Анализ коэффициентов вейвлет-спектров диагностических сигналов при повреждении ротора асинхронного двигателя

Abstract

Предложен диагностический коэффициент, являющийся функцией среднего значения суммы коэффициентов вейвлет-преобразования сигнала электродвижущей силы в обмотках статора асинхронного двигателя (АД) для области средних частот. Разработан метод его декомпозиции с использованием теории обратного z-преобразования. Предложенный метод позволил уточнить диагностические признаки повреждений стержней ротора АД в режиме самовыбега двигателя и определить степень повреждения ротора.

Purpose. Specification of diagnostic signs of the broken rotor bars in electromotive force (EMF) signal of induction motor (IM) by the analysis of the totality of wavelet-coefficients values with the use of diagnostic coefficients decomposition. Materials and results of the research. During the research reverse z-transform was used. After decomposition EMF signals in the stator winding elements are obtained and their wavelet transform is performed. The analysis of wavelet-spectra of the obtained signals resulted in singling out of information signs of broken bars. However, when EMF signals decomposition method is used, it is difficult to determine the degree of rotor breakage. That is why in this paper decomposition of diagnostic coefficient is proposed to simplify the procedure of singling out information signs of the broken bars. A diagnostic coefficient, being the function of the average value of wavelet-decomposition coefficients sum for medium-frequency domain is proposed. A method for its decomposition with the use of the theory of reverse z-transform is worked out. The proposed method made it possible to specify diagnostic signs of IM brok en rotor bars in the motor self-running-out mode and determine the degree of the rotor breakage. The method of decomposition of IM stator winding phase EMF signal with the use of the theory of reverse z-transform allowed one to specify the reliability of diagnostics of IM broken bars due to singling out information signs existing in EMF signals of one active side of the coil. Practical value. The proposed method can be applied to decomposition of the diagnostic coefficient obtained as a result of wavelet-analysis of the phase EMF experimental signal measured in the mode of self-running-out of the motor. Application of this method makes it possible to specify the diagnostic signs of IM broken rotor barsand to determine the degree of breakage.