Application of wavelet synchrosqueezed transforms to the analysis of infrasound signals generated by wind turbines

Abstract

The issues reported in this article concern the development of methods applied for measurement, processing, and analysis of infrasound signals generated in association with the operation of wind farms. In particular, the discussion involves the results of the analysis using synchrosqueezed wavelet transforms of infrasound noise emitted by a 2 MW wind turbine that have been recorded during its operation in actual conditions. To record infrasound signals, a wireless measurement system was used, consisting of a base station and three synchronized mobile recording stations. To identify the wavelet structures with the highest ratio of energy, the synchrosqueezed wavelet transforms were used, and the courses of six time runs representing instantaneous frequencies were determined. Application of this approach enables the selection of energy-dominant waveforms from the time-frequency images, whose assessment can be performed mainly in terms of qualitative measures. Application of the synchrosqueezed wavelet transform is an effective tool for the purposes of detection and selection in the designated wavelet structures for the recorded infrasound dominant frequencies for which the carried energy ranges have the highest value.