A novel online diagnosis of brushless generator rotary rectifier fault

The fault detection of rotary rectifier based on harmonic analysis has some deficiencies. A new fault diagnosis method is presented using fractal theory and dynamics. Firstly, the quantitative description of exciter field currentpsilas complexity and irregularity is performed by box dimension calculation. Then the exciter field current fluctuation range under noisy environments is obtained by dynamics. Finally, the fault detection and identification of rotary rectifier is implemented through fractal dimension-signal fluctuation range trajectory area. The analysis results of testing waveforms show that box dimension and dynamics can analyze the waveform characteristics of exciterpsilas field current synthetically, the diagnosis effect is remarkable. Because the arithmetic has a little working load, it is more feasible to realize online fault diagnosis.