Noise Source Localization Using a Compact Phased Array: Studies on a Full Scale Wind Turbine in a Wind Farm

Abstract

Locating the dominant noise sources on a wind turbine is an important problem in designing and developing low noise wind turbines. Previously very large microphone arrays were used to locate these sources. The primary focus of this paper is to show that using a compact and mobile microphone array with advanced beamforming algorithms, the noise sources can be successfully located and quantified. The results from the qualification experiments on the microphone array conducted in laboratory using synthetic noise sources show the differences between the various beamforming algorithms used in this study (both frequency and time domain algorithms). The initial experimental results on a full scale wind turbine reveal that it is indeed possible to locate the noise sources using a compact microphone array by successfully locating the two dominant noise sources on the wind turbine namely, aerodynamic noise near the blade tip and mechanical noise from nacelle.

Research status on aero-acoustic noise from wind turbine blades

“…TBL-TE noise is commonly considered to be the dominant noise source for modern large wind turbines [7,8,10]. Sometimes, inflow-turbulence noise may be prominent for some specific wind turbines [11,12].…”

“…It has many applications such as for high-speed train [37], flying aircraft [38 -39] and wind turbine [7,8,10,12,40,41], etc.. A microphone array is composed of a number of microphones, the phase and amplitude of acoustic signal received by each microphone may be different due to their different distance from noise sources. Then the difference of phase and amplitude can be employed by some post processing algorithm to locate the positions of noise sources.…”

Advances in Experimental Aeroacoustics

“…21, the peak source location is determined i.e; the scan point for which P j (i-1) obtains its maximum value P max (i-1) . Degraded source powers P j (i) without the influence of this peak source are written as (23) To find the G (i) , it is required that the source cross powers of any scan point with the peak location are determined entirely by G (i) . In other words…”

“…The Rayleigh limit was also tested experimentally using synthetic sources: experimental results reaffirmed the theoretical findings. Detailed results of the Rayleigh criterion are discussed in Ramachandran et al [23]. The first decision make before making the measurements was the location of the microphone array.…”

“…A comparison between the theoretical prediction and experimental result of the peak aerodynamic noise from the blade tip over one revolution. (reprinted from Ramachandran et al[23], 2012 Multi-Science Publishing Co. Ltd., used with permission). aerodynamic noise on the wind turbine blades as multiple peaks distributed along the length of the blades.…”