D.M. Zhuk – Associate Professor, 

Department «Systems of the Automated Designing», Bauman Moscow State Technical University

E-mail: zhuk@bmstu.ru 

T.M. Volosatova – Associate Professor, 

Department «Systems of the Automated Designing», Bauman Moscow State Technical University

E-mail: tamaravol@gmail.com

А.Yu. Spasenov – Assistant, 

Department «Systems of the Automated Designing», Bauman Moscow State Technical University

E-mail: a.spasenov@bmstu.ru

K.V. Kucherov – Assistant, 

Department «Computer systems and networks», Bauman Moscow State Technical University

Email: cvkucherov@yandex.ru

The article presents a method for analyzing multidimensional time series to assess the state of dynamical systems. The method can be used in the processing and interpretation of data of various physical nature to solve the problems of diagnosis and monitoring the status of complex technical systems. An approach is formulated to assess changes in the state of technical systems using modal linguistic analysis. The possibilities of using the window Fourier transform and the method of reducing the dimension of the space of characteristic features to obtain a character sequence describing the dynamics of changes in the states of the system under study are considered. Analysis and interpretation of the obtained sequences can significantly expand the diagnostic capabilities of complex technical systems evaluating. The basic ideas of the technology are presented, illustrating the main points that are important for the application of technology in order to identify defects in technical systems. Possible modifications of the approach are proposed for the analysis of multidimensional quasiperiodic time series with a trend.

The possibility of using the method of modal-linguistic analysis of multidimensional time series to solve the problem of assessing the state of complex technical systems is shown, which allows us to use this approach to identify various interpreted patterns in multidimensional time series. The results of using the proposed method for the analysis of time series describing fluctuations in a technical system with the appearance of a defect are presented. A possible modification of the proposed method is the use of alternative characteristic features, such as the result of applying the wavelet transform or empirical modal decomposition to the initial time series, showing high classification accuracy in solving the problem. The initial data for the analysis of many ongoing technical, socioeconomic and biomedical processes are often given in the form of quasiperiodic time series. Of particular interest is the use of pattern search methods in time series characterizing specific fragments with the subsequent extraction of characteristic features. This approach will significantly reduce the number of processed segments and will allow you to evaluate the dynamics of changes in characteristic features only specific fragments of the record. In the future, it is planned to expand this method for the analysis of quasiperiodic signals with a trend based on parametric modal analysis using the Prony method.

1. Bassvil' M. Banvenist A. Obnaruzhenie izmenenij svojstv signalov i dinamicheskih sistem. M.: Mir. 1989. 278 s. (in Russian).
2. Anciperov V.E. Mnogomasshtabnyj korrelyacionnyj analiz nestacionarnyh, soderzhashchih kvaziperiodicheskie uchastki signalov. Radiotekhnika i elektronika. 2008. T. 53. № 1. S. 73−85 (in Russian).
3. Rastrigin L.A. Adaptaciya slozhnyh sistem. Riga: Zinatne. 1981. 375 s.
4. Sachkov YU.V. Novaya filosofskaya enciklopediya / Pod red. V.S. Styopina. V 4-h tt. M.: Mysl'. 2001 (in Russian).
5. Dmitriev A.K., YUsupov R.M. Identifikaciya i tekhnicheskaya diagnostika: Ucheb. dlya vuzov. L.: MO SSSR. 1987.
6. Myshko V.V. Teoreticheskie osnovy i metody optimizacii analiza tekhnicheskogo sostoyaniya slozhnyh sistem. SPb: VKA im.  A.F. Mozhajskogo. 2013 (in Russian).
7. Homonenko A.D. Yakovlev E.L. K ocenke sostoyaniya slozhnyh ob"ektov s pomoshch'yu invariantov. Proceedings of 3rd Russian Conference «Mathematical Modeling and Information Technologies». 2016 (In Russian).
8. Butkovskij A.G. Strukturnaya teoriya raspredelennyh sistem. M.: Nauka. 1977 (In Russian).
9. Huan Huang, Baddour N. Bearing vibration data collected under time-varying rotational speed conditions. Data in Brief. 2018.
10. ZHuravlyov YU.I., Ryazanov V.V., Sen'ko O.V. Raspoznavanie. Matematicheskie metody. Programmnaya sistema. Prakticheskie primeneniya. M.: FAZIS. 2006. 176 s. (In Russian).
11. Braverman E.M., Muchnik I.B. Strukturnye metody obrabotki empiricheskih dannyh. M.: Nauka. 1983. 464 s. (in Russian).
12. Jurafsky D., Martin J.H. Speech and Language Processing: An Introduction to Natural Language Processing, Computational Linguistics and Speech Recognition. Prentice-Hall. 2008. 1024 p.
13. IMS bearings dataset. 2014. http://ti.arc.nasa.gov/tech/dash/pcoe/prognostic-data-repository.
14. Kuharenko B.G. Issledovanie po metodu Proni dinamiki sistem na osnove vremennyh ryadov. Trudy MFTI. 2009. T. 1. № 2 (in Russian).