A. Yu. Spasenov, Assistant, Postgraduate Student, Department "Systems of the Automated Designing", e-mail: a.spasenov@mail.ru, K. V. Kucherov, Assistant, Postgraduate Student, Department "Computer Systems, Complexes and Networks", e-mail: Cvkucherov@yandex.ru, T. M. Volosatova, Ph.D., Associate Professor, Department "Systems of the Automated Designing", e-mail: tamaravol@gmail.com, D. M. Zhuk, Ph.D., Associate Professor, Department "Systems of the Automated Designing", e-mail: zhuk_d@mail.ru, Bauman Moscow State Technical University, Moscow, Russian Federation

Analysis of Quasi-Periodic Time Series by a Structural-Modal Method for Monitoring and Diagnostics of Complex Technical Systems

The possibilities of using the structural-modal method for monitoring and diagnostics the technical states of a complex technical systems are presented. The main idea of the proposed method is to combine the methods of parametric modal decomposition of signals and thematic modeling methods used for soft clustering of attributes of segments of time series. The initial data for analysis are often presented as a quasiperiodic time series. Each period of the time series is a set of modal parameters. The first stage of the proposed approach is time series segmentation. Second stage is a parametric modal decomposition of the processes occurring in the identified signal segments. The final stage is creation of a probabilistic portrait of each state of the system. The structural-modal method allows to identify the dynamics of state of systems changes in automatic or semi-automatic modes, depending on the amount of a priori information about the analyzed complex technical system. The proposed method can be used to solve the problem of diagnosing and monitoring the condition of complex technical system in various application areas. The proposed approach can be used to create special mathematical software aimed at the automatic analysis of the state of complex technical systems.
Keywords: time series analysis, technical state, modal analysis, thematic modeling

P. 563–569