Wind Turbine Periodic Intermittent Fault Detection Based on Fractional Order Chaotic System

For the wind turbine vibration signal periodic intermittent fault detection problem, we put forward a new detection method: first to extract the wind turbine vibration original signal fundamental frequency based on Morlet wavelet analysis, and then get a new set of signal baseband removed as the fractional terminable vibration system input. Whether the wind turbine has periodic intermittent faults or not is judged by whether the phase trajectory is in a terminable state and the fault signal amplitude can be detected. The system is highly sensitive to the periodic signal but insensitive to the random perturbation. The experiment proves that the method has good function of early fault prediction.