment is accepted worldwide; it is very popular in industries using rotating machines involved. These techniques are relevant in environments where predicting a failure and preventing or mitigating its consequences will increase both profit and safety. Prognosis is the most critical part of this process and is now recognised as a key feature in maintenance strategies; the estimation of Remaining Useful Life (RUL) is essential when a failure is identified. The literature review identifies three basic ways to model the fault development process: with symbols, data, or mathematical formulations based on physical principles. The review discusses hybrid approaches to machinery diagnosis and prognosis; it notes some typical approaches and discusses their advantages and disadvantages.

Place, publisher, year, edition, pages
Luleå: Luleå tekniska universitet, 2014. p. 73
Series
Technical report / Luleå University of Technology, ISSN 1402-1536
National Category
Other Civil Engineering
Research subject
Operation and Maintenance
Identifiers
urn:nbn:se:ltu:diva-22358 (URN)29348b08-6532-4abc-b189-9f5be8e742cf (Local ID)978-91-7439-968-4 (ISBN)978-91-7439-969-1 (ISBN)29348b08-6532-4abc-b189-9f5be8e742cf (Archive number)29348b08-6532-4abc-b189-9f5be8e742cf (OAI)
Note
Godkänd; 2014; 20140602 (madmis)Available from: 2016-09-29 Created: 2016-09-29 Last updated: 2018-05-07Bibliographically approved
Saari, J., Mishra, M., Galar, D. & Johansson, C.-A. (2013). Applied methods of condition monitoring and fault detection for underground mobile machines (ed.). Paper presented at MPES2013 : Mine Planing and Equipment Selection 14/10/2013 - 19/10/2013. Paper presented at MPES2013 : Mine Planing and Equipment Selection 14/10/2013 - 19/10/2013.
Open this publication in new window or tab >>Applied methods of condition monitoring and fault detection for underground mobile machines
2013 (English)Conference paper, Oral presentation only (Refereed)
Abstract [en]

Condition monitoring is a health assessment technique worldwide accepted and very popular in many industries especially where there are rotating machines involved in the processes. These techniques may be relevant in environments where the prediction of a failure and the prevention and mitigation of its consequences increase the profit and safety of the facilities. The maintenance of underground mobile mining equipment is one of these scenarios. It has several problem areas: harsh environment, potential risks and distant location of workshops. When a machine breaks down, there are two ways to handle the repair. Either the equipment has to be repaired on site at the production area or taken to the workshop. The difficulties involved in moving this type of large equipment are substantial but it might be difficult or unsafe to repair the LHD on site (depending on where and why it fails). Therefore it is necessary to identify the critical components and monitor them properly in order to skip undesired shutdowns or stoppages. This paper describes the benefits of different CM techniques applied to a critical part of a LHD machine (the transmission) in order to detect the abnormal behavior if any, identify the fault and predict the degradation. These techniques will provide enough information to optimize the maintenance actions minimizing and mitigating the costly effects of unplanned actions.

National Category
Other Civil Engineering
Research subject
Operation and Maintenance
Identifiers
urn:nbn:se:ltu:diva-32028 (URN)6626b0e3-5aeb-42f2-8278-3d965c57d8d0 (Local ID)6626b0e3-5aeb-42f2-8278-3d965c57d8d0 (Archive number)6626b0e3-5aeb-42f2-8278-3d965c57d8d0 (OAI)
Conference
MPES2013 : Mine Planing and Equipment Selection 14/10/2013 - 19/10/2013
Note
Godkänd; 2013; 20140829 (madmis)Available from: 2016-09-30 Created: 2016-09-30 Last updated: 2018-05-07Bibliographically approved
Organisations
Identifiers
ORCID iD: ORCID iD iconorcid.org/0000-0002-9599-1016

Search in DiVA

Show all publications