DFIG defects diagnosis method for wind energy conversion chain

Abstract

This paper is an extension of research work originally presented in 2018 IEEE fifth International Congress on Information Science and Technology (CiSt). The research consists on developing method to diagnose electrical defects affecting wind turbine doubly-fed induction generator DFIG which constitutes a crucial part of wind energy conversion chain. First off all, we create a model of a non-defected wind conversion system based on mathematical equations introduced in Matlab Simulink. Then, we apply an indirect vector control stator field orientation in order to increase wind energy performance. With the aim of diagnosing the defects attacking wind turbine generator, we propose a method based on grouping of fast Fourier transform spectral analysis and Lissajous curves performed to generator stator and rotor currents. This diagnosis technique is applied to wind turbine in normal operation (non-defected generator) in order to have a reliable reference data for asynchronous generator behaviour. However, connected to the grid, wind turbine generator is affected by various faults occurring in electrical power networks. Therefore, the diagnosis method is applied also to a defected generator. Considering diversity of grid defects, we deal in the current paper with open stator supplying phases and open rotor feeding phases due to rotor side converter legs opening. Indeed, this diagnosis method allows diagnosing generator defects type and severity by comparing the resulting frequency spectrum analysis and Lissajous curves under abnormal condition operating to reference data obtained in case of non-defected generator. So, our proposed method contributes to DFIG defects identification and anticipation. The simulations had been accomplished using Matlab Simulink. These results proved the efficiency and effectiveness of the proposed DFIG diagnosis method for wind energy conversion chain.