ОЦЕНКА ТЕХНИЧЕСКОГО СОСТОЯНИЯ ЦЕНТРОБЕЖНОГО НАСОСНОГО АГРЕГАТА НА ОСНОВЕ МУЛЬТИФРАКТАЛЬНОЙ ПАРАМЕТРИЗАЦИИ ВЕЙВЛЕТ-КАРТИН

Abstract

Facility security of centrifugal pumping units have the great importance in process of continuous production at oil refining and end product transfer. In conditions of continuous production, where frequent stoppages of production equipment are not admissible, there is a necessity of using methods of non-destructive control. These methods allow to reveal latent defects of different parts of machines and to control their progress, which provides facility security and increases the service life of machines. Current state of vibration diagnostics services shows that the majority of instruments and systems, applied by vibration diagnostics services, is based on fast Fourier transform (spectral analysis) of an input signal. This method is applied due to its simplicity and universal use. To use existing spectral Fourier analysis methods is not enough for definite identification of defects therefore we must elaborate additional methods for primary analysis of vibration signal. Antecedent investigations proved that wavelet analysis can be used for rotary machines diagnostics on the example of revealing centrifugal compressor unit defects. Proved wavelet analysis perspective use needs fulfilling further investigations in relation quantitative assessment, performed with help of wavelet pattern multifractal parametrization. The examples of realization of the proposed mathematical processing of artificial and real signals on MATLAB software environment and MFRDROM multifractal parametrization program is represented. This paper proposes wavelet analysis, wavelet pattern multifractal parametrization as an additional method for monitoring and evaluation of machine technical state which tends to be mutually complementary with existing practical methods in order to increase exactness of identifying rotary equipment defects.