Abnormality determination method for wind power generation device, abnormality determination system for wind power generation device, and abnormality determination program for wind power generation device

An abnormality determination system for a wind power generation device, comprising:

• an acoustic detector that measures sound emitted by the wind power generation device and records acoustic data;
• a computer processor that:
• i. creates, based on the recorded acoustic data, deviation data from a time-averaged spectrum of the recorded acoustic data, and performs a spectrogram analysis on the deviation data, on a frequency axis and in a temporal axis space as a temporal change in a frequency characteristic by using the short-time Fourier transform or the wavelet transform;
• ii. detects, from the analysis result of the analysis means, only a signal component emitted from an abnormal portion of the wind power generation device in a time corresponding to rotation of the wind power generation device;
• and
• iii. determines that the wind power generation device is abnormal when the signal component detected by the detection means is greater than or equal to a certain threshold value, and ceases the acoustic detector from measuring sound emitted by the wind power when in the abnormal state.

A computer program stored in non-transitory medium for an abnormality determination device in an abnormality determination system for a wind power generation device, the computer program executable on a computer to execute:

• measuring sound emitted by the wind power generation device as acoustic data using an acoustic detector and recording the acoustic data in computer memory;
• creating, based on the acoustic data recorded using a computer processor, deviation data from a time-averaged spectrum of the recorded acoustic data, and performing a spectrogram analysis on the deviation data, on a frequency axis and in a temporal axis space as a temporal change in a frequency characteristic by using the short-time Fourier transform or the wavelet transform;
• detecting using the computer processor, from the analysis result in the analysis step, only a signal component emitted from an abnormal portion of the wind power generation device in a time corresponding to rotation of the wind power generation device;
• and
• determining, using the computer processor, that the wind power generation device is abnormal when the signal component detected is greater than or equal to a certain threshold value, and ceases the acoustic detector from measuring sound emitted by the wind power when in the abnormal state.