Портал психологических изданий PsyJournals.ru
Каталог изданий 108Рубрики 53Авторы 9084Новости 1792Ключевые слова 5095 Правила публикацииВебинарыRSS RSS


0,727 — двухлетний импакт-фактор

Моделирование и анализ данных

Издатель: Московский государственный психолого-педагогический университет

ISSN (печатная версия): 2219-3758

ISSN (online): 2311-9454

DOI: https://doi.org/10.17759/mda

Лицензия: CC BY-NC 4.0

Издается с 2011 года

Периодичность: 4 номера в год

Язык журнала: русский

Доступ к электронным архивам: открытый


Hidden Markov models in task of condition monitoring 743

Gałęzia A.
Institute of Automotive Engineering, Warsaw University of Technology, Warsaw, Poland

Radkowski S.
Institute of Automotive Engineering, Warsaw University of Technology


This paper describes hidden Markov models use in technical objects condition monitoring. A hidden Markov model is a statistical model in which it is assumed that a system is Markov process with unknown parame¬ters. Based on observations generated by the system, model parameters are extracted. Models described by these parameters are used for further analysis and prediction studies. The paper discusses relationships and differences between a visible Markov model and its hidden counterpart. In considered task hidden Markov models are used to detect changes in the technical state of examined object. The main problem of method developed is to elaborate method which will be able to detect significant changes in a system observation vector and ascribe a given observation to the present technical state. An example of using hidden Markov models for rolling bearings is presented.

Ключевые слова: Hidden Markov models, condition monitoring, vibro-acoustic signals, signal symbolization, bearings

Рубрика: Научная жизнь

Тип: научная статья

Ссылка для цитирования

Фрагмент статьи

Modern research is often aimed at building such a diagnostic system which will allow to determine the technical state of tested object. Correct and certain determination of technical state is important in condition monitoring and makes it possible to determine future operating strategies for minimiz­ing operating costs and failure hazards. Change of technical state is usually connected with changes in the object kinematics, development of failures, changes in cooperated kinematics pairs.

  1. Daw C.S., Finney C.E.A., Tracy E.R., “A review of symbolic analysis of experimental data”, http://www-chaos.engr.utk.edu/pap/crg-rsi2002.pdf.
  2. Ray A., “Symbolic dynamic analysis of complex systems for anomaly detection”, Signal Pro­cessing, Vol 84, 2004, pp 1115-1130.
  3. Galezia A., Radkowski S., “Problem of proper choice of discretization and quantization in sym­bolic transformation of signals”, Proceedings of the Third WCEAM-IMS, pp 509-514, 2008. 
  4. Galezia A., Radkowski S., “Usage of symbolic analysis in condition monitoring”, Proceedings of the fifth International conference on Condition Monitoring and Machinery Failure Prevention Technologies, UK, 2008.
  5. Rabiner L.R., “A tutorial on HMM and selected applications in speech recognition”, Proceed­ings of the IEEE, Vol 77, No 2, pp 257-286, February 1989.
  6. Ibe O.C., “Markov processes for stochastic modelling”, Elsevier Academic Press, Oxford, UK, 2008.
  7. http://www.cs.ubc.ca/~murphyk/Software/HMM/hmm_usage.html.
  8. http://www.igi.tugraz.at/lehre/CI/tutorials/HMM.zip.
О проекте PsyJournals.ru

© 2007–2020 Портал психологических изданий PsyJournals.ru  Все права защищены

Свидетельство регистрации СМИ Эл № ФС77-66447 от 14 июля 2016 г.


Creative Commons License Репозиторий открытого доступа     Рейтинг репозиториев Webometrics