Title :
The Method of Early Fault Identification in a Wind Turbine
Author :
Blagitko, B. ; Brigylevich, V. ; Jarmolovskyj, I.
Author_Institution :
Lviv Nat. Univ. of Ivan Franko
Abstract :
The method of early identification of the nonoperating wind electricity devices is suggested here. The main idea of this method consists in the assumption that the detection of a nonoperation is reflected by the detection of a periodic signal hidden in the noises with a long duration: practically its duration is longer than the duration of a measured succession and that is why one may consider this signal to be periodic with the constant amplitude and frequency. The principal work of the suggested system of early identification of the nonoperating devices consists in the usage of an operated Duffing oscillator. A small periodic signal hidden in the noise may be detected by Duffing oscillator, as a transition from the oscillation regime with a small amplitude to the regime with a high amplitude. Oscillator works as a highly sensitive frequency relay. For instance, the bearings damages have periodic character that is why the process of nonoperation may identify the detection of a periodic determined frequency signal in the informational signals. Vibroaccelerometers signals are used in the work, which are taken from the real equipments. Harmonic signal of a fixed frequency with different amplitude is used for nonoperation modeling to each of these signals. The aim of the research is to fix harmonic signal using at the correlation signal/noise smaller than 1. The results have shown the possible usage of the suggested method for the construction system of early debugging of nonoperating wind electricity aggregates at the signal-to-noise merit greater than -23 dB
Keywords :
machine bearings; oscillators; signal detection; wind turbines; correlation noise; correlation signal; early fault identification; frequency relay; frequency signal; harmonic signal; informational signals; nonoperating wind electricity devices; operated Duffing oscillator; periodic signal; vibroaccelerometers signals; wind turbine; Aggregates; Debugging; Fault diagnosis; Frequency measurement; Noise level; Noise measurement; Oscillators; Relays; Signal processing; Wind turbines;
Conference_Titel :
Mathematical Methods in Electromagnetic Theory, 2006 International Conference on
Conference_Location :
Kharkiv
Print_ISBN :
1-4244-0490-8
DOI :
10.1109/MMET.2006.1689782