ISSN 1991-2927
 

ACP № 1 (59) 2020

Author: "Anastasiia Valerevna Alekseeva"

/table>

Vladimir Nikolaevich Kliachkin, Doctor of Sciences in Engineering, Professor; graduated from the Faculty of Mechanics of Ulyanovsk Polytechnic Institute; Professor at the Department of Applied Mathematics and Informatics of Ulyanovsk State Technical University; an author of scientific publications in the field of reliability, statistical methods. e-mail: v_kl@mail.ruV.N. Kliachkin

Anastasiia Valerevna Alekseeva, Postgraduate Studentatthe Departmentof Applied Mathematics and Informatics of UlSTU; graduated from the Faculty of Information System and Technologies of UlSTU; a standardization engineer at the Ulyanovsk Design Bureau of Instrumentation, JSC; an author of scientific publications in the field of statistical methods of quality control. e-mail: age-89@mail.ruA.V. Alekseeva

Study of the efficiency of statistical control of hydro-unit vibrations59_3.pdf

To diagnose the technical condition of the hydraulic unit, vibration monitoring is carried out, as the level of vibration largely determines the quality of operation of the unit. When assessing the stability of vibrations, methods of statistical process control can be used. Many monitored indicators include both independent and correlated indicators. When monitoring correlated indicators, multivariate control methods are used. Mid-level monitoring of the process is based on the Hotelling algorithm. For the analysis of multivariate scattering, the generalized dispersion algorithm is used. The methodology and test results for analyzing the effectiveness of the generalized dispersion algorithm for controlling multidimensional vibration scattering are considered. Regression dependences of the average run length on the characteristics of the process disturbance were constructed, on the basis of which the quality of vibration diagnostics can be estimated. Initially, a lot of samples are constructed that are identical to the studied process of vibration of the hydraulic unit, that is, with a vector of average and covariance matrix corresponding to the training sample obtained in the real process. Based on the results of statistical tests, regression dependences of the average length of the series on the characteristics of the process violation were obtained, on the basis of which the quality of vibration diagnostics can be estimated.

Vibration stability, multivariate scattering, generalized dispersion, average run length.

2020_ 1

Sections: Mathematical modeling

Subjects: Mathematical modeling.

© FRPC JSC 'RPA 'Mars', 2009-2018 The web-site runs on Joomla!