Machine Learning-Based Tools for Wind Turbine Acoustic Monitoring

The identification and separation of sound sources has always been a difficult problem for acoustic technicians to tackle. This is due the considerable complexity that made up many contributions at different frequencies. Each specific frequency spectrum, but when sounds overlap it becomes discriminate between contributions. In this case, can be extremely useful have tool capable identifying operating conditions an source. study, measurements were noise emitted by wind turbine in vicinity sensitive receptor. To identify turbine, average spectral levels one-third octave bands used. A model based on support vector machine (SVM) was developed detection turbine; then artificial neural network used compare performance both models. high precision returned simulation models supports adoption these tools as characterization environments close turbines.