Tonality evaluation of wind turbine noise by filter-segmentation

Wind turbine (WT) noise is one of the major hindrances for the widespread use of wind energy. Noise tone can greatly increase the annoyance factor and the negative impact on human health. The national standard for measuring WT noise tonality IEC 61400-11 does not ensure consistency and accuracy in the tonality analysis of strongly non-stationary noise. The aim of this study is to explore a new tonality evaluating methodology that is based on noise separation and stationary processing. An effective Gabor filtering method is proposed to remove background noise embedded in the sound signal of operating wind turbine. One-minute wind-turbine sound signal without background noise is optimally divided into relatively stationary sub-segments, by making full use of the close relationship between the wind-turbine rotor speed and the tonal noise emitted from WT. The proposed method’s theoretical advantages over conventional evaluation in IEC are illustrated. Results of experiments for GOLDWIND 77/1500 WT are presented, which show the capability of the proposed methodology.

► The method is proposed to overcome the drawbacks of IEC method for WT tonality.
► Order components extraction is used to extract signals generated by wind turbine.
► Rotary speed based optimal segmenting is used to suppress “frequency ambiguity” of FFT.
► Results verify that the tones of different groups are unified into the same scale.